C S I R O L A N D a nd WAT E R
Quantifying Submarine Groundwater Discharge
and Nutrient Discharge into Cockburn Sound,
Western Australia
Anthony J. Smith, Jeffrey V. Turner, David E. Herne and Wayne P. Hick
CSIRO Land and Water, Perth
Technical Report 1/03, June 2003
A Technical Report to Coast and Clean Seas Project WA9911:
Quantifying Submarine Groundwater Discharge and Demonstrating
Innovative Clean-Up to Protect Cockburn Sound from Nutrient Discharge
Quantifying Submarine Groundwater Discharge and Nutrient Discharge into Cockburn Sound, Western Australia
A Technical Report to Coast and Clean Seas Project WA9911: Quantifying Submarine Groundwater Discharge and Demonstrating Innovative Clean-Up to Protect Cockburn Sound from Nutrient Discharge Anthony J. Smith, Jeffrey V. Turner, David E. Herne and Wayne P. Hick CSIRO Land and Water, Wembley, Western Australia 6913 Technical Report No. 01/03 ISSN 1446-6163 January 2003 ISBN No. 0 643 061134 CGS Report No. 104
© CSIRO Australia 2003
To the extent permitted by law, all rights are reserved and no part of this publication (including
photographs, diagrams, figures and maps) covered by copyright may be reproduced or copied in any
form or by any means except with the written permission of CSIRO Land and Water.
Important Disclaimer:
CSIRO Land and Water advises that the information contained in this publication comprises general
statements based on scientific research. The reader is advised and needs to be aware that such
information may be incomplete or unable to be used in any specific situation. No reliance or actions
must therefore be made on that information without seeking prior expert professional, scientific and
technical advice. To the extent permitted by law, CSIRO Land and Water (including its employees
and consultants) excludes all liability to any person for any consequences, including but not limited to
all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from
using this publication (in part or in whole) and any information or material contained in it.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge i and Nutrient Discharge into Cockburn Sound
TABLE OF CONTENTS
Table of Contents...........................................................................................................i
List of Tables ...............................................................................................................iv
List of Figures ..............................................................................................................vi
Executive Summary ......................................................................................................xi
Acknowledgements.....................................................................................................xx
1 Introduction ............................................................................................................1
1.1 Background.....................................................................................................1
1.2 This Study.......................................................................................................2
1.3 Report Format.................................................................................................2
2 Hydrogeology.........................................................................................................5
2.1 Regional Setting .............................................................................................5
2.2 Subcrop Formations......................................................................................10
2.2.1 Kardinya Shale Member........................................................................10
2.2.2 Pinjar and Waneroo Members...............................................................10
2.2.3 Rockingham Sand .................................................................................10
2.3 Superficial Formations .................................................................................14
2.3.1 Tamala Limestone.................................................................................14
2.3.2 Cooloongup Sand..................................................................................14
2.3.3 Becher Sand ..........................................................................................15
2.3.4 Safety Bay Sand ....................................................................................15
2.4 Groundwater Occurrence and Flow..............................................................15
2.4.1 Artesian Aquifer System .......................................................................15
2.4.2 Rockingham Aquifer.............................................................................16
2.4.3 Superficial Aquifer................................................................................16
2.5 Groundwater Bores.......................................................................................19
2.6 Groundwater Levels......................................................................................26
2.6.1 Regional Groundwater Table ................................................................26
2.6.2 Influence of Sea Level Variation...........................................................26
2.6.3 Influence of High-permeability Limestone ...........................................27
2.7 Seawater Intrusion ........................................................................................31
3 Field Investigations ..............................................................................................37
3.1 Northern Harbour..........................................................................................37
3.1.1 Hydrogeology........................................................................................39
3.1.2 Groundwater Levels ..............................................................................40
3.1.3 EC Surveys of Submarine Porewater ....................................................40
CSIRO Land and Water
ii Coast and Clean Seas Project No. WA9911
3.1.4 Groundwater Chemistry and Isotopic Composition..............................43
3.1.5 Porewater Chemistry.............................................................................43
3.1.6 Benthic Flux Measurements .................................................................50
3.1.7 Continuous Radon Monitoring in Northern Harbour............................57
3.1.8 Seismic Survey......................................................................................57
3.1.9 Groundwater Temperature and EC Profiling ........................................62
3.2 Challenger Beach..........................................................................................67
3.2.1 Hydrogeology........................................................................................67
3.2.2 EC Surveys of Submarine Porewater....................................................69
3.2.3 Porewater Chemistry.............................................................................70
3.3 BP Refinery ..................................................................................................73
3.3.1 Hydrogeology........................................................................................73
3.3.2 EC Surveys Submarine Porewater ........................................................76
3.3.3 Porewater Chemistry.............................................................................76
3.3.4 Benthic Flux Measurements .................................................................78
3.4 Cockburn Sound Submarine Porewater Survey ...........................................81
3.4.1 Ionic Chemistry.....................................................................................81
3.4.2 Calculations Of Groundwater Nutrient Concentrations........................81
3.4.3 Synthesis Of Results .............................................................................82
4 Quantifying SGD into Cockburn Sound ..............................................................95
4.1 Conceptual Models of SGD .........................................................................95
4.2 Previous Estimates of SGD ........................................................................100
4.2.1 Flownet Calculations ..........................................................................101
4.2.2 CTD Profiling .....................................................................................102
4.3 Direct Measurement of SGD......................................................................102
4.4 Estimation of SGD by Inshore Water Balance ...........................................103
4.4.1 Groundwater Flow Model...................................................................103
4.4.2 Groundwater Recharge .......................................................................105
4.4.3 Calibration ..........................................................................................105
4.4.4 Results.................................................................................................106
4.4.5 Conclusions.........................................................................................108
4.5 Estimation of SGD Based On Saltwater-Freshwater Relations..................115
4.6 Summary.....................................................................................................117
5 Quantifying Nutrient Discharge into Cockburn Sound......................................119
5.1 Mechanisms for Nutrient Discharge...........................................................119
5.2 Nitrogen Concentration in Groundwater ....................................................120
5.3 Significance Of Measured Concentrations .................................................123
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge iii and Nutrient Discharge into Cockburn Sound
5.4 Previous Estimates Of Nutrient Discharge .................................................125
5.5 Estimation Of Nutrient Discharge Based On SGD Rates From The Inshore
Water Balance Model And Measured Nutrient Concentrations In Submarine
Porewater ....................................................................................................127
5.5.1 Method ................................................................................................127
5.5.2 Results.................................................................................................128
5.6 Estimation Of Nutrient Discharge Based On SGD Rates From The Inshore
Water Balance Model And Groundwater Nutrient Concentrations Measured
By Appleyard (1994) ..................................................................................135
5.6.1 Method ................................................................................................135
5.6.2 Results.................................................................................................136
5.7 Synthesis of Results....................................................................................141
5.7.1 Uncertainty..........................................................................................144
5.7.2 Composite Estimate ............................................................................146
References.................................................................................................................149
Appendix A: Aquifer Tidal Propagation at Northern Harbour.................................153
The Tidal Method...............................................................................................154
Fourier Analysis .................................................................................................156
Discrete Fast Fourier Transform (FFT)..............................................................157
Fremantle Tide Data...........................................................................................157
Fremantle Tide Data...........................................................................................158
Aquifer Tidal Propagation..................................................................................168
Appendix B: Benthic Flux Meters ............................................................................173
Lee-type..............................................................................................................174
Ultrasonic ...........................................................................................................174
Heat-pulse...........................................................................................................175
Appendix C: Models of Tide and Wave Affects on Benthic Flux............................179
Effect of Sea Level Dynamics ............................................................................180
Tide-Forced Model.............................................................................................181
Wave-Forced Model...........................................................................................182
CSIRO Land and Water
iv Coast and Clean Seas Project No. WA9911
LIST OF TABLES
Table 1: Estimated hydraulic properties of the Superficial Formations inshore
from Cockburn Sound ..............................................................................18
Table 2: Relation between aquifer thickness and water table elevation at the
position of the saltwedge toe based on the Ghyben-Herzberg
approximation...........................................................................................32
Table 3: Electrical conductivity of groundwater samples from saltwater
monitoring bores.......................................................................................33
Table 4: Completion details of intercomparison monitoring bores ............................39
Table 5: Bulk-ground EC surveys by Steiglitz (2001b); measurement times and
ocean water levels.....................................................................................41
Table 6: Bulk-ground EC surveys by CLW; measurement locations and times ........42
Table 7: Groundwater chemistry and Isotopes ...........................................................44
Table 8: Northern Harbour porewater chemistry........................................................45
Table 9: Benthic flux measurements; SCOR SGD intercomparison..........................51
Table 10: Bulk-ground EC surveys; measurement locations and times .....................69
Table 11: Challenger Beach porewater chemistry......................................................71
Table 12: Bulk-ground EC surveys; measurement locations and times .....................76
Table 13: BP Refinery porewater chemistry...............................................................77
Table 14: Porewater chemistry ...................................................................................85
Table 15: Estimates of Submarine Groundwater Discharge into Cockburn
Sound and related sites ...........................................................................100
Table 16: Cockburn Groundwater Area model: conductance and external head
values for the general head coastal boundary cells .................................105
Table 17: Cockburn Groundwater Area model: calibrated hydraulic
conductivity values .................................................................................106
Table 18: Cockburn Groundwater Area model: model water balance for low-
recharge scenario ....................................................................................107
Table 19: Cockburn Groundwater Area model: model water balance for high-
recharge scenario ....................................................................................107
Table 20: Summary of nutrient concentrations in groundwater in Cockburn
Sound area and related sites....................................................................121
Table 21: Default trigger values for physical and chemical stressors for south-
west Australia for slightly disturbed inshore marine ecosystems
(ANZECC, 2000);...................................................................................123
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge v and Nutrient Discharge into Cockburn Sound
Table 22: Receiving water quality protection criteria for discharge of
groundwater from aquifer dewatering (WRC, 1999)..............................124
Table 23: Estimates of nutrient discharge into Cockburn Sound via SGD...............126
Table 24: Nitrogen loads for nominated sections of coastline..................................128
Table 25: Nitrogen loads for nominated sections of coastline..................................136
Table 26: Estimates of total nitrogen discharge into Cockburn Sound via
groundwater, and equivalent average values of SGD and nitrogen
concentrations in groundwater ................................................................143
Table 27: Composite estimate of total nitrogen discharge into Cockburn Sound
via groundwater ......................................................................................147
Table 28: Summary of tidal analysis for K1 tidal constituent ..................................170
Table 29: Summary of tidal analysis for O1 tidal constituent ..................................170
CSIRO Land and Water
vi Coast and Clean Seas Project No. WA9911
LIST OF FIGURES
Figure 1: Study location map........................................................................................3
Figure 2: Regional Geomorphology .............................................................................6
Figure 3: Groundwater Flow Systems of the Superficial Aquifer ................................7
Figure 4: Subcrop Formations beneath the Superficial Aquifer ...................................8
Figure 5: Coastal Surface Geology...............................................................................9
Figure 6: Hydrological cross section (Section A) through the northern ends of
Garden Island and Cockburn Sound .........................................................11
Figure 7: Hydrological cross section (Section B) through the southern end of
Garden Island and James Point.................................................................12
Figure 8: Hydrological cross section (Section C) through the southern ends of
Garden Island and Cockburn Sound .........................................................13
Figure 9: Groundwater bore locations ........................................................................21
Figure 10: Combined thickness of Quaternary and Tertiary sediments (m);..............22
Figure 11: Thickness of Tamala Limestone (m);........................................................23
Figure 12: Thickness of basal clay (m);......................................................................24
Figure 13: Thickness of Quaternary sands (m);..........................................................25
Figure 14: Water table map of the Superficial Aquifer inshore from Cockburn
Sound based on 1998 water level data......................................................29
Figure 15: Tidal fluctuations in groundwater levels inshore from Northern
Harbour from November 2000 to June 2001; sampling interval for
water levels is 1 hour ................................................................................30
Figure 16: Hydrological cross section (Figure 5, Section D) through the
Northern Harbour study site depicting two possible types of
interaction between the coastal saltwedge and saline water in Lake
Coogee; .....................................................................................................35
Figure 17: Northern Harbour (SGD intercomparison) study site ...............................38
Figure 18: Bulk-ground EC measurements by Steiglitz (reproduced from:
Steiglitz, 2001b); ......................................................................................46
Figure 19: R-Transect; bulk-ground EC measurements (reproduced from:
Steiglitz, 2001b)........................................................................................47
Figure 20: E-Transect; bulk-ground EC measurements (reproduced from:
Steiglitz, 2001b)........................................................................................48
Figure 21: B-Transect; bulk-ground EC measurements (reproduced from:
Steiglitz, 2001b)........................................................................................48
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge vii and Nutrient Discharge into Cockburn Sound
Figure 22: Bulk-ground EC measurements; additional Northern Harbour
transects; ...................................................................................................49
Figure 23: Site map of shore-based experimental design during the SGD
intercomparison; (reproduced from: Krupa et al., 2000) ..........................53
Figure 24: Benthic flux measurements recorded during the SGD
intercomparison; East Transect, station E1...............................................54
Figure 25: Benthic flux measurements recorded during the SGD
intercomparison; East Transect, station E2...............................................54
Figure 26: Example heat pulse data from Northern Harbour indicating seawater
flow into sediment; benthic flux from T2-T3 pair ≈ - 36 cm.d-1,
benthic flux from T1-T3 pair ≈ - 21 cm.d-1 ..............................................55
Figure 27: Example heat pulse data from Northern Harbour indicating SGD;
benthic flux from T4-T5 pair ≈ 20 cm.d-1, benthic flux from T4-T6
pair ≈ 9 cm.d-1 ...........................................................................................55
Figure 28: Benthic flux measurements from Northern Harbour on 8/5/01 from
11:24 AM to 5:16 PM; measurements are every two minutes..................56
Figure 29: Seismic tracks in Northern Harbour and Cockburn Sound
(reproduced from: Steiglitz, 2001b)..........................................................59
Figure 30: Northern Harbour seismic features (reproduced from: Stieglitz,
2001a) .......................................................................................................60
Figure 31: Cockburn Sound seismic features (reproduced from: Stieglitz,
2001a) .......................................................................................................61
Figure 32: Groundwater temperature profiles on 29/2/00 in selected Northern
Harbour monitoring bores; profiles are indicated over the full
depths of bores ..........................................................................................63
Figure 33: Groundwater EC profiles on 29/2/00 in selected Northern Harbour
monitoring bores; profiles are indicated over the screened sections
of bores only .............................................................................................64
Figure 34: Groundwater temperature profiles on 28/11/00 in selected Northern
Harbour monitoring bores; profiles are indicated over the full
depths of bores ..........................................................................................65
Figure 35: Groundwater EC profiles on 28/11/00 in selected Northern Harbour
monitoring bores; profiles are indicated over the screened sections
of bores only .............................................................................................66
Figure 36: Challenger Beach study site ......................................................................68
Figure 37: Bulk-ground EC measurements; Challenger Beach transects; ..................72
Figure 38: BP Refinery study site ...............................................................................75
CSIRO Land and Water
viii Coast and Clean Seas Project No. WA9911
Figure 39: Bulk-ground EC measurements; BP Refinery transects; ..........................79
Figure 40: Benthic flux measurements from BP Refinery on 12/5/01 from
8:56 AM to 11:10 AM; measurements are every two minutes.................80
Figure 41: Porewater sampling locations ...................................................................89
Figure 42: Porewater ionic chemistry.........................................................................90
Figure 43: Porewater chloride-EC relation.................................................................90
Figure 44: Porewater dilution index ...........................................................................91
Figure 45: Groundwater nitrogen concentrations [mg.l-1] estimated from
submarine porewater analyses ..................................................................92
Figure 46: Groundwater phosphorus concentrations [mg.l-1] estimated from
submarine porewater analyses ..................................................................93
Figure 47: Conceptual models of SGD into Cockburn Sound ...................................98
Figure 48: Cockburn Groundwater Area flow model; finite difference grid,
boundary conditions and hydraulic conductivity zones..........................109
Figure 49: Cockburn Groundwater Area flow model; base of Superficial
Formations ..............................................................................................110
Figure 50: Cockburn Groundwater Area flow model; groundwater abstraction
and artificial recharge locations and rates...............................................111
Figure 51: Cockburn Groundwater Area flow model; groundwater recharge
rates for the low-recharge scenario, recharge is expressed as a
percentage of mean annual rainfall of 0.87 m/yr ....................................112
Figure 52: Cockburn Groundwater Area flow model; groundwater recharge
rates for the high-recharge scenario, recharge is expressed as a
percentage of mean annual rainfall of 0.87 m/yr ....................................113
Figure 53: Cockburn Groundwater Area flow model; calibrated water table
elevations and model computed SGD for the low-recharge and
high-recharge scenarios ..........................................................................114
Figure 54: Abrupt saltwater interface profiles from the solution of Glover
(1959)......................................................................................................116
Figure 55: Last-recorded, total nitrogen concentrations in groundwater [mg.l-1]; ...122
Figure 56: Model simulated SGD for low-recharge scenario...................................129
Figure 57: Model simulated SGD for high-recharge scenario..................................130
Figure 58: Estimated nitrogen load into Cockburn Sound based on model
simulated SGD (low-recharge scenario) and groundwater nutrient
concentrations estimates from porewater sampling;...............................131
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge ix and Nutrient Discharge into Cockburn Sound
Figure 59: Estimated nitrogen load into Cockburn Sound based on model
simulated SGD (high-recharge scenario) and groundwater nutrient
concentrations estimates from porewater sampling;...............................132
Figure 60: Estimated phosphorus load into Cockburn Sound based on model
simulated SGD (low-recharge scenario) and groundwater nutrient
concentrations estimates from porewater sampling;...............................133
Figure 61: Estimated phosphorus load into Cockburn Sound based on model
simulated SGD (high-recharge scenario) and groundwater nutrient
concentrations estimates from porewater sampling;...............................134
Figure 62: Estimated nitrogen load into Cockburn Sound based on model
simulated SGD (low-recharge scenario) and groundwater nutrient
concentrations measured by Appleyard (1994).......................................137
Figure 63: Estimated nitrogen load into Cockburn Sound based on model
simulated SGD (high-recharge scenario) and groundwater nutrient
concentrations measured by Appleyard (1994).......................................138
Figure 64: Estimated phosphorus load into Cockburn Sound based on model
simulated SGD (low-recharge scenario) and groundwater nutrient
concentrations measured by Appleyard (1994).......................................139
Figure 65: Estimated phosphorus load into Cockburn Sound based on model
simulated SGD (high-recharge scenario) and groundwater nutrient
concentrations measured by Appleyard (1994).......................................140
Figure 66: Estimates of total nitrogen discharge into Cockburn Sound via
groundwater ............................................................................................142
Figure 67: Estimates of total nitrogen discharge into Cockburn Sound via
groundwater with data plotted according to equivalent average
values of SGDs and nitrogen concentrations in groundwater.................143
Figure 68: Tidal data and Power Spectral Density for Fremantle for the period
1/9/2000 to 16/1/2001; sampling interval is 10 minutes.........................160
Figure 69: Cockburn Sound tidal data and ‘repaired’ Cockburn Sound tidal
data for the period 25/11/2000 to 3/1/2001; sample interval is 10
minutes....................................................................................................161
Figure 70: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH1a for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................162
Figure 71: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH2a for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................163
CSIRO Land and Water
x Coast and Clean Seas Project No. WA9911
Figure 72: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH2b for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................164
Figure 73: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH3 for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................165
Figure 74: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH4 for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................166
Figure 75: Hydrograph and Power Spectral Density for groundwater
monitoring bore NH5 for the period 25/11/2000 to 27/12/2000;
sampling interval is 10 minutes ..............................................................167
Figure 76: Non-dimensional plot of normalised lag verses tidal efficiency for
groundwater monitoring bores NH1 to NH5 ..........................................171
Figure 77: Heat-pulse BFM ......................................................................................176
Figure 78: Example heat-pulse data from laboratory calibration; flow rate =
2.8 mm.s-1 ...............................................................................................177
Figure 79: Calibration curve for the heat-pulse BFM using inner thermistor
pairs T2-T3 and T4-T5 ...........................................................................177
Figure 80: Calibration curve for the heat-pulse BFM using outer thermistor
pairs T1-T3 and T4-T6 ...........................................................................178
Figure 81: Tide-pumping model; amplitudes of tidal groundwater fluxes across
the coastal boundary calculated for typical values of tidal range and
aquifer properties ....................................................................................184
Figure 82: Wave-pumping model; amplitudes of wave-induced groundwater
fluxes across the seabed calculated for typical values of wave height
and aquifer properties .............................................................................185
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xi and Nutrient Discharge into Cockburn Sound
EXECUTIVE SUMMARY
The Study
During the last ten years, an apparent divergence between estimates of nutrient
inputs into Cockburn Sound and trends in marine water quality has been observed.
Chlorophyll-a levels in Cockburn Sound have only improved slightly during this time
despite an estimated 70% reduction in the total anthropogenic nitrogen load into the
sound.
D.A Lord & Associates (2001) estimated that approximately 70% of the
anthropogenic nitrogen load into Cockburn Sound was contributed by submarine
groundwater discharge (SGD); however, the estimates of groundwater inputs were
subject to large errors. To assess management options for improving the water
quality in Cockburn Sound, improved estimates of SGD rates and associated nutrient
loads are required.
This study involved an integrated approach, combining a range of measurements
and techniques, to estimate groundwater nutrient loadings and identify major
nutrient inputs into Cockburn Sound. The investigation was conducted by CSIRO
Land and Water as part of the Coast & Cleans Seas Project WA9911, which involved
two components:
1. Mapping and quantifying submarine groundwater discharge and nutrient
discharge into Cockburn Sound, and
2. At one “hot spot”, demonstrating innovative clean up of a groundwater
nutrient plume using a biologically active, in situ treatment wall.
This report is a full account of the activities that were undertaken to accomplish
component one of the study. Component two activities are reported separately by
Patterson et al. (see references).
Hydrogeology of Cockburn Sound
The Perth Basin contains up to 12,000 meters of marine and continental sediments.
The stratigraphic sequence to a depth of around 2,000 metres below the present land
surface elevation contains Jurassic and Cretaceous age sediments that are overlain
by a relatively thin veneer of late Tertiary to Quaternary age sediments. These
CSIRO Land and Water
xii Coast and Clean Seas Project No. WA9911
upper-most sediments are known, collectively, as the superficial formations. The
present study is focused on estimation of SGD from these formations.
The saturated extent of the superficial formations forms a mostly unconfined aquifer
system called the Superficial Aquifer. Recharge to the aquifer occurs by percolation
of rainfall through well-drained, sandy soils. There is some leakage of groundwater
to the confined aquifer system beneath but most unutilised groundwater discharges
eventually to the Indian Ocean, Swan-Canning River estuary system and other
regional surface drains. Diffuse replenishment and lateral drainage of shallow
groundwater results in a number of distinct flow systems that can be recognised as
mounds in the shallow groundwater table.
The Cockburn Sound study area lies mostly on the western edge of the Jandakot
Mound and partly within the Safety Bay Mound. The superficial formations in this
area have a saturated thickness of around 25 to 30 metres and unconformably
overlie an erosional surface of Cretaceous sediments that act as a low-permeability
base to the Superficial Aquifer. Groundwater in the Superficial Aquifer flows
generally in a westerly direction and discharges to the near shore marine
environment along the coastline of Cockburn Sound. There are virtually no surface
drains because the surface sands are permeable enough to prevent significant
surface runoff.
The coastal strip of aquifer is characterised by very high transmissivity and very
small hydraulic gradients due to secondary porosity in the Tamala Limestone. In
general, groundwater in the upper sand aquifer moves downward into the
underlying, high-permeability limestone, which acts as a regional drainage layer
and conducts groundwater toward the coast. Locally, groundwater in the Tamala
Limestone can be confined by a clay layer at the base of the overlying sands, which
restricts drainage of the sands at these locations.
Site Investigations
Field investigations were concentrated at three locations along the Cockburn Sound
coast:
Northern Harbour
Challenger Beach
BP Refinery
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xiii and Nutrient Discharge into Cockburn Sound
Methodologies for SGD estimation included physical, geochemical, isotopic and
geophysical approaches.
Results from four type of benthic flux (seepage) meter revealed that benthic flux
across the seabed was highly variable and strongly influenced by sea level
fluctuations at time scales ranging from seconds to months. Results indicated that
terrestrial SGD across the seabed, and tide-forced and wave-forced benthic fluxes,
were of a similar order of magnitude and could not be effectively distinguished from
one another. To attain an estimate of the terrestrial component of SGD would
require a long time series of benthic flux measurements that span the broad
frequency spectrum of wave and tide phenomena. Due to transient effects of sea
level change and the patchy distribution of SGD, it was concluded that benthic flux
meters were not suited to obtaining regional scale estimates of groundwater
discharge.
Measurements of electrical conductivity (EC) of submarine porewater revealed
different patterns of groundwater discharge at different locations along the
coastline. A narrow shoreline zone of SGD—as indicated by fresher porewater—
was typical at most locations. Fresh porewater was also detected in patches further
from the shore (the maximum distance from shore of sampling was approximately
56 m). Porewater composition typically varied from that of pure seawater to
relatively fresh groundwater over distances across the seabed of only several meters.
It seemed likely that groundwater was also discharging further offshore but beyond
the limits of the EC transects.
Continuous monitoring of the seawater radon inventory in Northern Harbour
revealed a quasi-inverse relation between diurnal tides and groundwater discharge
rates estimated by this method. This relation was consistent with diurnal
groundwater discharge rates measured by the benthic flux meters.
Seismic surveys in Northern Harbour and Cockburn Sound revealed a number of
stationary water-column features that were consistent with fluid migration from the
seabed. It was beyond the scope of the investigation to determine whether the
observed density anomalies in the water column were due to fresh groundwater
discharge, gas migration from seabed sediment, or another cause. Nevertheless, the
results were highly suggestive of focused SGD, possible through offshore solution
features in the coastal limestone. Although seismic techniques cannot be used to
CSIRO Land and Water
xiv Coast and Clean Seas Project No. WA9911
calculate discharge rates, they have been used successfully at other locations in
Australia to map the occurrence of focused SGD.
Submarine Porewater Survey
The above investigations were complimented by an onshore (groundwater) and
offshore (porewater) survey of nutrient concentrations along the Cockburn Sound
foreshore between Woodman Point and Palm Beach in Rockingham.
Approximately 20 groundwater samples were collected near Northern Harbour and
80 porewater samples were collected between Woodman Point and Palm Beach
during the period September 2001 to June 2002. Porewater and Groundwater
samples indicated N concentrations up to 130 mg.l-1. Porewater samples were
obtained from depths of approximately 0.3-0.5 m below the seabed and within 10
metres of the shoreline using an EC probe to locate the freshest porewater and a
spear probe to extract the sample. All samples were analysed for nutrient
concentrations, chloride concentration, electrical conductivity and pH.
Ionic balances on 18 samples indicated that the dominant porewater chemistry was
that of seawater diluted by various portions of fresh groundwater. The portion of
fresh groundwater contributing to a porewater sample was estimated based on the
chloride concentration of the sample. The approximate nutrient concentration in
groundwater was then determined by assuming all nutrients in the porewater
mixture were derived from the groundwater component.
High nutrient concentrations in submarine porewater detected along the shoreline
south of James Point were consistent with results from previous groundwater
investigations, which revealed high levels of nutrient contamination in this area.
Total nitrogen concentrations greater than 100 mg.l-1 were measured in porewater
less than 0.5-m below the seabed and greater than 50-m offshore. Total nitrogen
concentrations in groundwater were measured at 200 mg.l-1.
High nitrogen concentrations were also detected along a 40-m section of the
shoreline north of James Point and adjacent the HIsmelt facilities. Total nitrogen
concentrations in groundwater of up to 740 mg.l-1 were estimated based on
submarine porewater concentrations of up to 220 mg.l-1. The source of
contamination is unknown; however, there is evidence from groundwater monitoring
of significantly elevated nitrate levels in the inshore aquifer.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xv and Nutrient Discharge into Cockburn Sound
Slightly elevated nutrient concentrations were detected along a section of shoreline
north of Challenger Beach, corresponding approximately to the location of the
Naval Base caravan park. Nutrient enriched groundwater might be the result of
leaching from septic tanks, although the source of contamination has not been
investigated. The average total nitrogen concentration of groundwater in this area
was approximately 8 mg.l-1.
At remaining locations along the shoreline, the average total nitrogen concentration
of groundwater was approximately 4 mg.l-1. Contrary to previous groundwater
investigations, nutrient enriched groundwater was not detected in the Northern
Harbour porewater samples even though high nutrient levels are observed in
groundwater. This provided further evidence that the Northern Harbour plumes
probably discharge further offshore and beyond the shoreline strip of seabed from
which the porewater samples were collected.
Submarine Groundwater Discharge
Several conceptual models of groundwater discharge into Cockburn Sound are
feasible based on the principle that groundwater in the Superficial Aquifer flows
generally toward ocean along pathways of least resistance. In a homogeneous
aquifer, flow is mainly toward the shoreline because it represents the shortest travel
distance to the nearest point of discharge. In an inhomogeneous aquifer,
groundwater may discharge further from the shore if either a low-conductivity
sediment layer restricts vertical flow toward the shoreline, or high-conductivity
zones, such as fractures and solution features, focus SGD to other locations. Both of
these possibilities occur in Cockburn Sound.
Absence of offshore hydrogeological data at a scale that is relevant to SGD is the
main limitation to identifying which conceptual model of groundwater discharge is
most likely at particular locations along the coast. In particular, the presence and
extent of clay aquitard between the Tamala Limestone and overlying sands is poorly
described across most of the study area. In addition, the scale and structure of
secondary porosity in the coastal limestone is unknown.
Previous estimates and implied values of SGD rates based on groundwater flownet
analyses have varied between 0.7 and 8 m3.d-1.m-1 (i.e. by an approximate order of
magnitude). Similar values of SGD were derived by different investigators but the
associated calculations of groundwater velocity were often based on significantly
different values of hydraulic conductivity and hydraulic gradient.
CSIRO Land and Water
xvi Coast and Clean Seas Project No. WA9911
A groundwater flow model was applied to quantify the aquifer water balance inshore
from Cockburn Sound and estimate regional values of SGD for each 100-m section
of the coast. Modelled discharges varied spatially along the length of the coastline
due to the assigned variation in hydraulic connection between the aquifer and ocean
in the model, distributed groundwater recharge, spatial variability in aquifer
transmissivity and pumping, and focusing of groundwater flow toward coastal
embayments.
Total average SGD into Cockburn Sound between Woodman Point and the southern
boundary of the model was estimated to be 2.8-5.5 m3.d-1.m-1. The estimated range
in average SGD corresponds to low and high estimates of groundwater recharge.
An assumption of high groundwater recharge implies a larger SGD rate, and visa
versa.
Differences assigned to in the hydraulic connection between the aquifer and ocean in
the model had the largest affect on the spatial distribution of SGD. The modelled
distributions of SGD are not corroborated against independent data—since none
exists—and are considered to have large uncertainty. The modelling provided
regional-scale estimates of SGD rates, which cannot take account of the detailed
hydrogeology and associated flow paths that control SGD at the local scale.
Nutrient Discharge
Nutrients in the Superficial Aquifer enter Cockburn Sound by SGD of terrestrial
groundwater and aquifer ‘flushing’ by seawater. The mobile, un-utilised fractions of
nutrients that enter the inshore aquifer are eventually discharged to the marine
environment by advection with regional groundwater flow. It is probable that
seawater movement into and out of coastal sediments also plays a role in
transporting nutrients to the submarine environment. Seawater entering the aquifer
during periods of rising sea level can contact and mix with nutrient enriched
groundwater and then drain back to the ocean during a subsequent period of falling
sea level. This exchange mechanism may be important in locations where the
aquifer beneath a contaminant source is regularly flushed by saltwater flows.
Nitrogen is the nutrient of most concern and has been monitored more intensive than
phosphorus. To date, significant point source emissions of phosphorus into
Cockburn Sound have not been detected and ambient discharges are relatively small.
Previous estimates of groundwater nitrogen loads into Cockburn Sound have varied
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xvii and Nutrient Discharge into Cockburn Sound
in the range 180 to 450 tonnes.yr-1 during the period 1978 to 2000. The most recent
estimate by D.A. Lord (2000) was 220 tonnes.yr-1.
Uncertainty in estimates of nutrient loads is non-quantifiable and stems from
uncertainty in the SGD rates and spatial interpolation of nutrient concentrations in
the aquifer. In particular, it is unclear how SGD rates and nutrient concentrations
measured in groundwater bores should be combined to estimate nutrient loads. In
previous estimates, it has been assumed that nutrients in the upper sand aquifer
travel laterally through the sand and discharge at the coast. This conceptual model
ignores vertical draining of groundwater into the underlying, high-permeability
limestone and is highly unlikely except where the limestone is confined by a clay
aquitard that restricts vertical flows.
To improve our understanding of groundwater nutrient transport into Cockburn
Sound, annual SGD nutrient loads computed in this study and previous studies need
be compared with estimates of mass loadings at the source locations (e.g. based on
loss of product or leaching rates). It is unknown whether these estimates have been
made. Since porewater sampling affords a reasonably high level of confidence in
estimates of nutrient concentrations at the location of SGD, inconsistencies between
nutrient mass loadings at source and discharge locations are likely the results of
errors in the estimated SGD rates.
Spatial variability in groundwater recharge and flow, control the transport
pathways and travel times for nutrients to move from source areas inshore from
Cockburn Sound to the ocean. The presence of preferred flow paths mean that some
areas of the aquifer are likely to contain less active—slower flowing—areas of
groundwater compared with other parts of aquifer, which contain the preferential
flow paths. Nutrients that are mobilized into relatively inactive areas of the flow
system are likely to take longer to reach Cockburn Sound. In addition, because the
flushing time is relatively longer, these areas of aquifer are also likely to contain
nutrients for longer periods following contamination events.
Estimates of nutrient mass fluxes into Cockburn Sound were calculated for the
following combinations of SGD rates and nutrient concentrations:
CSIRO Land and Water
xviii Coast and Clean Seas Project No. WA9911
SGD Rates Nutrient
Concentrations
Nitrogen Load
[tonnes.yr-1]
Phosphorus Load
[tonnes.yr-1]
Model - low recharge Porewater 144 2.8 Model - high recharge Porewater 305 5.6 Model - low recharge Appleyard (1994) 317 1.3 Model - high recharge Appleyard (1994) 585 2.5
A composite estimate of 234±88 tonnes.yr-1 total nitrogen load into Cockburn Sound
was calculated based on the following assumption:
• The model-derived values of SGD were judged the most realistic because
they incorporated the inshore groundwater balance. Groundwater discharge
rates used by Appleyard (1994) thus appear to be comparatively small.
• Groundwater nutrient concentrations are likely to have changed significantly
since 1994. The submarine porewater survey provides the only current and
contemporaneous set of groundwater nutrient concentrations along the
Cockburn Sound coast. This data set includes previously undetected areas of
nutrient-enriched groundwater that discharge from north of James Point. In
addition, porewater samples were collected at the location of groundwater
discharge. More samples were collected along the coastline compared to
previous surveys of groundwater bores, and the density of sampling was
increased at locations where significant nutrient contamination was detected.
• Because the Northern Harbour plumes were not detected by the porewater
survey, a previous estimate of nitrogen mass flux into Northern Harbour
(PPK 2000, Table 3.2, Scheme 4) was adopted. The estimate allows for a
51% reduction in nitrogen loading to the harbour due to plume recovery
operations.
• Other factors of importance in determining effective management strategies
for reducing environmental impacts include estimation of the time of arrival
and discharge of on-shore groundwater contaminants into the near-shore
environment, improvement of the management of on-shore groundwater
resources, and better understanding of seawater intrusion. The latter factor
is regarded as a coupled but inverse process to SGD. However, the central
question of establishing a causal link between SGD-derived nitrogen input
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xix and Nutrient Discharge into Cockburn Sound
and the extent of, trigger levels and conditions for occurrence of algal
blooms in the near shore environment remains unanswered even though a
high level of confidence can now be claimed in terms of the net SGD flux and
nitrogen loads into Cockburn Sound.
A potential inadequacy of the porewater survey was the possibility that porewater
samples collected within 10 m of the shoreline were not representative of
groundwater discharging further from the shore. This uncertainty cannot be
resolved without a more detailed investigation of local-scale SGD processes.
CSIRO Land and Water
xx Coast and Clean Seas Project No. WA9911
ACKNOWLEDGEMENTS
Funding Agencies
CSIRO Land and Water
National Heritage Trust Coast & Clean Seas Program (Project No. WA9911)
Centre for Groundwater Studies
Kwinana Industries Council
Department of Industry and Technology (former Dept. of Commerce and Trade)
In-kind Contributions
Assistance received from the following individuals and organisations is gratefully
acknowledged:
• The Water and Rivers Commission of Western Australia; hydrological and
hydrogeological data supplied from the state water resource database
• Department of Transport Maritime Division; tide data for Fremantle
• PPK Environment and Infrastructure; groundwater level and nutrient data for
bores in the Jervoise Bay area.
• Steve Krupa* (South Florida Water Management District); seepage
measurements (“Krupaseep”) and environmental data from the Northern Harbour
SGD intercomparison
• Makoto Taniguchi* (Nara University of Education, Japan); electrical conductivity
and temperature profiles for groundwater bores in the Jervoise Bay area
• Christopher Smith*, Daniel O’Rourke* and Ronald Paulsen* (Cornell
Cooperative Extension, New York); seepage measurements (ultrasonic) from the
Northern Harbour SGD intercomparison
• Simon Nield (Nield Consulting, Western Australia); documentation and
computer modelling files for the Cockburn Groundwater Area groundwater flow
model
• Thomas Stieglitz* (James Cook University, Queensland); data and reports on
porewater electrical conductivity and seismic surveys from the Northern Harbour
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge xxi and Nutrient Discharge into Cockburn Sound
SGD intercomparison, and design details and parts for construction of porewater
EC probe*
• The Scientific Committee on Oceanic Research (SCOR) and the Land-Ocean
Interactions in the Coastal Zone (LOICZ) project of IGBP, with support of the
International Oceanographic Commission (IOC) for financial support of travel
and equipment costs for the SGD intercomparison at Cockburn Sound
• Bill Burnett and Mike Lambert (Florida State University, USA); seepage
measurements (continuous radon monitoring) from the Northern Harbour SGD
intercomparison
• Claus Otto and Lorraine Bates (CSIRO Land and Water); bore completion details
for groundwater monitoring bores on Garden Island
• Gerald Watson and Robert Woodbury (CSIRO Land and Water); technical and
logistical support
Steering Committee Members
Chris Barber (Centre for Groundwater Studies) - Chairman
Mat Vanderklift (Ministry for Planning) - WA Coast and Clean Seas Coordinator
John Braid – former WA Coast and Clean Seas Coordinator
Greg Davis (CSIRO Land and Water)
Steve Morris (Department of Industry and Technology)
Mark Nener (Water Corporation of Western Australia)
Tony Smith (CSIRO Land and Water)
Bradley Patterson (CSIRO Land and Water)
Steve Appleyard (Water and Rivers Commission)
Phil Manning (Water and Rivers Commission)
Phil Hine (Department of Environmental Protection)
Rod Lukatelich (BP Oil) - Kwinana Industries Council representative
* Supported by the SCOR-LOICZ-IOC SGD Intercomparison Programme
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 1 and Nutrient Discharge into Cockburn Sound
1 INTRODUCTION
1.1 BACKGROUND
The waters of Cockburn Sound are the most intensively used marine system in
Western Australia (Cockburn Sound Management Council, 2001). Decline in the
health of seagrass in Cockburn Sound first became apparent in the 1950s and by the
mid-1990s approximately 80% of the original meadows had been lost (Department
of Environmental Protection, 1995). A major study in the late 1970s (Department of
Conservation and Environment, 1979) attributed seagrass death to eutrophication and
smothering of plants by epiphytic algae, which were thriving in the nutrient-enriched
waters. Waste disposal by major industries, ocean discharge of sewerage effluent,
and inshore contamination of groundwater that is known to discharge via submarine
groundwater discharge (SGD) into Cockburn Sound were identified as the main
factors contributing to nutrient enrichment. Management responses to reduce
nutrient inputs appeared initially to have been successful and measurable
improvements in marine water quality had occurred by the early-1980s. However,
by the 1990s, chlorophyll-a concentrations had re-deteriorated and were close to the
worst levels experienced in the 1970s (Department of Environmental Protection,
1995; D.A Lord & Associates, 2001). In addition, episodic marine algal blooms
(diatoms) have been observed.
During the last ten years, there has been an apparent divergence between estimates of
nutrient inputs into Cockburn Sound and trends in marine water quality.
Chlorophyll-a levels have only improved slightly during this period despite an
estimated 70% reduction in the total anthropogenic nitrogen load into the sound.
This has raised concerns regarding the accuracy of estimates of nutrient loads,
particularly the quantity of nutrients that enter Cockburn Sound via submarine
groundwater discharge (SGD). Recently, D.A Lord & Associates (2001) estimated
that approximately 70% (219 tonnes.yr-1) of the anthropogenic nitrogen load into
Cockburn Sound was contributed by SGD.
To assess management options for improving water quality in Cockburn Sound, the
accurate determination SGD rates and associated nutrient loads is essential. It
remains a difficult strategic problem to link estimates of SGD nitrogen inputs into
CSIRO Land and Water
2 Coast and Clean Seas Project No. WA9911
Cockburn Sound to the frequency and magnitude of algal blooms and their
subsequent impact on the marine environment of Cockburn Sound.
1.2 THIS STUDY
The objective of this study was to use an integrated approach involving a range of
measurements and techniques to estimate groundwater nutrient loadings and identify
major inputs into Cockburn Sound.
The investigation was conducted by CSIRO Land and Water as part of the Coast &
Cleans Seas Project WA9911, entitled: Quantifying Submarine Groundwater
Discharge and Demonstrating Innovative Clean-Up to Protect Cockburn Sound from
Nutrients.
This joint project involved two components:
1. Mapping and quantifying submarine groundwater discharge and nutrient
discharge into Cockburn Sound, and
2. At one “hot spot”, demonstrating innovative clean up of a nutrient plume by
installing a biologically active, in situ treatment wall across the direction of
groundwater flow.
A full account of the work undertaken to accomplish the first component is set out in
this report. Details of the insitu remediation trial are reported separately by Patterson
et al. (2002).
1.3 REPORT FORMAT
This report is divided into five sections. Section 1 (this section) provides a brief
background to the study, describes the study motivations, and sets out its objectives.
Section 1 presents an overview of the regional hydrogeology and reviews existing
hydrological data that is relevant to the assessment of SGD. Section 1 details field
investigations and results. Sections 0 and 5 present assessments of SGD and nutrient
discharge into Cockburn Sound, respectively. Conclusions and major finding are
summarised in the Executive Summary at the front of the report.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 3 and Nutrient Discharge into Cockburn Sound
��
��
���
��
�
�
������������
��������
�� ����
�����������������������
�������������
��������������
������������
���������
� � �� � �
! �
"����������
���������
�
� � � � ��������
� ����������������������� ������������
#$%%%%
#$%%%%
#&%%%%
#&%%%%
'%%%%%
'%%%%%
(')%%%% (')%%%%
('#%%%% ('#%%%%
(''%%%% (''%%%%
('*%%%% ('*%%%%
('(%%%% ('(%%%%
�������������
�������������
Figure 1: Study location map
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 5 and Nutrient Discharge into Cockburn Sound
2 HYDROGEOLOGY
2.1 REGIONAL SETTING
The Perth Basin lies to the west of the Darling Fault (Figure 2) and contains up to
12,000 meters of marine and continental sediments. The stratigraphic sequence to a
depth of around 2,000 metres below the present land surface elevation contains
Jurassic (210-140 Ma) and Cretaceous (140-65 Ma) age sediments that are overlain
by a relatively thin veneer of late Tertiary to Quaternary age sediment. The Tertiary
and Quaternary deposits, known as the superficial formations, vary in thickness up to
a maximum of approximately 110 meters collectively.
The saturated extent of the superficial formations forms a mostly unconfined aquifer
system that is called the Superficial Aquifer. Recharge to the aquifer occurs through
direct vertical infiltration of rainfall, whereas groundwater drains primarily by lateral
flow through the aquifer. There is some leakage of groundwater to the confined
aquifer system beneath but most unutilised groundwater discharges eventually to the
Indian Ocean, Swan-Canning River estuary system and other regional surface drains.
The result is a number of distinct regional groundwater mounds and flow systems
that are located between the regional discharge boundaries (Figure 3). SGD from the
Superficial Aquifer is the topic of this report.
The Cockburn Sound study area lies mostly on the western edge of the Jandakot
Mound and partly within the Safety Bay Mound. Over most of the study area, the
superficial formations have a saturated thickness of around 25 to 30 metres and
unconformably overlie an erosional surface of Cretaceous sediments that act as a
low-permeability base to the Superficial Aquifer (Figure 4). In the southern part of
the study area, the superficial formations are underlain by Rockingham Sand, which
occupies an erosional channel in the Cretaceous sediments. This is known to be
more than 100 metres deep in places.
Exchange of groundwater between the Superficial Aquifer and Cretaceous-Jurassic
artesian aquifer system varies regionally. Leakage between aquifers is only crudely
defined at local scales.
CSIRO Land and Water
6 Coast and Clean Seas Project No. WA9911
��
��
���
��
�
�
���������
��������
���������������
�����
��������
�� ����
�����������������������
�������������
� � � � �� � �
�
�����������
�
� � � ����������
������������������������������������� !!"#
�������������$�����%�������������&���'����������������(�����)��������
��������������&��*�����&����$�����+��������%,����������
-����.�����-����������
./�/0�
������
������
������
������
�����
�����
! "���� ! "����
! ����� ! �����
! ���� ! ����
! #���� ! #����
! !���� ! !����
Figure 2: Regional Geomorphology
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 7 and Nutrient Discharge into Cockburn Sound
��
��
���
��
�
�
����
���
����
����
���
����
����
���
��� ����
����
���������
��������
�������� ��� ����
����
���� ���
�������������
������ ������
�������������
��� ������
�������������
�������
������������
��������
�� ����
�����������������������
�������������
� � � � �� � �
�
�����������
�
� � � � ����������
������������� ���� !�����������"���� ����#$$%&'������������"���� ����#$$%&
���� !�������������������������������������� ���(��������
��� ����������� ���������� ���������������������
��� ��������� )�(������!�����������"�&
*+ +,�
������
������
������
������
�����
�����
! "���� ! "����
! ����� ! �����
! ���� ! ����
! #���� ! #����
! !���� ! !����
Figure 3: Groundwater Flow Systems of the Superficial Aquifer
CSIRO Land and Water
8 Coast and Clean Seas Project No. WA9911
�������������� ����������������� ������������������ ������������������������
�������������������
������� ���������������������� ���������� �������������������������������� � �����������������! � �����"��������������! �
�#$#%&
��
��
���
��
�
�
���������
��������'
��������(
)*)
+,*),*)
+-)*)
+.)*)
+-,*)
-)*)
-,*)
.)*)
+.,*)
+/)*)+/,*)
,*)
+,*)
+/)*)������������
��������
�� ����
�����������������������
�������������
� � � � �� � �
�
�����������
�
. ) . 0 ��������
&�����������1���������&��������-22,�#�������������������&��������-22,�
������
������
������
������
�����
�����
! "���� ! "����
! ����� ! �����
! ���� ! ����
! #���� ! #����
! !���� ! !����
Figure 4: Subcrop Formations beneath the Superficial Aquifer
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 9 and Nutrient Discharge into Cockburn Sound
��
��
���
��
�
�
���������
��������
���������������
�����
��������
�� ����
�����������������������
�������������
� � � � �� � �
�
�����������
�
� � � ����������
����������������������������������������������������� !�"
�����������������������#����������$%������������#����������$"&������%��������
���$��&������%��������"���$����������"�'��(�$�(����
������
������
������
������
������
�����
�����
! "���� ! "����
! ����� ! �����
! ���� ! ����
! #���� ! #����
! !���� ! !����
Figure 5: Coastal Surface Geology
CSIRO Land and Water
10 Coast and Clean Seas Project No. WA9911
2.2 SUBCROP FORMATIONS
The superficial formations mostly overlie Cretaceous sediments, which include the
Kardinya Shale Member of the Osborne Formation and the Pinjar and Waneroo
Members of the Leederville Formation. A different situation exists in the southern
part of Cockburn Sound where the Rockingham Sand underlies the superficial
formations. The interpreted regional extent of subcrop units (Davidson 1995) is
based on drilling data from the mainland and offshore on Garden Island.
Hydrogeological cross sections through Garden Island and Cockburn Sound are
presented in Figure 6 to Figure 8. The following hydrogeological descriptions of the
subcrop units are based mainly on the synthesis of Davidson (1995).
2.2.1 KARDINYA SHALE MEMBER
The Kardinya Shale Member consists of moderately to tightly consolidated,
interbedded siltstones and shales, which are typically dark green to black, and
includes thin interbeds of mostly fine-grained sandstone. It has relatively low
permeability and is considered to be an aquitard the separates the Superficial Aquifer
from the deeper Leederville Aquifer. Leakage through the Kardinya Shale Member
is considered negligible compared to aquifer flows.
2.2.2 PINJAR AND WANEROO MEMBERS
The Pinjar and Waneroo Members are upper units within the Leederville Aquifer and
consist of discontinuous, interbedded sandstones, siltstones and shale of marine and
non-marine origin. The subcrop of these two units beneath the superficial formations
represents an area of direct contact between the Superficial and Leederville Aquifers.
The intercomparison study area is a region of upward hydraulic gradient where there
is potential for the Leederville Aquifer to discharge to the Superficial Aquifer.
2.2.3 ROCKINGHAM SAND
The Rockingham Sand occupies what is thought to be an erosional channel incised
into the Cretaceous sediments and which probably extends from Rockingham to
offshore beneath the southern end of Garden Island. The unit consists of slightly
silty, medium-grained to coarse-grained sand of shallow marine origin. The
maximum thickness of Rockingham Sand is approximately 110 metres at the
southern end of Cockburn Sound in the Rockingham area.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 11 and Nutrient Discharge into Cockburn Sound
Coc
kburn
Sound
0m
-40
-80
-120
-160
Superf
icia
lA
quifer
Aquitard
Leederv
ille
Aquifer
Lake
Coo
gee
Jerv
oise
Bay
Par
melia
Ban
kB
asin
Unit
Par
melia
Ban
k
Gard
en
Isla
nd
Ho
loce
ne
sed
iment
Kar
din
yaShal
eM
em
ber
(Osb
orn
eFo
rmat
ion)
Henle
ySan
dst
one
Mem
ber
(Osb
orn
eFo
rmat
ion)
Pin
jar
Mem
ber
(Leed
erv
ille
Fo
rmat
ion)
Wan
nero
oM
em
ber
(Leed
erv
ille
Fo
rmat
ion)
Tam
ala
Lim
est
one
Sp
ear
wo
od
San
d
Kar
din
yaShal
eM
em
ber
(Osb
orn
eFo
rmat
ion)
?
?
?
?
??
SO
UR
CE
:
This
cro
ssse
ctio
nis
bas
ed
on
dat
afr
om
Dav
idso
n(1
99
5)
and
1:5
0,0
00
Envi
rom
enta
l
Geo
logy
Map
Sheets
(GSW
A1
98
6,1
98
5)
Fre
man
tle
and
Ro
ckin
gham
Saf
ety
Bay
San
d
Figure 6: Hydrological cross section (Section A) through the northern ends of Garden Island and Cockburn Sound
CSIRO Land and Water
12 Coast and Clean Seas Project No. WA9911
Kar
din
yaShal
eM
em
ber
(Osb
orn
eFo
rmat
ion)
Aq
uit
ard
Sup
erf
icia
l
Aq
uifer
Leed
erv
ille
Aq
uifer
Jam
es
Poin
t
Par
melia
Ban
kB
asin
Unit
Par
melia
Ban
k
Gard
en
Isla
nd
Ho
loce
ne
sed
iment
Kar
din
yaShal
eM
em
ber
(Osb
orn
eFo
rmat
ion)
Pin
jar
Mem
ber
(Leed
erv
ille
Fo
rmat
ion)
Wan
nero
oM
em
ber
(Leed
erv
ille
Fo
rmat
ion)
Tam
ala
Lim
est
one
?
?
?
?
??
0m
-40
-80
-12
0
-16
0
?
?
??
Ro
ckin
gham
San
d
Kin
gsP
ark
Fo
rmat
ion
(aq
uit
ard
)
Coc
kburn
Sound
Sp
ear
wo
od
San
d
SO
UR
CE
:
This
cro
ssse
ctio
nis
bas
ed
on
dat
afr
om
Dav
idso
n(1
99
5)
and
1:5
0,0
00
Envi
rom
enta
l
Geo
logy
Map
Sheets
(GSW
A1
98
6,1
98
5)
Fre
man
tle
and
Ro
ckin
gham
Saf
ety
Bay
San
d
Saf
ety
Bay
San
d
Figure 7: Hydrological cross section (Section B) through the southern end of Garden Island and James Point
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 13 and Nutrient Discharge into Cockburn Sound
0m
-40
-80
-12
0
-16
0
Sup
erf
icia
l
Aq
uifer
Leed
erv
ille
Aq
uifer
Roc
kingh
am
Par
melia
Ban
kB
asin
Unit
Par
melia
Ban
k
Gard
en
Isla
nd
Ho
loce
ne
sed
iment
Kar
din
yaShal
eM
em
ber
(Osb
orn
eFo
rmat
ion)
Wan
nero
oM
em
ber
(Leed
erv
ille
Fo
rmat
ion)
Tam
ala
Lim
est
one
?
???
Ro
ckin
gham
San
d
Kin
gsP
ark
Fo
rmat
ion
(aq
uit
ard
)
Coc
kburn
Sound
Saf
ety
Bay
San
d
Ro
ckin
gham
San
d
Tam
ala
Lim
est
one
??
Lake
Coo
loon
gup
SO
UR
CE
:
This
cro
ssse
ctio
nis
bas
ed
on
dat
afr
om
Dav
idso
n(1
99
5)
and
1:5
0,0
00
Envi
rom
enta
l
Geo
logy
Map
Sheets
(GSW
A1
98
6,1
98
5)
Fre
man
tle
and
Ro
ckin
gham
?
?P
inja
rM
em
ber
(Leed
erv
ille
Fo
rmat
ion)
Figure 8: Hydrological cross section (Section C) through the southern ends of Garden Island and Cockburn Sound
CSIRO Land and Water
14 Coast and Clean Seas Project No. WA9911
2.3 SUPERFICIAL FORMATIONS
Individual stratigraphic units that comprise the superficial formations in the
Cockburn Sound area are—in order of deposition—the Tamala Limestone,
Cooloongup Sand, Becher Sand and Safety Bay Sand. These are known collectively
as the Kwinana Group.
2.3.1 TAMALA LIMESTONE
The Tamala Limestone unit is a calcareous eolianite that unconformably overlies
Cretaceous sediments and Rockingham Sand. It contains various proportions of
quartz sand, fine-grained to medium-grained shell fragments and minor clay lenses.
The limestone typically exhibits secondary porosity in the form of numerous solution
channels and cavities. The average base elevation of Tamala Limestone in the
intercomparison study area is around -25 to -30 metres Australian Height Datum
[mAHD]; see Figure 4.
Garden Island is an offshore outcrop of Tamala Limestone and is part of a former
‘drowned’ dune ridge called the Garden Island Ridge (Seale et al. 1988) which runs
northwest through Point Peron, Garden Island, Carnac Island and Rottnest Island. A
roughly parallel sand and limestone ridge, called the Spearwood Ridge, is located
onshore from the coast along the contact between the Tamala Limestone and Safety
Bay Sand. The area between these ridges is known as the Warnbro-Cockburn
Depression, in which Tamala Limestone also outcrops as submarine reef.
Passmore (1970, Figure 37) depicts a section through the Tamala Limestone and
overlying sediments from the southern end of Garden Island to the mainland across
Cockburn Sound. The section is along a similar line to Section C in this report and
is based on drilling logs from Fremantle Port Authority Line 1 No. 7 (FPA7), Line 10
No. A (FPA10A) and No. F1 (FPAF1). The base of the Tamala Limestone occurs at
approximately –25 to –30 mAHD at these locations, which is consistent with drilling
logs from the mainland.
2.3.2 COOLOONGUP SAND
The Cooloongup Sand unit consists of fine-grained to coarse-grained feldspathic
quartz sand of grey and yellow-brown colour, with variable amounts of shell material
(up to 25%). It unconformably overlies the Tamala Limestone.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 15 and Nutrient Discharge into Cockburn Sound
2.3.3 BECHER SAND
Becher Sand originated in the near-shore marine environment and consists of grey,
fine-grained to medium-grained quartz and skeletal sand that is mostly structureless
and bioturbated. It unconformably overlies the Tamala Limestone and
unconformably underlies the Safety Bay Sand. Becher Sand extends along the
coastal margin and is typically 10 to 15 metres thick. The base of the unit can locally
contain a layer of silty calcareous clay that is rich in shell fragments.
2.3.4 SAFETY BAY SAND
Safety Bay Sand unconformably overlies Tamala Limestone and Becher Sand and
consists of cream, unlithified, calcareous fine-grained to medium-grained quartz sand
and shell fragments. Traces of fine-grained, black heavy minerals are also present.
Safety Bay Sand is clearly visible along the coastal margin as white, aeolian sand
dunes, which extend offshore into Cockburn Sound as submarine bank units.
2.4 GROUNDWATER OCCURRENCE AND FLOW
2.4.1 ARTESIAN AQUIFER SYSTEM
Artesian aquifers of the Perth Basin comprise a multi-layered aquifer system that is
recharged by slow downward leakage of groundwater through overlying formations.
This system extends offshore beneath the seabed and probably discharges to the
ocean at least several kilometers from the coast—possibly through offshore faults
(Davidson, 1995). Nevertheless, the precise mechanism(s) for artesian groundwater
discharge and locations of discharge are unknown.
It is relevant to note that offshore in the central and southern parts of Cockburn
Sound, the upper members of the Leederville Aquifer—the Pinjar and Waneroo
Members—subcrop directly beneath the Tamala Limestone, which is only 10 to 15
metres thick in this area. The average piezometric head of the Leederville Aquifer in
the same area is around 5 metres above the water table elevation of the Superficial
Aquifer (Davidson, 1995). This indicates that there is potential for artesian
groundwater discharge to occur in some parts of Cockburn Sound; however, it is not
investigated in this study.
CSIRO Land and Water
16 Coast and Clean Seas Project No. WA9911
2.4.2 ROCKINGHAM AQUIFER
The Rockingham Aquifer is defined in association with the Rockingham Sand and is
locally confined by discontinuous clay lenses at the base of the superficial
formations. Confinement of groundwater in the Rockingham Aquifer is indicated by
propagation of ocean tidal fluctuations more than 150 metres inland from the
shoreline (Passmore, 1970). This suggests either very high aquifer transmissivity—
allowing high rates of tidal groundwater flow—or propagation of the tidal signal
through elastic expansion and compression of the aquifer under confined or semi-
confined conditions. Based on the sediment type, the latter explanation is more
likely.
The ocean saltwedge can penetrate deeper into the Rockingham Aquifer because the
sediments are thicker and deeper than in the Superficial Aquifer. The bottom part of
the aquifer contains seawater to an elevation of around –65 mAHD, while the top 40
metres contain groundwater of salinity less than 1,000 mg.l-1 (Davidson, 1995). Un-
utilised groundwater from the Safety Bay Mound discharges to the ocean over the
top of the saltwedge.
The Rockingham Aquifer is assumed to have similar hydraulic conductivity to sands
in the superficial formations.
2.4.3 SUPERFICIAL AQUIFER
Groundwater in the Superficial Aquifer flows generally in a westerly direction and
discharges to the near shore marine environment along the coastline at Cockburn
Sound and Safety Bay (Figure 3). The aquifer is recharged by infiltration and
percolation of rainfall. There are virtually no surface drains because the surface
sands are permeable enough to prevent significant surface runoff.
Table 1 is a compilation of estimates of aquifer hydraulic properties for the
Superficial Formations inshore from Cockburn Sound. Note that the coastal strip is
characterised by very high transmissivity, due to secondary porosity in the Tamala
Limestone, and has very small horizontal hydraulic gradients. Several kilometers
inland from the coast is a relatively narrow band of lower permeability sediments
that run roughly parallel to the coastline along the contact between the Bassendean
Dunes and Tamala Limestone (see Figure 14). A dramatic change in hydraulic
gradient is the main evidence for the existence of this flow barrier. The East Beeliar
Wetlands, a north-south chain of lakes and wetlands located approximately 5
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 17 and Nutrient Discharge into Cockburn Sound
kilometres from the coast, are a surface expression of higher groundwater levels that
are elevated behind this barrier. Water levels in these lakes are up to 18 metres
above mean sea level.
Horizontal hydraulic gradients in the coastal strip are as small as 0.0002—equivalent
to a 1 metre change in water level over a 5 kilometre horizontal distance—but are
larger than 0.01 across some parts of the flow barrier. Thus, hydraulic gradients
within this region of the aquifer vary by a factor of around 50. Nield (1999)
attributed these steep gradients to the presence of clay in the Tamala Limestone, as
indicated in hydrogeological logs from a dense network of monitoring bores in the
vicinity of Aloca’s residue disposal areas. The fact that horizontal hydraulic
gradients vary by a factor of approximately 50 indicates that there is a similar order
of magnitude variation in the Superficial Aquifer transmissivity and hydraulic
conductivity. Nield (1999) estimated that the hydraulic conductivity of the
Superficial Aquifer in the Cockburn Groundwater Area varies spatially in the range 6
to 1,660 m.d-1, for the assumption of low rainfall recharge, and 8 to 3,000 m.d-1 for
the assumption of high rainfall recharge.
Along the coastal margin of the superficial formations, groundwater in the Tamala
Limestone is locally confined where the shelly clay layer at the base of the Becher
Sand is present. This layer, commonly referred to as ‘basal clay’ or ‘shell bed’ in
drilling logs, is in the order of 1-metre thick. Vertical hydraulic head differences
between groundwaters in the Tamala Limestone and Safety Bay Sand, plus
propagation of tidal fluctuations hundreds of metres into the Tamala Limestone,
provide evidence that the limestone is confined or semi-confined in some locations.
Tidal fluctuations are rapidly attenuated inshore from the coast in the Safety Bay
Sand and Becher Sand units because groundwater is essentially unconfined. Davis et
al. (1994) observed diurnal tidal fluctuations in piezometric head of up to 0.4 metres
in the Tamala Limestone within the BP Refinery, whereas water level fluctuations in
the overlying sands were negligible at the same locations.
CSIRO Land and Water
18 Coast and Clean Seas Project No. WA9911
Table 1: Estimated hydraulic properties of the Superficial Formations inshore from Cockburn Sound
Reference/Source Property/Value Location/Method
Bodard (1991) T = 190 – 235 m2/d k = 10 – 30 m/d S = 0.3 [1] (unconfined) B = 10 m
Safety Bay Sand
Bodard (1991) T = 1,700 – 2,600 m2/d k = 100 – 250 m/d n = 30% S = 0.3 ± 1.5 (unconfined) S = 0.02 (confined) B = 10 m
Tamala Limestone
Walker (1994) k = 100 – 250 m/d (aquifer unspecified) BP Refinery, slug testing of 20 bores
Davidson (1995) k = 6 – 50 m/d Cockburn Sound area, superficial aquifer
Nield (1999) T = 40,000 m2/d (superficial) Alcoa Kwinana Refinery; pump test
Nield (1999) k = 400 – 1,660 m/d (low recharge scenario) k = 800 – 3,000 m/d (high recharge scenario)
Cockburn Sound coastal strip; Cockburn Groundwater Area model calibration, superficial aquifer
PPK (2000) T = 13,000 m2/d (JBTB1 early-time data) T = 9,400 m2/d (JBTB1 late-time data) T = 19,900 m2/d (JBMB9S recovery) T = 23,800 m2/d (JBMB9S early-time data) S = 2.0E-06 (as above) T = 54,200 m2/d (JBMB9D early-time data) S = 1.4E-01(as above) T = 9,000 m2/d (JBTB2 early-time data) T = 28,800 m2/d (WPM5C early-time data) S = 9.0E-02 (as above) T = 39,700 m2/d (JBMB1early-time data) S = 1.0E-01(as above) T = 25,000 m2/d (adopted mean value) S = 1.0E-01(adopted mean value)
Inshore from Northern Harbour; constant rate pump tests on bores JBTB1 and JBTB2 (observation bores JBMB9S, JBMB9D, WPM5C, JBMB1), superficial aquifer
Nield in PPK (2000) k = 900 m/d (superficial) Northern Harbour; groundwater model calibration
Smith & Hick (2001)
k = 53 - 174 m/d (Safety Bay Sand) Northern Harbour, tidal method
n: porosity k: saturated hydraulic conductivity i: hydraulic gradient B: aquifer saturated thickness T = kB: aquifer transmissivity S: aquifer storage coefficient
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 19 and Nutrient Discharge into Cockburn Sound
2.5 GROUNDWATER BORES
A relatively large number of groundwater bores have been installed into the
Superficial Aquifer adjacent to Cockburn Sound (Figure 9), mainly for water supply
and groundwater monitoring purposes. Water and Rivers Commission (WRC) holds
information about these bores in the State Water Resources Information System
(SWRIS/WIN) and AQWABase. Some Kwinana industries also collect groundwater
data for abstraction and environmental monitoring purposes.
In this study, bore information was gathered from the following sources:
• Water and Rivers Commission - SWRIS/WIN and AQWABase databases.
• Bodard, J.M., Assessment of the BP refinery site, Kwinana, Western Australia:
Volume I Hydrological framework and site work recommendations, BHP
Research, Melbourne Laboratories, 1991.
• BP Refinery groundwater database.
• Davis, G.B.; C. D. Johnston, and B. M. Patterson, Deep drilling data, multiport
installation specifications, groundwater chemistry and soil gas composition for
the first bioremediation field trial at the Kwinana refinery site, Report to BP
(Kwinana), Report No. 94/34, September 1994.
• Rust PPK Environment and Infrastructure, Kwinana Works Groundwater
Monitoring Manual, prepared for Wesfarmers CSBP Ltd., May 1996,
98H065A:PR2:0639:RevA, 1996.
• PPK Environment and Infrastructure, Jervoise Bay Project Groundwater
Recovery Plan, report prepared for Department of Commerce and Trade, March
2000, 98L084A:PR2:5426, 2000.
• CSIRO Land and Water – Garden Island monitoring bores.
Three hundred and fifty bores were registered in the SWRIS/WIN database within the
search window 374000 to 387000 mE and 6425000 to 6445000 mN (AMG zone 50).
One hundred of these were registered for WRC and the rest on behalf of other
agencies. Seven hundred and sixty-four bores were recorded in AQWABase for the
same search window. All tabular information was collated and organized within a
spreadsheet, then imported into ArcView for spatial analysis and display.
CSIRO Land and Water
20 Coast and Clean Seas Project No. WA9911
Shallow groundwater bores in Figure 9 are those bores that do not fully penetrate the
Superficial Formations. These are denoted as “logged” if a record of stratigraphy
was made during drilling, otherwise they are denoted as “unlogged”. “Fully
penetrating” bores are those bores drilled fully through the Superficial Formations
and intersecting Cretaceous sediment. Bores that partially penetrate the Superficial
Formations are referred to as “shallow bores”. Thicknesses of stratigraphic units
were determined from drilling logs if this information was available.
The main purpose of reviewing drilling information was to refine the description of
regional stratigraphy (see Section 2.3) and develop a more detailed conceptual
hydrogeological model at a scale relevant to the investigation of SGD. Nevertheless,
it was found that the amount and quality of information available was insufficient to
significantly refine the regional model at local scales. Detailed groundwater
investigations have been carried out in some areas, for example, adjacent to Northern
Harbour and beneath the BP Refinery, but this level of information does not extend
across the study area. Only a relatively small number of the bores intersect
Cretaceous sediment (Figure 9 and Figure 10) and total thicknesses of the Superficial
Formations are known only at these locations.
Shallow bores range in depth from only a few metres to almost the base of the
Superficial Formations. Many of these are unlogged and do not provide stratigraphic
information. Interpretation and interpolation of stratigraphy based on the available
drilling information is made more difficult by the mixed quality of the bore logs.
Dependent on the drilling method purpose, these vary from detailed descriptions of
stratigraphy and sediment types to coarse descriptions of texture.
A summary of the stratigraphic information derived from the drilling logs is
presented in Figure 10 to Figure 13. These depict total thicknesses of Quaternary-
Tertiary sediments (Figure 10), Tamala Limestone (Figure 11), basal clay (Figure 12)
and Quaternary sands consisting of Safety Bay Sand and Becher Sand (Figure 13). A
greater than symbol (>) is used to indicate that a particular unit was intersected
during drilling but was not fully penetrated. For example, “>7” indicates that there is
at least 7 metres of the particular sediment type but the total thickness is unknown at
the drilled location. The symbol “A” is used to indicate that a unit is apparently
absent and “U” indicates no data (i.e. unknown).
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 21 and Nutrient Discharge into Cockburn Sound
���� ��
���� ��
��
������
��
����
��������������
��
�� ��
������
��
������
��
����
����
������
����
��
��������
��
��
��
��
��
��
����������������
��
����
����
��
����
��������
��������
����������������������
��������������
��������������
����
��
����
���� ������ ����
�� ��
��������
�� ������ ������
���� ���� ��
������������
����������
��
��������
������������
��
��
��
����
��
������
��������������
��
���� ����
���� ������
�� ��
��
����
��
����������
��
��
������ ��
��
��
����������
����
����
�� ����
��
��
��
����
��
��
��
��
��
����������������������������������
��
����
��
��
����
��
��
�� ��
��
��
��
��
��
��
��
��
��
��
��
��
����
����
������ ����
��
������
��
��
������
��
����
��
����
��
�� ����
���� ��
�� �� ���� ��
��������
����
����
��������
������
������
����
������
��
��
��
���� ��������
��
��
����
��
����
��
��
��
������������
����
��
����
����
��
��
����
��
��
��
��
��
����
��
��
��
��
�� ��
��
��
��
��
��
��
��
��
����
��
����������
����
��
��
��
��
����������
������������
��
��
��
��
��
��
��
��
��
��
����
��
������ ���� ��
���� ���� ���������� �������� ������ ����������
��
����
�� ��
��
��
��
����
��������������
��
������������������������ ����������������������������������
��
������
�������� ��������
������������ ���������� ����������
��������
��
��
������ ����������
���� ����
��
��������
��
��
������������
������
������ ��
��������������
�� �� ��
�� �� ��
��
��
��
��
��
������������
��������������������
��
������
������������
����
��
��
��
��������
��
����
��������
��
��
��
��
��
��������
��
������������
����������
������������
��
����
��
��������
����������
��
��
����
����
����
���� ��
��
��������������������������������������������������������������������������������
��
����������
������
��
������
��
��
��
��
����
����������������������������������������������������������
��
����
������������������������������������������������������������������������
��
��
����
����
��
��
����
��
��
������������
��
��
�������� ��
��������
����������
��
��
��
��
��
����
������
��
����
��
������������������������������������
������������������
��������������������
��
������
��
��
����
��
��
����������
��
��
���������� ��
��������
��
������
����������
��������
��������
����
����
��
��
��
��
��
������
��
��
��
��
����
��
��
��
������
�� ��
��
��
����������������
����
����
������
������
��
��
���� ����
��
��
������
��
������ ��
��������
��
��
��������
��
��
����
������
����
��
��
��
��
��
��
��
��
�� ��
��
��
����
��
����
��
��
��
����
��
�� ��
��
����
������������
��
��������
��
����
����
��
����
��
��
�� ��
��
������
��
����
��
��
����
��
��
��
��
��
������������
������������������������������������
��
������
����
������������ ����������������
��������
������
������
����
����
��������������
��
����
������
����
��
����
��
��
��
��
��������
��
��
��
������
����
������
��
��������������������
����������������������������������������������������������
������������������������������������������
��
����
����
��
����
��
��
������������
��
����
������
����
������
��
����
����
����
��
����
��
��
��
��
������ ��
������
��
��
��������
��
��
����
������
����
��
����
��
��
��
��
��
��
��
��
��
��
��
����
������
��
��
�� ��
��
���������
�
� ��
�
� � � ��
�
�����������
���������
�������������
�
��� � ��� � ��� �������
�������������������������������������������� ����
������ �������!����
�"����#�!��������$���
�"����#�!����������$���
��������������������%��&������$'����������&�����$�(������'�����
���$��(������'���������$������������#�� �$ ���
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 9: Groundwater bore locations
CSIRO Land and Water
22 Coast and Clean Seas Project No. WA9911
������������
��������
�� ����
�����������������������
�������������
��
����
����
��
��
��
����
��������
����
��
��
��
��
�� �� ��
��
��
��
��
����
��
����
��
��
����
��
��
���
�������
���
���
���
���� ���
��
���
���������
��
���
��
��� ���
������
���
�����
��
��
���������
��
��
��
���
��
�� �
���
��
���
��
���
���
���
���
������
���
���
���
��
���
��
������
�
����
��
���
��
���
��
���
���
���
���
��
���
��
���
��
��� ����
���
������
����
���
������
��������
���
�������
����
���
��
����
���
���
����
���
��������
����
���
�������
����
���
����
��������
����
�������
��
���
�
�� �� �� � � ��������
�������������������� ���������� ����!"���� ���#�$%�&
��� ���������� ����!"��'��"�(�"����)
* ��������#���(�����)&
+������* �������
���)�#+������* �������&
���)�#����� ��&
�,��-�)�-�� �
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 10: Combined thickness of Quaternary and Tertiary sediments (m); greater than (>) symbols indicate partial thicknesses determined from shallow bore logs, total thicknesses are unknown at these locations
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 23 and Nutrient Discharge into Cockburn Sound
������������
��������
�� ����
�����������������������
�������������
�
��
�
��
��
������
��
����
����
��
�� ��
����
����
��
��
�
��
�
��
��
��
���
�
����
��
����
�����
���
�� ��
����
��
��
���
��� ��
���
��
���
�����
�����
��
��
���
��
��
���
������
���
������
���������������
���
����
���
����
����
���
���
������
���
���
������
���������
����
����
������
�
��� � ��� � ��� ���������
������������������������ �����������!����"#��������$�%&�'
�� �����������!����"#��(��#�)�#����*
+���������$���)�����*'
,������+��������
���*�$,������+��������'
���*�$���� ���'
�-��.�*�.����
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 11: Thickness of Tamala Limestone (m); greater than (>) symbols indicate partial thicknesses determined from shallow bore logs, total thicknesses are unknown at these locations; “A” indicates the unit is absent; “U” indicates no data or unknown
CSIRO Land and Water
24 Coast and Clean Seas Project No. WA9911
������������
��������
�� ����
�����������������������
�������������
��
�
�� ���
�
�
��
��
����
���
��
��
��
������
��������
����
�����
����� ���
���
����
��
��
����
��
���
�
��� � ��� �� � � ��������
������������������������ ���������� ����!"���� ���#�$%�&
��� ���������� ����!"��'��"�(�"����)
* ��������#���(�����)&
+������* �������
���)�#+������* �������&
���)�#����� ��&
�,��-�)�-�� �
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 12: Thickness of basal clay (m); greater than (>) symbols indicate partial thicknesses determined from shallow bore logs, total thicknesses are unknown at these locations; “A” indicates the unit is absent; “U” indicates no data or unknown
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 25 and Nutrient Discharge into Cockburn Sound
������������
��������
�� ����
�����������������������
�������������
�
�
���
��
�����
�
���
�
��
���
� ��
�� �
�
�� �
�
�
���
�
�
��
�
��
�
�
�
�
�
�
��
�
��
�
��
��
� �
� ��
�
�
�
���
�
�
�
�
����
����
����
��������
���
��
��
��
��
��
��
�� ��
�
�
����
����
��
�� ��
��
��
��
��
��
��
����
��
���
��
��
�
��
�� ��
����
�� ���
��
���
������
��
���
���
��
��� ���
��
��� �
� �
�
�� �� �� � ����������
��������������� � �� !��� �� ����"����#$��������%�&'�(
� !��� �� ����"����#$��)��$�*�$��� +
,������ ��%���*���� +(
-������,������ �
�� +�%-������,������ �(
�� +�%����!���(
�.��/�+�/����
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 13: Thickness of Quaternary sands (m); greater than (>) symbols indicate partial thicknesses determined from shallow bore logs, total thicknesses are unknown at these locations; “A” indicates the unit is absent; “U” indicates no data or unknown
CSIRO Land and Water
26 Coast and Clean Seas Project No. WA9911
2.6 GROUNDWATER LEVELS
2.6.1 REGIONAL GROUNDWATER TABLE
Figure 14 is a water table map for the Superficial Aquifer inshore from Cockburn
Sound and presents 1998 water level data compiled by Nield (1999). Water levels
are close to their minimum seasonal values, with maximum values being
approximately 0.4 metres higher. At locations of multiple depth bores, water levels
from the lower part of the Superficial Formations were used.
Extremely small hydraulic gradients occur along the coastal margin of the aquifer.
The 1-m water table elevation is generally a minimum of 2 km inshore from the
coast but is typically up to 4 km inshore. Hydraulic gradients within the coastal strip
of the Superficial Aquifer are in the order of .0002 (.02%) to .0005 (.05%), which
suggests that extremely high-permeability sediments occur in this area. Estimates of
hydraulic conductivity of Tamala Limestone vary up to values greater than
3,000 m.d-1 (see Table 1 and Sections 2.3.1 and 2.4.3).
2.6.2 INFLUENCE OF SEA LEVEL VARIATION
Because the coastal margin of the Superficial Formations contains highly permeable
sediments, shallow groundwater levels are strongly influenced by changes in sea
level at daily (tidal), seasonal and interannual time scales. In general, longer time-
scale sea level changes propagate further into the aquifer from the coast and affect a
larger part of the groundwater system. The ocean saltwedge—which in the study
area is more like a saltwater ‘tongue’—is known to extend up to 2 km inshore. An
extended rise or fall of sea level causes a corresponding rise or fall in both the
elevation of the saltwater interface and shallow groundwater levels.
Nield (PPK, 2000) compared monthly average sea levels to groundwater levels in the
Woodman Point monitoring bore WPM5, located approximately 200 m inshore from
the coast at Northern Harbour. The data depicts a strong correspondence between
sea level and groundwater level at monthly, annual and interannual time scales
during the twelve-year period from 1987 to 1999.
Walker (1994) investigated aquifer tidal propagation beneath the BP Refinery using
five pairs of monitoring bores located at distances of 10 m, 750 m and 1,280 m from
the coast. Diurnal tidal fluctuations in groundwater levels of approximately 2 cm
were measured at the site furthest from the coast. This is consistent with either high
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 27 and Nutrient Discharge into Cockburn Sound
permeability sediments or semi-confinement of Tamala Limestone by overlying
basal clay.
Nield (pers. comm.) found detectable tidal fluctuations in groundwater levels up to
1 km from the coast beneath Alcoa’s alumina refinery, which is located directly
south of Challenger Beach (see Figure 1).
More recently, Smith and Hick (2001) analysed aquifer tidal propagation at Northern
Harbour to obtain estimates of aquifer diffusivity—the ratio of transmissivity to
aquifer storage coefficient (see Appendix A and Table 1). The dominant diurnal
frequencies (K1 and O1 tidal constituents) were found to propagate more than 150 m
into the aquifer during each 24-hour tidal cycle. Since the basal clay is not present at
this location, these results indicate the presence of high-permeability coastal
sediments.
2.6.3 INFLUENCE OF HIGH-PERMEABILITY LIMESTONE
It is clear from the above information that the Tamala Limestone has a large tidal
efficiency due to high permeability associated with secondary porosity. Tidal
efficiency is further increased in areas where the limestone is semi-confined by
overlying clay. Efficient tidal propagation through the limestone effectively extends
the coastal sea-level boundary inshore and underneath the coastal sands. This effect
is clearly illustrated by water level data collected in several piezometers inshore from
the beach at Northern Harbour.
Figure 15 depicts groundwater levels recorded in two sets of nested piezometers that
were installed at Northern Harbour prior to an SGD intercomparison experiment (see
Section 3.1). Drilling and piezometer completion details are reproduced in
Section 3.1.1. The piezometer pairs were installed in an interdunal depression at
distances of approximately 180 m (NH1) and 95 m (NH2) from the shoreline. One
deep piezometer, approximately 17 m deep and slotted over the bottom 6 m, and one
shallow piezometer, approximately 5 metres deep and slotted over the bottom 3 m,
was installed at each location. The bottoms of the deep piezometers were located
just above the top of the limestone.
In late January 2001, a temporary trench was excavated adjacent to the piezometer
locations as part of works being undertaken for the Northern Harbour development.
Dewatering of the trench caused a draw down of shallow water table of
approximately 0.7 m near to piezometer nest NH2. Though the dewatering discharge
CSIRO Land and Water
28 Coast and Clean Seas Project No. WA9911
was initially directed away from the piezometer locations, it was subsequently re-
directed into the interdunal depression that contained the piezometers. This resulted
in local surface flooding and a ‘draw up’ of the shallow water table in this area.
A relevant observation is that, despite deviations in shallow water levels of
approximately 1.2 m at NH2b and 0.8 m at NH1b, the piezometric heads in both of
the deep piezometers were unaffected and continued to reflect sea level.
At this location, it is apparent that the basal clay is absent and the limestone behaves
like a prescribed head boundary condition at the base of the Quaternary sands. Sea
level fluctuations are propagated laterally though the limestone and then upward into
the overlying sediment. This data suggest that, where coastal sands overly high-
permeability limestone, groundwater movement through the sands is likely to be
predominantly vertical with minor lateral flow. In other words, groundwater in the
sand drains downward into the underlying limestone. This hypothesis is supported
by the fact that hydraulic gradients in the sands are much flatter than would be
expected if flow were predominantly horizontal since the hydraulic conductivity of
sand is significantly less than the conductivity of limestone.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 29 and Nutrient Discharge into Cockburn Sound
��������������� ����������������� ��������������������������
� � � � �������
�
���
� ��
�
�
�
��
�
�
�
��
�
�
�
�
����
���
� �
��
�
�
�
��
�
�
�
�
�
��
�
�
�
�
�
�
�
�
�
�
�
�� �
�
�
�
�
��
�
�
����
�
�
�
����
��� ��
�������
��
�
����
�
�
�
��
�
�
��
��
�
�
�
�
�
�
������
���
�����
� ��
�
���� �!� "�� �
������������
��������
�� ����
�����������������������
�������������
#
�$
�
�
%
�$��
���#
�%
��
#&' %&''&�
�&$
�&�
�&'
�&��&�
�&�
'&'
�&$
�&�
�&�'&%�&%
�&�
'&�
�&��&�
�&%
�&�
�&�
�&� �&��&�
�&��&$
�&�
�&�
�&�
�&%
�&�
�&��&'
�&�
%&�
�&��&�
�&'
�&�
�&$�&�
�&%�&�
�&�
�&$
�&%
�&%
�&$
�&$�&� �&�
�&��&%�&%�&%�&��&�
�&�
��&$��&'��&�
�%&�
��&� ��&#
��
�'&�
�� �
��&#
�$&'
��&%��&%
��&�
��&�
��&' ��&�
��&�
��&�
��&�
��&�
�#&�
�#&�
��&'��&� ��&�
��&��%&�
��&���&�
��&#�%&�
��&�
��&�
������������ ����� �� ��(�
���
���� ����������)�&���*+��(���,
�������, ��(!-�.�� �
�!�����/�����,�
������
������
������
������
������
������
������
������
�� ���� �� ����
������� �������
������� �������
������� �������
������� �������
���� ������������� �!� �"�� �0� ��1���"�� ����� ��������������'���*+��
23�3��
Figure 14: Water table map of the Superficial Aquifer inshore from Cockburn Sound based on 1998 water level data (source: Nield, 1999)
CSIRO Land and Water
30 Coast and Clean Seas Project No. WA9911
Figure 15: Tidal fluctuations in groundwater levels inshore from Northern Harbour from November 2000 to June 2001; sampling interval for water levels is 1 hour
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 31 and Nutrient Discharge into Cockburn Sound
2.7 SEAWATER INTRUSION
The saltwedge within the Superficial Formations adjacent to Cockburn Sound is
known to extend up to 2 km inshore. High permeability of Tamala Limestone results
in a long, flat tongue of saltwater rather than a classically shaped curved saltwedge.
This is commonly observed as layer of saline water at the base of the Superficial
Formations. Where the Tamala Limestone is confined above by basal clay that
extends to the coast, a second, much smaller saltwedge is present in the sand units
overlying the clay. Only a few bores in the Cockburn Sound study area intersect the
saltwedge and enable direct observation of saltwater intrusion. Unless a bore is
installed specifically for this purpose, it is normally undesirable to drill into or below
the saltwater-freshwater interface. As a result, most groundwater bores are screened
above the interface.
Knowing the position and shape of the saltwater-freshwater interface can be useful
for inferring information about SGD. For example, if groundwater discharges
through offshore geological structures such as limestone reefs—rather than directly
along the shoreline—then the position of the saltwedge and groundwater levels along
the shoreline should indicate this. Other aspects of the hydrogeology such as the
groundwater discharge velocity, aquifer permeability and shoreline geometry also
affect the position and shape of the saltwedge. Unfortunately, the potential for
conducting these types of analyses is hindered in this study due to the extremely flat
hydraulic gradients and the long, flat shape formed by intruded seawater.
Where the depth of the saltwater-freshwater interface adjacent to Cockburn Sound is
known, the interface position is reasonably well predicted by the Ghyben-Herzberg
approximation. This is a simple hydrostatic relation based on the density difference
between freshwater and seawater. It predicts that, in a freshwater aquifer, the depth
to an abrupt saltwater-freshwater interface will be approximately 40 times the height
of the groundwater table above mean sea level. For example, if the water table
elevation at a particular location is 0.5 m above mean sea level, the Ghyben-
Herzberg approximation predicts that the depth to the saltwater-freshwater interface
below mean sea level will be approximately (0.5 × 40 m) 20 m. If the aquifer is less
than 20-m thick at that location, then the freshwater head is sufficiently large to
prevent saltwater intrusion.
Adjacent to Cockburn Sound the base of the Superficial Formations is around -20 to
-25 m AHD. According to the Ghyben-Herzberg approximation, the inshore extent
CSIRO Land and Water
32 Coast and Clean Seas Project No. WA9911
of saltwater intrusion should correspond broadly to the 0.6 m AHD elevation contour
of the groundwater table. It is clear from Figure 14 that this can be a considerable
distance inshore. More generally, Table 2 can be used to estimate the location of the
saltwedge toe based on the saturated thickness of aquifer and water table elevation.
Tabulated values are derived from the Ghyben-Herzberg approximation, with values
in the second column being equal to values in the first column divided by 41.
Table 2: Relation between aquifer thickness and water table elevation at the position of the saltwedge toe based on the Ghyben-Herzberg approximation.
Aquifer Saturated Thickness
[m]
Water Table Elevation at
Saltwedge Toe [m AMSL]
15 .36 20 .48 25 .61 30 .73 35 .85 40 .98 45 1.1 50 1.2
Monitoring bores that are known to intercept the saltwedge in the study area are the
Mayor Road multiport bores (MR4M to MR6M) and the Cockburn Saltwater
Interface monitoring bores (CSI1/97, CSI2/97 and CSI3/97). The Mayor Road bores
are located directly east and south of Lake Coogee adjacent the northern end of
Cockburn Sound (Figure 17). The Cockburn Saltwater Interface bores are located
closer to the coast near Challenger Beach (Figure 36). Table 3 clearly shows that the
aquifer contains saltwater at the bottom ports of the Mayor Road monitoring bores,
which are located between 25.5 m and 27.5 m below ground level. The top of casing
elevations are less than 4 m AHD and therefore the saltwater interface in this area is
estimated to be around -21 to -24 m AHD.
It is worthwhile to note that Lake Coogee is an expression of the groundwater table
but is saline. Salinities in the lake have varied seasonally in the range 10,000 to
45,000 mg.l-1 of total dissolved solids (Hirschberg, 1990). The type of interaction
that occurs between the coastal saltwedge and saline water in Lake Coogee has not
been clearly established. Figure 16 depicts two possibilities. In case “a”, Lake
Coogee is convectively unstable due to the density contrast between saline lake water
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 33 and Nutrient Discharge into Cockburn Sound
and relatively fresher groundwater. This results in interaction between the lake
‘saltwedge’ and coastal saltwedge. The flow arrows indicate that the saltwedges are
in fact convection cells in which saltwater circulates from the sea (or lake) into the
aquifer and back out. In case “b”, the lake is convectively stable. A low-
permeability lakebed lining prevents natural convection beneath the lake. The
coastal saltwedge passes directly underneath and does not interact with it. In both
cases, Lake Coogee is a discharge lake with respect to the regional groundwater flow
and therefore accumulates solutes derived from aerosol fallout, which is high close to
the coast.
Table 3: Electrical conductivity of groundwater samples from saltwater monitoring bores
EC @ 25 deg. C [µS.cm-1]
Bore ID
Top of Casing
Elevation [mAHD]
Depth to Inlet [m] Min Max Mean Readings
MR4M 1.97 3.5 2,100 7,260 4,007 10 6.5 1,870 1,910 1,890 10 13.0 1,440 1,920 1,803 10 19.0 1,920 3,110 2,144 10 25.5 29,700 32,700 31,411 9 MR5M 3.78 5.0 669 1,210 889 9 10.0 728 1,730 1,411 11 15.0 1,340 1,840 1,448 10 21.0 1,850 17,000 3,919 10 26.0 17,000 20,600 19,600 10 MR6M 3.14 4.0 1,910 3,500 2,543 9 9.0 1,280 1,470 1,385 10 15.0 1,220 1,520 1,357 10 21.0 2,670 3,200 2,855 10 26.0 25,400 27,800 26,650 10 MR7M 3.57 5.0 919 1,630 1,377 10 10.5 1,700 2,030 1,953 10 16.0 2,570 2,900 2,831 10 21.5 4,340 5,310 4,822 10 27.5 30,000 32,100 31,070 10
EC - uncompensated [µS.cm-1]
Bore ID
Top of Casing
Elevation [mAHD]
Depth to Inlet [m] Min Max Mean Readings
CSI1/97 11.14 30.1-32.1 44,700 66,000 50,220 5 CSI2/97 11.2 18.9-20.9 7,730 10,030 8,732 5 CSI3/97 19.68 25.3-27.3 1,400 1,600 1,538 5
Source: WRC SWRIS/WIN database
CSIRO Land and Water
34 Coast and Clean Seas Project No. WA9911
Cockburn Saltwater Interface monitoring bores CSI1/97 and CSI2/97 are located
together approximately 150 m from the coast. Seawater intrusion is detectable at
about -9 m AHD in CSI2/97, which appears to sample a mixture of seawater and
terrestrial groundwater. At approximately -19 m AHD in CSI1/97 the aquifer
appears to contain seawater only. Monitoring bore CSI3/97 is located further
inshore, approximately 350 m from the coast. The aquifer is apparently unaffected
by saltwater intrusion at -6 m AHD at this location.
Gerbaz (1999) monitored groundwater levels and the saltwater interface position
beneath the BP Refinery in eleven monitoring bores. The saltwedge was intercepted
in nine bores and the average depth below water table to the saltwater interface was
between 17 m and 19 m; however, the distances between the coast and the
monitoring bores were not reported. The extent of saltwater intrusion beneath the
refinery was probably affected by groundwater pumping from the refinery production
bores and was observed to be greater in the middle of the refinery area than at the
northern and southern boundaries.
Passmore (1970) describes saltwater intrusion into the Rockingham Aquifer at the
southern end of Cockburn Sound.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 35 and Nutrient Discharge into Cockburn Sound
MR
5M
MR
4M
NH2
NH1
Nor
ther
nH
arbou
rLa
keC
ooge
e
Tam
ala
Lim
est
one
Kar
din
yaSh
ale
Mem
ber
(Osb
orn
eFo
rmat
ion)
Ho
rizo
nta
lsc
ale
1:1
00
00
Vert
ical
scal
e1
:20
00
Vert
ical
exag
gera
tio
nx5
Sand
unit
s
NH3NH4NH5
Sand
unit
s
MR
5M
MR
4M
NH2
NH1
Nor
ther
nH
arbou
rLa
keC
ooge
e
Tam
ala
Lim
est
one
Kar
din
yaSh
ale
Mem
ber
(Osb
orn
eFo
rmat
ion)
Ho
rizo
nta
lsc
ale
1:1
00
00
Vert
ical
scal
e1
:20
00
Vert
ical
exag
gera
tio
nx5
Sand
unit
s
NH3NH4NH5
Sand
unit
s
(a)
(b)
Figure 16: Hydrological cross section (Figure 5, Section D) through the Northern Harbour study site depicting two possible types of interaction between the coastal saltwedge and saline water in Lake Coogee; (a) Convectively unstable lake, (b) Convectively stable lake
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 37 and Nutrient Discharge into Cockburn Sound
3 FIELD INVESTIGATIONS
Field-based activities were concentrated at three locations along the coast of
Cockburn Sound (Figure 1):
Northern Harbour
Challenger Beach
BP Refinery
In November and December 2000, Northern Harbour was the site of an international
scientific SGD intercomparison experiment (SCOR-LOICZ-IOC) that contributed
valuable information about groundwater-discharge processes that operate in
Cockburn Sound. In addition to the site being of interest because of known SGD
occurrence, this area was of particular interest because two groundwater-nutrient
plumes discharge into Northern Harbour (PPK 1999).
Challenger beach is located approximately mid way between Northern Harbour and
BP Refinery. There is limited industrial development inshore from this location and
Tamala Limestone is fully exposed along the shoreline. Groundwater discharge is
thought to be greater in this area because the limestone aquifer is direct contact with
the ocean.
BP Refinery is located along a stretch of foreshore where nutrient contamination of
groundwater is known to occur at high levels. A beach site adjacent to the Small
Boat Harbour at BP Refinery was selected for the in situ remediation trial that was
conducted as a counterpart to this investigation (see Section 1.2).
A major field component of this study was a nutrient survey of submarine porewater
conducted along Cockburn Sound foreshore between Woodman Point and Palm
Beach in Rockingham. The results of the survey are reported in Section 3.4 and
provided the basis of estimates of nutrient mass flux in Section 5.
3.1 NORTHERN HARBOUR
A map of the Northern Harbour and hydrogeological cross-section through the coast
are depicted in Figure 17 and Figure 16, respectively.
CSIRO Land and Water
38 Coast and Clean Seas Project No. WA9911
�
����
������
�����
��
���
�����
����
�����
�����
� ��
��
������
������
����!
"#���
"�"
����
����
�����
��$
�%!&
'���
�����
�"���
��()*�
�+�,-
�'��
�..�
$����
&/�
�0����
0���
���
�$ ���
����1&
��
��
��
��
������
����
��
��
��
����
��
������
��
�� ��
��
��
����
��
����
�� ��
��
��
��
��
��
����
��
��
��
��
��
��
��
��
��
����
��
��
��
��
��
��
��
����������
��
��
��
������������
����
��
��
�� ��
��
��
��
��
������ ������
����
��
��
��
��
�� ��
�� ����
����
��
����
��
��
��
����
��
��
�� ��
��
��
��
��
� �
��
��
�� ��
��
� � � �� �
� �� �� �
� �� �
� �� �
� �
����
����
�����
�2��
����
����
�2��
�����
����
����
��������
�2��
�����
��
��
�
��
��
3�
3�
3� 3�
��
��
3�
3�
� 3�
��
��3�
3�
��
3� ��
���
3�
�
3�
����"�
4/�4/
�
%)�%
%)5%
%)2%
%)�%
���
�����
�
4/��
4/��
4/�5
4/��
4/��
4/�3
6'%'�
� 6'%'� 6'%
'��
6'%'�
�
6'%'�
� 6'%'�
�
6'%'�
6'%'�
6'%'7
�6'�
'�
6'%'2
6'%'�
3
(.%
�'
�'��
6'%'�
5
�'��
�'8�'
7
�'�
�'�
�'��
�'2
�'� �'
5
(.%
�9
(.%
�9 (.%
3'
�� %
�4:��.
�*4�
�4��
��
!;.�
�)��
*�4�
/��!
��
6'�'�
��������
���
�����
��
��
������
� ������
����
���
������
� �����
�����
���
��
����
�������
�����
� �
����
��
������
�������
���
����
����
�
�����
����
�
����
��
����
��
����
��
����
��
����
��
����
��
����
��
����
��
����
��
����
���
����
���
����
���
����
���
����
���
����
���
Figure 17: Northern Harbour (SGD intercomparison) study site
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 39 and Nutrient Discharge into Cockburn Sound
3.1.1 HYDROGEOLOGY
The bottom of the Superficial Formations in the Woodman Point area is intersected
at an elevation of around -30 mAHD (see Section 1). Thus, the saturated thickness
of the Superficial Aquifer is approximately 30 metres. The Kardinya Shale Member
of the Osborne Formation subcrops beneath the Superficial Formations and form a
low-permeability base. Regional water table gradients are small and indicate that the
aquifer has very high transmissivity in this area. Typical water table slopes are in the
order of 0.0003 (.03%).
Two piezometer nests (NH1 and NH2) were installed at Northern Harbour to
monitor inshore groundwater levels during the SGD intercomparison trial. Their
locations are indicated in Figure 17 and Figure 16. At each location, there is one
deep piezometer, approximately 17 metres deep and slotted over the bottom 6
metres, and one shallow piezometer, approximately 5 metres deep and slotted over
the bottom 3 metres. Bore completion details are presented in Table 4.
Table 4: Completion details of intercomparison monitoring bores
Bore ID
Latitude [Degrees, decimal minutes]
Longitude [Degrees, decimal minutes]
Depth [mBGL]
Screen Length
[m]
Ground Surface RL [mAHD]
Top of Pipe RL [mAHD]
NH1a (Deep) -32 08.356 115 45.863 16.44 6.0 1.414 1.721 NH1b (Shallow) -32 08.358 115 45.863 5.44 3.0 1.444 1.944 NH2a (Deep) -32 08.399 115 45.840 18.76 6.0 1.046 1.454 NH2b (Shallow) -32 08.397 115 45.840 5.30 3.0 1.009 1.412
Drilling and piezometer installations were carried out by CLW on 15th November
2000 using an auger drill and solid auger. Because the water table was close to the
surface and most drilling was through fully saturated and unconsolidated sediment,
this method did not allow accurate logging of the stratigraphy. It was possible to
determine the depth to limestone based on drilling resistance and drill ‘chatter’ but
the predominantly sandy sequence above the limestone could only be described
generally. Some interpretation of stratigraphy was possible during retrieval of the
auger sections by examining sediments adhered to the auger blades.
The first attempt to install the deep piezometer at location NH2 failed during drilling,
when the auger plug became stuck and the top of the hole ‘slumped’ due to the
CSIRO Land and Water
40 Coast and Clean Seas Project No. WA9911
drilling spoil presumably entering cavities in the limestone. A new deep hole
(NH2a) was drilled within 2 m of the abandoned hole.
Limestone was encountered at approximately 14 mBGL (elevation 13 mAHD) in the
abandoned hole at NH2 and at approximately 19 mBGL (elevation 18 mAHD) in
NH2a. The depth to limestone at NH1a is approximately 18 mBGL (elevation
17 mAHD).
In general, the sand sequence above the limestone became coarser and muddier with
increasing depth, varying from Safety Bay Sand at the surface to coarse and very
course, dark grey muddy sand above the limestone. The muddy sand contained
rounded sand grains up to 2 mm diameter, shell fragments and sparse plant fibre.
The bottom of NH2a contained yellow, coarse and very coarse muddy sand, with
rounded sand grains up to 3mm diameter.
The interpreted stratigraphic sequence at the site is Tamala Limestone, ?Cooloongup
Sand, Becher Sand and Safety Bay Sand.
3.1.2 GROUNDWATER LEVELS
Groundwater levels in all four piezometers were monitored using capacitive water
level probes (Dataflow Systems) and model 392 data recorders (Dataflow Systems).
Water levels were logged initially at 10-minute intervals during the period of the
SGD intercomparison experiment but the sampling frequency was subsequently
decreased to once per hour, which was adequate to detect and analyse tidal
fluctuations. The collected data is described and analysed in Section 2.6.3 and
Appendix A.
3.1.3 EC SURVEYS OF SUBMARINE POREWATER
The electrical conductivity (EC) of sediment porewater was measured and mapped in
Northern Harbour during the SGD intercomparison. (Stieglitz, 2001b). The
technique employed a conductivity probe (Stieglitz et al., 2000) that was pushed into
the sediment to obtain values of bulk-ground electrical conductivity, as a measure of
the combined electrical conductivity of the sediment and its porewater.
Measurements correspond to a spherical domain, which was approximately 4 cm in
diameter for the instrument used.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 41 and Nutrient Discharge into Cockburn Sound
The EC probe is constructed as a Wenner array of four ring-electrodes, with the two
inner current electrodes generating an electrical field in the sediment and the two
outer voltage electrodes measuring the voltage potential. In use, both current and
voltage are measured with conventional multimeters and the ratio of current to
induced voltage is proportional to the bulk-ground EC.
The bulk-ground EC of sediment containing pure seawater is less than the EC of
seawater alone because non-conducting sediment grains affect in situ measurements
made with the EC probe. Thus, the sediment porosity also affects bulk-ground EC.
Nevertheless, the salinity of the porewater can be estimated—if the sediment
physical properties are assumed relatively uniform—by calibrating measurements
against bulk-ground EC for sediment that is known to contain pure seawater.
EC transect locations and results from the SGD intercomparison (Stieglitz, 2001b)
are depicted in Figure 18 to Figure 21. Survey times are given in Table 5. Note that
Figure 18 is a map of the small section of beach inside of Northern Harbour where
shore-based activities during the intercomparison were focused. The two inshore
stations on E-transect were groundwater-monitoring sites NH1 and NH2; see Figure
17. Note, also, that a bulk-ground EC value of around 15 to 16 mS.cm-1 indicates
that the sediment contained mostly seawater. The EC of seawater is approximately
55 mS.cm-1.
Table 5: Bulk-ground EC surveys by Steiglitz (2001b); measurement times and ocean water levels
R-Transect
Date Time [WST]
Water level [mAHD]
Origin of depth [mAHD]
04/12/00 1330 0.6 0.8 04/12/00 1700 0.8 0.8 05/12/00 1730 0.8 0.8
B-Transect
04/12/00 1330 0.6 0.6
E-Transect
04/12/00 1330 0.6 0.6
Figure 19 and Figure 20 depict bulk-ground EC readings along R-transect and E-
transect, which are both perpendicular to the shore. Both transects extend
approximately 8 m offshore and to a sediment depth of around 1 m below the seabed.
CSIRO Land and Water
42 Coast and Clean Seas Project No. WA9911
A vertical sampling interval of 0.1 m was used. The data indicate that, at the times
of these surveys, there were shoreline zones of relatively fresh SGD approximately
5-6 m wide. Figure 21 depicts bulk-ground EC along B-transect, which is parallel to
the shoreline. This profile indicates that SGD was higher at the southern end of the
beach, near to the location of R-transect.
Since the SGD Intercomparison experiment, additional porewater EC surveys in
Northern harbour have been conducted at the locations, dates and time indicated in
Table 6 and Figure 17. Results, which are depicted in Figure 22, show EC profiles
that were obtained at a uniform depth of approximately 0.5-m below the seabed.
Transect lengths were limited by seabed slope and depth of seawater since
measurements could only be taken in up to chest-deep water. An attempt was made
to extend transects further offshore by taking measurements from a boat but this
proved impracticable due to the difficulty of maintaining a stationary position over
measurement locations.
Table 6: Bulk-ground EC surveys by CLW; measurement locations and times
Northern Harbour
Location (MGA94) Name
Easting [m] Northing [m] Date
Time (WST)
Tide Stage
NHT1 383390 6443299 27/8/01 11:30 Low NHT2 383414 6443265 27/8/01 11:40 Low NHT3 383357 6443314 27/8/01 11:50 Low NHT4 383302 6443323 27/8/01 12:00 Low NHT5 383365 6443317 15/11/01 13:30 - NHT6 383417 6443282 15/11/01 14:00 -
Figure 22 indicates a zone of shoreline SGD approximately 5-m wide at transect
locations NHT5 and NHT6. There is also evidence of fresher porewater and SGD
further offshore at these locations. It is unlikely that all groundwater from the
superficial formations discharges to such a narrow zone along shoreline, which at
some locations appears to be only several metres wide. Seismic data collected
during the SGD intercomparison provides additional evidence that groundwater also
discharges into Northern Harbour further offshore (see Section 3.1.8).
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 43 and Nutrient Discharge into Cockburn Sound
3.1.4 GROUNDWATER CHEMISTRY AND ISOTOPIC COMPOSITION.
Figure 17 shows the locations of groundwater chemistry and isotopic compositions
from selected bores in the proximity of Northern Harbour.
Table 7 is a summary of the major ion chemistry data and stable isotope
compositions of the sampled groundwaters.
Table 7 shows and confirms the elevated nitrogen (nitrate and ammonium)
concentrations of up to 130 and 21 mg.L-1, respectively, that have previously been
reported at Cockburn Sound. Enriched stable isotope compositions and its
correlation with chloride show the influence of mixing of seawater intrusion with the
shallow groundwater discharging toward the coast.
3.1.5 POREWATER CHEMISTRY
Blue bars in Figure 22 indicate locations along porewater EC transects where
porewater samples were collected. The samples were obtained from approximately
0.5 m below the seabed using a spear probe and all samples were analysed for major
ions, nutrients, EC and pH. Analyses were performed by the Chemistry Centre of
Western Australia and results are reported in Table 8.
These results form part of a more extensive survey of porewater nutrients along the
Cockburn Sound foreshore and are discussed in more detail in Section 3.4.
Table 7: Groundwater chemistry and Isotopes
Bore Date Time Water Level Field Field Field Field Field Field Field Field Anion DeuteriumSampled Depth to TOC EC Salinity TDS Temp. pH Eh DO DO Alkalinity CO3 Ca Cl EC HCO3 K Mg N_NH3 N_NO3 Na P_SR SO4_S Si Balance pH sample/smow
m µS/cm mg/L C mV % mg/L mg/L mg/L mg/L mg/L mS/m mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L % (delta)JBMB10 30/11/2000 15:14 12.1 1,405 0.5 553 21.9 7.1 -7 2 1 300 <2 112 200 137 366 16 18 10 2.8 144 <0.01 77 4.2 1.3 7.6 -10.4JBMB13 30/11/2000 15:40 10.2 1,664 0.7 654 21.7 7.1 -9 2 0.2 300 <2 131 260 161 366 17 31 3.3 8.1 175 <0.01 108 3.8 3 7.8 -14.1JBMB3 30/11/2000 16:16 4.9 795 0.1 311 21.1 7.3 -18 67 5.9 213 <2 94 86 77 259 5 14 <0.01 5 56 0.01 37 2.7 3.4 7.9 -16.7JBMB16 30/11/2000 16:37 12.6 3,060 1.4 1212 22.1 6.8 7 2 0.2 425 <2 163 640 296 519 35 48 23 1.4 391 <0.01 155 3.6 -0.1 7.7 -14.5JBMB14 1/12/2000 - - 1,659 0.7 648 21.2 7.2 -11 53 4.7 250 <2 127 300 160 305 11 24 <0.01 8.4 183 <0.01 90 3 1.7 7.9 -13.6JBMB4 1/12/2000 11:03 4.3 2,790 1.3 1096 21.7 7.0 -2 2 0.3 300 <2 161 590 268 366 24 43 7.3 11 355 0.01 148 3.2 1.8 7.6 -14.1JBMB7 1/12/2000 10:20 5.3 4,880 2.5 1928 22.3 6.8 7 2 0.2 425 <2 198 1,200 472 519 47 77 21 0.03 693 <0.01 234 2.8 0.3 7.6 -12.2
JBMB9D 1/12/2000 10:43 4.1 50,100 32.6 - 24.1 7.3 -17 5 0.4 263 <2 426 18,000 5,200 320 377 1,210 11 <0.01 9740 0.02 2,480 2.1 -0.8 7.4 4.1NH1A 30/11/2000 - 1.3 2,680 1.2 1052 21.1 7.5 -30 2 0.2 245 <2 96 560 282 299 27 62 <0.01 5.2 371 <0.01 204 3.5 2.8 7.9 -13.0NH1B 30/11/2000 - 1.5 872 0.2 338 20 7.7 -42 1 0.1 263 <2 44 79 86 320 8 45 0.33 <0.01 77 0.01 50 3.5 4.9 8.0 -17.7NH2A 29/11/2000 - 1.0 3,530 1.7 1380 21.1 7.4 -23 5 0.4 238 <2 127 840 339 290 26 72 <0.01 6.8 515 <0.01 218 3.9 2.7 7.8 -13.3NH2B 29/11/2000 16:50 1.0 2,180 0.9 843 19.5 7.9 -52 3 0.2 263 <2 47 390 205 320 20 84 <0.01 <0.01 286 0.01 194 2.5 4.5 7.9 -12.3NH3 30/11/2000 - - 1,048 0.3 403 19.3 7.8 -45 4 0.3 263 <2 54 140 101 320 7 41 <0.01 <0.01 107 <0.01 43 3.3 3.6 8.0 -8.5NH4 30/11/2000 - - 1,377 0.5 541 21.7 7.8 -44 2 0.2 313 <2 60 210 134 381 9 47 <0.01 <0.01 171 0.01 86 3.8 1.6 8.0 -8.9NH5 30/11/2000 - - 1,657 0.7 652 21.3 7.7 -40 3 0.2 313 <2 75 280 161 381 11 50 <0.01 0.1 207 0.01 115 4.2 1.6 8.0 -8.2
Test (A.L.) 1/12/2000 11:12 15.2 4,690 2.5 1900 26.8 6.5 27 2 0.2 1390 <2 338 710 490 1,690 80 63 130 0.51 441 3.4 33 4.4 -5.6 7.3 -10.6WPM5C 30/11/2000 14:45 11.0 1,767 0.7 694 21.7 7.0 -5 6 0.6 288 <2 153 290 173 351 7 34 0.02 21 176 0.01 49 4.3 5 7.5 -14.2
CC1 (200m) 1/12/2000 - - - - - - - - - - 138 <2 23 720 256 168 20 36 0.16 0.01 473 0.01 106 10 -0.4 7.6 -23.8CC2 (400m) 1/12/2000 - - - - - - - - - - 188 <2 18 670 250 229 16 14 <0.01 0.09 517 0.01 58 7.4 2 7.8 -23.5
NH6 6/12/2000 10:44 1.3 52,100 33.5 22.1 7.9 -53 2 0.2 - - - - - - - - - - - - - - - - -NH7 6/12/2000 10:43 1.0 52,100 33.4 22.2 7.9 -53 2 0.2 - - - - - - - - - - - - - - - - -
Table 8: Northern Harbour porewater chemistry
CCWA ID Name Date CO3 Ca Cl EC @ 25 °C
[mS.m-1] HCO3 K Mg N_NH3 N_NO2 N_NO3 N_total Na P_SR P_total S_SO4
Ionic Balance
[%] pH
01E0594/012 NHT5-3m 15-Nov-01 <2 163 5600 1680 262 123 399 0.52 <0.02 0.01 0.59 3070 0.03 0.03 618 1 8 01E0594/013 NHT6-1m 15-Nov-01 <2 178 1800 665 281 47 133 0.11 <0.02 <0.01 0.19 1030 0.01 0.02 361 2.3 7.4 01E0594/014 NHT6-5m 15-Nov-01 <2 174 1600 562 281 44 125 0.19 <0.02 <0.01 0.26 972 0.01 0.08 352 4.3 7.4
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered); N_total
– total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (infiltered) 3. FIA – Flow Injector Analysis
CSIRO Land and Water
46 Coast and Clean Seas Project No. WA9911
Reproduced from: Steiglitz (2001)
NH1
NH2
Figure 18: Bulk-ground EC measurements by Steiglitz (reproduced from: Steiglitz, 2001b); transect locations in Northern Harbour during the SCOR SGD intercomparison experiment
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 47 and Nutrient Discharge into Cockburn Sound
R-Transect.
0 1 2 3 4 5 6 7 8
-1.5
-1.0
-0.5
0.0
Dep
th (
m)
0m AHD
beach surface
samplingpoints
0 1 2 3 4 5 6 7 8
-1.5
-1.0
-0.5
0.0
Dep
th (
m)
water level
04/12/00 ca. 1330 WST
0 1 2 3 4 5 6 7 8
-1.5
-1.0
-0.5
0.0
Dep
th (
m)
04/12/00 ca. 1700 WST
water level
0 1 2 3 4 5 6 7 8
Distance along transect (m)
-1.5
-1.0
-0.5
0.0
Dep
th (
m)
05/12/00 ca. 1730 WST
approximate position ofNY Mafia's seepage meter
water level
0.0 1.5 3.0 4.5 6.0 7.5 9.0 10.5 12.0 13.5 15.0
Bulk Ground Conductivity (mS/cm)
Figure 19: R-Transect; bulk-ground EC measurements (reproduced from: Steiglitz, 2001b)
CSIRO Land and Water
48 Coast and Clean Seas Project No. WA9911
E-Transect.
Bulk Ground Conductivity (mS/cm)
0.0 1.5 3.0 4.5 6.0 7.5 9.0 10.5 12.0 13.5 15.0
E1(seepage
meter)
0 1 2 3 4 5 6 7 8 9 10
-1.5
-1.0
-0.5
0.0
Dep
th (
m) beach surface
samplingpoints
0m AHD
0 1 2 3 4 5 6 7 8 9 10
Distance along transect (m)
-1.5
-1.0
-0.5
0.0
Dep
th (
m)
06/12/00 ca. 1045 WST
water level
Figure 20: E-Transect; bulk-ground EC measurements (reproduced from: Steiglitz, 2001b)
B-Transect.
Bulk Ground Conductivity (mS/cm)
0.0 1.5 3.0 4.5 6.0 7.5 9.0 10.5 12.0 13.5 15.0
00 10 20 30 40 50 60 70 80 90
-1.0
-0.5
0.0
Dep
th (
m)
0m AHD
beach surface (waterline)
samplingpoints
R-Transect East Transect West Transect
00 10 20 30 40 50 60 70 80 90
Distance along beach (m)
-1.0
-0.5
0.0
Dep
th (
m)
04/12/00 ca. 1730 WST
R-Transect East Transect West Transect
Figure 21: B-Transect; bulk-ground EC measurements (reproduced from: Steiglitz, 2001b)
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 49 and Nutrient Discharge into Cockburn Sound
NHT2 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
NHT1 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
NHT3 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
NHT4 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
NHT5 (middle of beach) - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
NHT6 (east end of beach) - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
Figure 22: Bulk-ground EC measurements; additional Northern Harbour transects; blue bars indicate locations at which porewater samples were collected
CSIRO Land and Water
50 Coast and Clean Seas Project No. WA9911
3.1.6 BENTHIC FLUX MEASUREMENTS
Several different designs of benthic flux meter (BFM)—or seepage meter—were
deployed in Northern Harbour during the SGD intercomparison and on subsequent
occasions. Basic knowledge of the instrument designs is required to make
meaningful interpretations of results. Therefore, measurement principles and
operation procedures for each type of meter are described briefly in Appendix B.
Benthic flux measurements from Northern Harbour are reported below.
Results from SGD Intercomparison
Instrument locations in Northern Harbour during the SGD intercomparison are
depicted in Figure 23. Benthic flux measurements were made over a ten-day period
from 27th November to 6th December, during which the daily mean sea level fell by
around 0.4 m. This was superimposed by diurnal tidal fluctuations that varied in
amplitude from approximately 0.1 m to 0.35 m, with the amplitude of tidal variation
generally decreasing during the measurement period. Further detail about local tides
is contained in Appendix A.
It is now apparent that most of the measurement stations were located beyond the
narrow shoreline zone of SGD that was identified subsequently during profiling of
porewater EC (see Section 3.1.3). While this did not invalidate the experimental
design of the intercomparison—which aimed to compare the performance of
different meters—it did mean that the measured fluxes were probably seawater
exchanges between the sea and aquifer. This notion was supported by the fact that
discharging water collected by Lee-type meters (Appendix B) was saltwater with
seawater EC. This also highlights the advantage of carrying out a beach-probe EC
survey to provide guidance on the deployment of seepage meters (Section 3.1.3).
A summary of benthic flux measurements from the SGD intercomparison is
presented in Table 9. Operation of Lee-type meters was labor intensive and
measurements could only be carried out during the day. The ultrasonic BFM was
deployed continuously at station E2, and for a short period at E1. Time-averaged
values of benthic flux were automatically calculated and recorded every 15 minutes.
This meant that variation in benthic flux at time scales smaller than 15 minutes was
not investigated.
The best measurement sequences were obtained at stations E1 and E2 on East
Transect (Figure 24 and Figure 25); both were located relatively close to shore but
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 51 and Nutrient Discharge into Cockburn Sound
beyond the shoreline zone of SGD. Each plot depicts sea level in Northern Harbour
(denoted as “Tide”) and benthic flux measurements obtained using Lee-type (Lee,
1977) and ultrasonic (Paulsen et al., 1997) meters. Lee-type benthic flux
measurements varied between -3 cm.d-1 and 33 cm.d-1 at station E1, and generally
increased in magnitude during the monitoring period. This increase was apparently
in response to falling sea level and the associated increase in the local hydraulic
gradient toward the sea. Negative values of benthic flux during 27th November
indicate that sea water was ‘recharging’ the aquifer when the sea level was at its
maximum level during the measurement sequence. There is also evidence that
benthic flux responded dynamically to the diurnal tide, decreasing on a rising sea
level and increasing on a falling sea level. There was less variation in benthic flux
measured by the Lee-type meter at station E2 (5 to 14 cm.d-1) though the trend of
increasing benthic flux in response to falling sea level was detectible.
Table 9: Benthic flux measurements; SCOR SGD intercomparison
Meter Type Range of recorded fluxes
[cm.d-1] Averaging Period
Lee-type (drum & bag) -3 to 32 1-3 hours Krupaseep (heat pulse) 1 to 3 - CLW heat-pulse BFM -60 to 60 0.2-10 seconds Ultrasonic -12 to 61 15 minutes
In general, it is apparent from the results that benthic flux is variable and strongly
influenced by sea level. There is a quasi-inverse relation between sea level and
measured benthic flux, which varied from -12 cm.d-1 to 27 cm.d-1 at station E1 and
-12 cm.d-1 to 61 cm.d-1 at station E2.
Heat-pulse BFM Measurements
Subsequent to the SGD intercomparison, benthic flux measurements were made in
Northern Harbour using a heat-pulse BFM developed by CSIRO Land and Water
(Appendix B). On these occasions, care was taken to deploy the instrument within
the narrow shoreline zone of SGD. The results indicated that water fluxes across the
sediment-water interface were highly variable in response to sea level fluctuations at
time scales ranging from seconds to days. In addition, the tide-driven and wave-
driven fluxes were a similar order of magnitude to the predicted magnitude of
terrestrial SGD. Thus, the task of separating SGD measurements into terrestrial-
groundwater and marine-water components is difficult. This would require very long
CSIRO Land and Water
52 Coast and Clean Seas Project No. WA9911
time series of benthic flux measurements that span the broad frequency spectrum of
wave and tide phenomena.
Figure 26 and Figure 27 are examples of the complex heat pulse data obtained under
field conditions in Northern Harbour, in comparison with the results obtained under
laboratory conditions (Appendix B). Water depths, which were detected by a
pressure transducer on the BFM, are also plotted. They reveal subtle details about
water level variation during the 66-second measurement period. Figure 26 indicates
that, after the heater fired at 20 seconds, the water level at the measurement location
remained relatively constant for the next 30-40 seconds. Small-scale fluctuations in
water level at time scales of 1-2 seconds during this period caused only minor
perturbations in benthic flux. At the time of measurement, the direction of benthic
flux was into the sediment. Figure 27 is more typical of the data obtained from
Northern Harbour and reveals the influence on benthic flux of ocean swells, which
have typical periods of around 10 to 15 seconds in Cockburn Sound. At the time the
heater fired, the direction of benthic flux was out of the sediment at a rate of around
10-20 cm.d-1 (0.001-0.002 mm.s-1). Subsequently, several reversals in flow direction
occurred that resulted in multiple temperature peaks at each of the thermistor
locations. Determining the arrival time of the temperature front at thermistors
immediately after the heater fired was the preferred method for computing the
direction and magnitude of benthic flux.
Time-series of benthic flux measurements collected in Northern Harbour at two-
minute intervals (Figure 28) indicate that the direction of benthic flux can be into or
out of the sediment dependent on the water level above the instrument location at the
time of measurement. The effect of ocean swells on benthic flux appears to be
significant even under relatively calm conditions, such as those inside of Northern
Harbour. Benthic flux rates of between –60 and 60 cm.d-1 were measured.
Analytical models of tide and wave affects on benthic flux, which are presented in
Appendix C, also corroborate these results.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 53 and Nutrient Discharge into Cockburn Sound
Figure 23: Site map of shore-based experimental design during the SGD intercomparison; (reproduced from: Krupa et al., 2000)
CSIRO Land and Water
54 Coast and Clean Seas Project No. WA9911
E1 - East Transect
-20.0
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
26/11/00 27/11/00 28/11/00 29/11/00 30/11/00 01/12/00 02/12/00 03/12/00 04/12/00 05/12/00 06/12/00 07/12/00
Time
Ben
thic
Flu
x (c
m/d
ay)
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Tid
e (m
AH
D)
Ultrasonic Lee-type Tide
Figure 24: Benthic flux measurements recorded during the SGD intercomparison; East Transect, station E1
E2 - East Transect
-20.0
-10.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
26/11/00 27/11/00 28/11/00 29/11/00 30/11/00 01/12/00 02/12/00 03/12/00 04/12/00 05/12/00 06/12/00 07/12/00
Time
Ben
thic
Flu
x (c
m/d
ay)
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Tid
e (m
AH
D)
Lee-type Ultrasonic Tide
Figure 25: Benthic flux measurements recorded during the SGD intercomparison; East Transect, station E2
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 55 and Nutrient Discharge into Cockburn Sound
BFM Data - Northern Harbour8/5/01, 11:46 AM
2100
2120
2140
2160
2180
2200
2220
2240
0 10 20 30 40 50 60 70
Time [s]
Tem
p [
cou
nts
]
0
50
100
150
200
250
300
Wat
er D
epth
[co
un
ts]
T1 T2 T3 T4 T5 T6 Depth
Figure 26: Example heat pulse data from Northern Harbour indicating seawater flow into sediment; benthic flux from T2-T3 pair ≈ - 36 cm.d-1, benthic flux from T1-T3 pair ≈ - 21 cm.d-1
BFM Data - Northern Harbour8/5/01, 11:52 AM
2110
2120
2130
2140
2150
2160
2170
2180
2190
2200
0 10 20 30 40 50 60 70
Time [s]
Tem
p [
cou
nts
]
0
50
100
150
200
250
300
Wat
er D
epth
[co
un
ts]
T1 T2 T3 T4 T5 T6 Depth
Figure 27: Example heat pulse data from Northern Harbour indicating SGD; benthic flux from T4-T5 pair ≈ 20 cm.d-1, benthic flux from T4-T6 pair ≈ 9 cm.d-1
CSIRO Land and Water
56 Coast and Clean Seas Project No. WA9911
Figure 28: Benthic flux measurements from Northern Harbour on 8/5/01 from 11:24 AM to 5:16 PM; measurements are every two minutes
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 57 and Nutrient Discharge into Cockburn Sound
3.1.7 CONTINUOUS RADON MONITORING IN NORTHERN HARBOUR
During the SGD intercomparison, researchers from the Department of Oceanography
at Florida State University conducted continuous radon monitoring of the water
column off a small pontoon anchored approximately 25 m offshore in Northern
Harbour. They observed a diurnal signal in radon concentration that was highest at
low tide indicating higher groundwater discharge during that period. Based on
modelling of the inventory of radon in the water column they calculated an average
discharge of fresh groundwater of 10 ± 7 cm.d-1 in the harbour. This result was
reported at the SCOR/LOICZ Working Group 112 meeting held in Catania (Sicily)
following the intercomparison. At the time of writing this report a detailed report on
the continuous radon monitoring work was not available.
Radon (Rn222) is a soluble radionuclide gas that is naturally occurring in
groundwater. It is formed from the decay of Radium, which comes from the decay
of Uranium. Groundwater becomes enriched with radon through contact with
sediments that contain traces of uranium and radium. Because it has a relative short
half-life (3.82 days), radon is ‘lost’ relatively quickly following groundwater
discharge to the surface, where the decayed radon production is not supported
through contact with minerals comprising the aquifer matrix. Because the decay rate
of radon is well known, and if the radon inventory in the water column is monitored
and its concentration in groundwater is known, it is possible to estimate the rate of
groundwater discharge into the water column. Importantly, this method provides an
independent estimate of SGD that does not involve aquifer properties or any aspects
of the inshore groundwater balance.
3.1.8 SEISMIC SURVEY
A seismic survey in a small area of Northern Harbour and one seismic line outside of
the harbour in Cockburn Sound were conducted during the SGD intercomparison
(Figure 29). Thomas Stieglitz (Stieglitz, 2001a) from the Marine Geophysical
Laboratory of James Cook University conducted this work using a 3.5 KHz ORE
Systems seismic profiler mounted of the boat Tartan II.
Note that this technique cannot be used to calculate seepage rates but has been used
successfully at other locations in Australia to map the occurrence of focused SGD
(e.g. Wonky Holes off the coast of Queensland).
CSIRO Land and Water
58 Coast and Clean Seas Project No. WA9911
The survey in Cockburn Sound revealed a number of stationary ‘features’ that were
consistent with fluid of different density to seawater migrating from the seabed.
These features are indicated on the sections of seismic chart depicted in Figure 30
and Figure 31. Numbered features might be indicative fresh groundwater discharge
from solution features and fractures in the Tamala Limestone; however, it was
beyond the scope of this survey to determine whether the observed density anomalies
were due to fresh groundwater discharge, gas migration from seabed sediment or
possibly some other reason. Most features were recorded more than once (e.g.
features 1, 2, 3, 4 and 6) indicating they might be permanent. Because limited data
was gathered, the number and positions of recorded features may not be
representative of the seabed in Northern Harbour and Cockburn Sound.
Generally, this type of seismic feature is recorded where there is a density difference
that affects the reflection of sound; e.g. large schools of fish, a cloud of suspended
sediment or fluid with a different density compared to the surrounding seawater.
Usually, recordings of fish shoals on seismic charts are not as straight and symmetric
as the seismic features recorded in Northern Harbour and Cockburn Sound. In
addition, fish shoals are normally mobile in shallow waters, which prevents repeated
recordings.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 59 and Nutrient Discharge into Cockburn Sound
115.755 115.760 115.765
-32.155
-32.150
-32.145
-32.140
NORTHERN HARBOUR Seismic track 01/12/00
La
titu
de
So
uth
Longitude East
115.755 115.760 115.765
-32.155
-32.150
-32.145
-32.140
NORTHERN HARBOUR Seismic track 01/12/00
La
titu
de
So
uth
Longitude East
115.755 115.760 115.765
-32.155
-32.150
-32.145
-32.140
NORTHERN HARBOUR Seismic track 30/11/00
La
titu
de
So
uth
Longitude East
115.755 115.760 115.765
-32.155
-32.150
-32.145
-32.140
NORTHERN HARBOUR Seismic track 29/11/00
L
atit
ud
e S
ou
th
Longitude East
Figure 29: Seismic tracks in Northern Harbour and Cockburn Sound (reproduced from: Steiglitz, 2001b)
CSIRO Land and Water
60 Coast and Clean Seas Project No. WA9911
Figure 30: Northern Harbour seismic features (reproduced from: Stieglitz, 2001a)
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 61 and Nutrient Discharge into Cockburn Sound
Figure 31: Cockburn Sound seismic features (reproduced from: Stieglitz, 2001a)
CSIRO Land and Water
62 Coast and Clean Seas Project No. WA9911
3.1.9 GROUNDWATER TEMPERATURE AND EC PROFILING
Temperature and EC profiling of selected groundwater monitoring bores in the
Northern Harbour area was carried out on two occasions as part of the SGD
intercomparison. The results of this work, which are reproduced in Figure 32 to
Figure 34, provide useful information about the position of the saltwater-freshwater
interface in the area. Note that profiles were obtained using a 0.5-m vertical
sampling interval.
EC values are indicated over screened sections of bores only but the temperature
profiles are indicated over the full depths of bores, based on the assumption that
water inside a bore casing is approximately in thermal equilibrium with groundwater
directly outside the bore. This is supported by the fact that abrupt temperature
changes between screened and unscreened section of bores were not observed.
The data indicate that groundwater EC is around 2 mS.cm-1 near the water table and
begins to increase with depth from around -10 mAHD. However, the saltwater-
freshwater interface was not detected in the measured bores. Groundwater EC was
generally less than 6 mS.cm-1 in all bores at screened sections. One profile indicated
an abrupt increase in EC inside bore JBMB7 on 28/11/00; at approximately
-14 mAHD, EC increased from less than 5 mS.cm-1 to greater than 13 mS.cm-1. The
accuracy of this data is questionable because the profile measured in nearby
monitoring bore JBMB9D does not indicate the same increase in salinity despite
being several metres deeper than JBMB7.
The general conclusion is that the saltwater-freshwater interface is below -18 mAHD
in the Northern Harbour area. Relatively small increases in salinity start from a
depth of around -10 mAHD and are apparently associated with relatively minor
mixing and dispersion above the saltwater-freshwater interface.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 63 and Nutrient Discharge into Cockburn Sound
Northern Harbour Groundwater Bores(Date: 29/2/00)
-25
-20
-15
-10
-5
0
5
16 18 20 22 24 26 28 30
Water Temperature (ºC)
Ele
vati
on
(m
AH
D)
JBMB1
JBMB2
JBMB3
JBMB4
JBMB5
JBMB7
JBMB8
JBMB9D
JBMB12
JBMB13
JBMB15
OB10
OB11
OB12
JBTB1
JBTB2
Figure 32: Groundwater temperature profiles on 29/2/00 in selected Northern Harbour monitoring bores; profiles are indicated over the full depths of bores
CSIRO Land and Water
64 Coast and Clean Seas Project No. WA9911
Northern Harbour Groundwater Bores(Date: 29/2/00)
-25
-20
-15
-10
-5
0
5
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
EC (µS/cm)
Ele
vati
on
(m
AH
D)
JBMB1
JBMB2
JBMB3
JBMB4
JBMB5
JBMB7
JBMB8
JBMB9D
JBMB12
JBMB13
JBMB15
OB10
OB11
OB12
JBTB1
JBTB2
Figure 33: Groundwater EC profiles on 29/2/00 in selected Northern Harbour monitoring bores; profiles are indicated over the screened sections of bores only
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 65 and Nutrient Discharge into Cockburn Sound
Northern Harbour Groundwater Bores(Date: 28/11/00)
-25
-20
-15
-10
-5
0
5
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Water Temperature (ºC)
Ele
vati
on
(m
AH
D)
JBMB3
JBMB4
JBMB7
JBMB9D
JBMB10
JBMB13
JBMB14
JBMB16
WPM5C
NH1A
NH2A
NH3
NH4
NH5
Figure 34: Groundwater temperature profiles on 28/11/00 in selected Northern Harbour monitoring bores; profiles are indicated over the full depths of bores
CSIRO Land and Water
66 Coast and Clean Seas Project No. WA9911
Northern Harbour Groundwater Bores(Date: 28/11/00)
-25
-20
-15
-10
-5
0
5
0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
EC (µS/cm)
Ele
vati
on
(m
AH
D)
JBMB3
JBMB4
JBMB7
JBMB9D
JBMB10
JBMB13
JBMB14
JBMB16
WPM5C
NH1A
NH2A
NH3
NH4
NH5
Figure 35: Groundwater EC profiles on 28/11/00 in selected Northern Harbour monitoring bores; profiles are indicated over the screened sections of bores only
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 67 and Nutrient Discharge into Cockburn Sound
3.2 CHALLENGER BEACH
Challenger Beach (Figure 36) is located at the northern tip of the Cape Peron Safety
Bay Sand (Figure 5). At its northern end, the beach pinches out and Tamala
Limestone is exposed as a small coastal cliff that runs northward toward Woodman
Point. This strip of coastline is thought to have relatively high SGD because high-
permeability limestone is in direct contact with the sea. It is also in the middle of the
embayment between Woodman Point and James Point where SGD is probably
focused. In general, coastal embayments focus SGD toward their centers because the
coastline extends further into the coastal aquifer at this location and groundwater
generally flows toward the nearest point on the coast.
3.2.1 HYDROGEOLOGY
The base of the Superficial Formations in the area of Challenger Beach is at an
elevation of around -25 mAHD. Saturated thickness of the Superficial Aquifer is
approximately 25 m and regional water table gradients are around 0.0003 (.03%)
toward the coast. There is evidence of basal clay at the base of the coastal sand units
(Figure 12) but there is insufficient data to interpolate the spatial distribution or
thickness of clay. The Kardinya Shale Member of the Osborne Formation subcrops
beneath the Superficial Formations.
CSIRO Land and Water
68 Coast and Clean Seas Project No. WA9911
��
��
��
��
��
����
��
��
��
��
��
��
��
��
��
��
��
������
��
��
��
��
��
����
��
����
��
����
�� ����������
��
��
��
��
����
����
��
����
��
��
������������ ������������������������������������������������������������
����
��
��
��
��
��
��
��
������
�������� ����������������������������������
����
� �� � � � � � � �� �
�
��
��
�
��
��
�
��
��
��
��
��
����
����
��
��
��
��
��
��
��
��
��
��
����
��
�
��
��
��
��
��
��
��
��
��
��
��
��
��
�
��
��
��
������
����
����
���
�� ��
���
���
��
���� ����
���
����
����
����
����������
�����
�����
����
���
!"�
�#����
�����
$����%
&'(��
&)�&��#�"��
#�����
��(���
*��+
%,��
���#��
��(�&
����-
.���!
/0!�
���
�1��
�1��1�
2(#�)
�����
�3�##��
��&����
�(&��4
���
��
���##�
5�4�
���*#�
���6,
��
���##�
5�4�
���*�&
#����6,
��
7���5
�����%
�����
&����
� �
����
�����
(����&��
���*��
),
������
��
��
��
��
� �
��
� �
��
���
��
���
��
���
��
���
��
���
��
���
��
� �
���
� �
���
� �
���
� �
���
Figure 36: Challenger Beach study site
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 69 and Nutrient Discharge into Cockburn Sound
3.2.2 EC SURVEYS OF SUBMARINE POREWATER
Locations where porewater EC surveys were conducted are indicated in Figure 36
and the dates and times of surveys are listed in The variable distribution of porewater
salinity and occurrence of relatively fresher porewater at significant distances
offshore, suggest that a considerable portion groundwater discharge may take place
via preferred pathways associated with solution features and fractures in the
limestone.
Table 10. Results are depicted in Figure 37. See Section 3.1.3 for additional
information about this technique.
Based on transect CBT7 it is evident that fresh groundwater discharges a significant
distance offshore relative to aquifer thickness. Relatively fresh porewater was
observed at 33-m from the shoreline and it seems probable that groundwater may
have been discharging further offshore. Transects could only be measured in up to
1.5-m deep water, which limited the length of transect at each location. Note that
distinct shoreline zones of SGD did not occur at each transect—for example,
compared transects CBT3 and CBT6.
The variable distribution of porewater salinity and occurrence of relatively fresher
porewater at significant distances offshore, suggest that a considerable portion
groundwater discharge may take place via preferred pathways associated with
solution features and fractures in the limestone.
Table 10: Bulk-ground EC surveys; measurement locations and times
Challenger Beach
Location (MGA94) Name
Easting [m] Northing [m] Date Time (WST) Tide Stage
CBT1 384395 6438453 27/8/01 13:00 Rising CBT3 384357 6438730 27/8/01 13:20 Rising CBT4 384385 6438578 27/8/01 13:30 Rising CBT5 384389 6438410 27/8/01 13:40 Rising CBT6 384397 6438197 27/8/01 13:55 Rising
CBT7 (Boat Ramp) - - 15/11/01 08:30 Falling
CSIRO Land and Water
70 Coast and Clean Seas Project No. WA9911
3.2.3 POREWATER CHEMISTRY
Blue bars in Figure 37 indicate locations along porewater EC transects where
porewater samples were collected. Samples were obtained from approximately
0.5 m below the seabed using a spear probe and all samples were analysed for major
ions, nutrients, EC and pH. Results are reported in Table 11 and discussed in more
detail in Section 3.4.
Table 11: Challenger Beach porewater chemistry
CCWA ID Name Date CO3 Ca Cl EC @ 25 °C
[mS.m-1] HCO3 K Mg N_NH3 N_NO2 N_NO3 N_total Na P_SR P_total S_SO4
Ionic Balance
[%] pH
01E0594/015 CBT7-8m 15-Nov-01 <2 151 5400 1540 262 130 360 0.04 <0.02 3.3 3.5 3240 0.03 0.07 805 2.5 7.9 01E0594/016 CBT7-19m 15-Nov-01 <2 144 7300 2240 268 176 496 0.03 0.29 1.7 1.8 4540 0.09 0.13 1080 3.7 8 01E0594/017 CBT7-25m 15-Nov-01 <2 61 4000 1320 293 96 246 0.03 0.38 2.2 2.3 2590 0.07 0.16 589 2.7 8.2 01E0594/018 CBT7-33m 15-Nov-01 <2 137 3800 1220 293 91 233 0.04 0.38 2.3 2.5 2430 0.11 0.15 564 3.6 8.2
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered); N_total –
total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (infiltered) 3. FIA – Flow Injector Analysis
CSIRO Land and Water
CBT1 (boat ramp) – 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
CBT3 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
CBT4 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
CBT5 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
CBT6 - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
CBT7 (boat ramp) - 0.5m
0 2 4 6 8
10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
Figure 37: Bulk-ground EC measurements; Challenger Beach transects; blue bars indicate locations at which porewater samples were collected
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 73 and Nutrient Discharge into Cockburn Sound
3.3 BP REFINERY
Figure 38 shows the Small Boat Harbour at BP Refinery, located directly south of
James Point (Figure 1) near to the boundary between BP Refinery and CSBP. A
regional-scale hydrogeological cross section through this part of the coast is depicted
in Figure 7.
3.3.1 HYDROGEOLOGY
Numerous bores have been drilled within the BP Refinery and Bodard (1991a,
1999b) conducted a detailed review of the hydrogeology utilising this drilling
information. Following is a summary of Bodard’s analysis of lithological data from
approximately 45 boreholes:
• The Leederville Formation subcrops beneath the Superficial Formations and is
generally intersected at around 30-35 m below ground surface.
• The stratigraphic sequence of ?Rockingham Sand, Tamala Limestone, basal clay,
Becher Sand and Safety Bay Sand can be mapped across the refinery site above
the Leederville Formation.
• Rockingham Sand is typically 5-7 m thick and is first intersected between 23 m
and 25 m below ground surface.
• Tamala Limestone overlies Rockingham Sand. It varies in thickness generally
between 7 m and 11 m mainly due to relief on the top surface of the limestone.
There is a distinctive knoll in this relief that is around 10 m higher than the
surrounding limestone. The knoll is approximately 800 m wide at its base and
centered just inshore from the beach on the southwest tip of James Point.
Groundwater seepage has been observed on some occasions from a small outcrop
of the limestone knoll, located on the shoreline immediately west of the refinery.
• Basal clay is consistently found at the base of the Becher Sand. The clay typical
is 1-m thick and up to 2-m thick in places. It appears to infill the lower parts of
depressions in the upper surface of Tamala Limestone and pinches out around the
limestone knoll; it is absent where the top of the Tamala Limestone is above
approximately -8 m AHD. On the other hand, the offshore distribution of basal
clay is unknown.
• Becher Sand is typically 8 m thick and first intersected at around 5-6 m below
ground surface.
CSIRO Land and Water
• Safety Bay Sand is typically 5 m thick.
Hydraulic head in the sand aquifer is up to 0.5 m higher than in the Tamala
Limestone due to the basal clay acting as an aquitard. This means that water levels
in the sand aquifer are relatively more influenced by vertical recharge. Water levels
greater than 1 m AHD and significantly above the regional aquifer water table
(Figure 14) are observed in the sand aquifer in the refinery area. This suggests that
shallow ‘mini-mounds’ develop above the basal clay in response to local
groundwater recharge and that shallow groundwater flow may reflect these gradients
rather than the regional hydraulic gradient.
Tidal propagation in the limestone is more efficient than in other areas due to the
clay layer acting as a confining ‘lid’. Tidal fluctuations are propagated by elastic
expansion and compression of the aquifer and aquifer diffusivity—the ratio of
transmissivity to aquifer storage coefficient—is increased (see Appendix A for
additional information).
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 75 and Nutrient Discharge into Cockburn Sound
�
����
������
�����
��
���
�����
����
�����
�����
� ��
��
������
������
����!
"#���
"�"
����
����
�����
��$
�%!&
'���
�����
�"���
��()*�
�+�,-
�'��
����
��
��
����
��
��
��
��
��
��
����
����
����
��������
��������
��
�� ��
����
��
����
����
����
����
�� ��
�� ���� ��
��
��
����
��
�� ��
�� ��
������
��
��
�� ���� ��
��
����
��
��
��
��
����
����
����
��
��
��
��
��
�� ��
��
��
��
������
����
��
���� ����
������������
����
��������
��
��
��
��
��
����
����
����
��
�� ��
����
��������
�� ��
��������
�� ��
������
���� ��
����
��
��
����
��
��
����
��
� �����
��
��
���� ������ �� �� �� ����
��
����
����
����
��
��
��
���� ����
��
��������
��
��
��
��
��
��
� �� � � �� �� �
�����
����
'.��
�/��
'.�0 '.
���/
��'.
�)1�"
����
����'�
�2�
�3���
�
�
�
�
4�����
�"
���"
��3��
��
������
5�3�
��$��
��6&
��
������
5�3�
��$�"
����
6&��
.��5
���!
7���
"��
� �
����
�����
�����
"����
$���&
������
��
��
�
�
��
��
��
��
���
���
Figure 38: BP Refinery study site
CSIRO Land and Water
3.3.2 EC SURVEYS SUBMARINE POREWATER
Locations of EC surveys, dates and times of surveys and results are presented in
Figure 38, Table 12 and Figure 40, respectively. See Section 3.1.3 for additional
information about this technique.
A shoreline zone of SGD was observed inside the Small Boat Harbour (transects
BPT1 and BPT2). In addition, there was evidence of focused SGD up to 56 m
offshore—the maximum length of transect. Porewater EC’s varied from pure
seawater to almost freshwater values within distances as small as several metres.
This pattern might reflect shoreline SGD from the sand aquifer and offshore SGD
from the limestone via preferred pathways.
The pattern of SGD south of the harbour (transects BPT3-BPT5) was less distinct.
Table 12: Bulk-ground EC surveys; measurement locations and times
BP Refinery
Name Location (MGA94) Date Time (WST) Tide Stage
Easting [m] Northing [m]
BPT1 (Small Boat Harbour)
382988 6432706 12/11/01 12:40 Low
BPT2 (Small Boat Harbour)
382988 6432706 13/11/01 08:50 High
BPT2 382966 6432638 13/11/01 10:35 Falling BPT3 383000 6432568 13/11/01 10:55 Falling BPT5 383000 6432568 15/11/01 10:50 Falling
3.3.3 POREWATER CHEMISTRY
Blue bars in Figure 39 indicate locations along porewater EC transects where
porewater samples were collected. Samples were obtained from approximately
0.5 m below the seabed using a spear probe and all samples were analysed for major
ions, nutrients, EC and pH. All analyses were performed by the Chemistry Centre of
Western Australia. Results are reported in Table 13 and discussed in more detail in
Section 3.4. Note that there is extensive inshore contamination of groundwater by
nutrients in this area and this is reflected by high ammonia concentrations in the
porewater samples.
Table 13: BP Refinery porewater chemistry
CCWA ID Name Date CO3 Ca Cl EC @ 25 °C
[mS.m-1] HCO3 K Mg N_NH3 N_NO2 N_NO3 N_total Na P_SR P_total S_SO4
Ionic Balance
[%] pH
01E0594/001 Seawater 12-Nov-01 <2 410 19000 5170 159 432 1290 5.4 0.08 1.2 6.6 10600 <0.01 <0.01 2790 0.9 8.3 01E0594/002 BPT1-5m 12-Nov-01 <2 118 1000 552 378 29 122 100 <0.02 <0.01 100 664 <0.01 0.08 714 -4.8 7.6 01E0594/004 BPT2-1m 13-Nov-01 <2 9 1800 657 214 4 165 2.3 0.92 12 15 1150 0.01 0.06 453 -1 8.3 01E0594/005 BPT2-12m 13-Nov-01 <2 113 1300 585 384 42 137 100 <0.02 <0.01 100 830 0.02 0.04 776 -3.7 7.6 01E0594/006 BPT2-19m 13-Nov-01 <2 206 4400 1560 421 115 356 85 <0.02 <0.01 85 2720 0.06 0.07 1140 1.5 7.7 01E0594/007 BPT2-31m 13-Nov-01 <2 171 4800 1720 409 142 437 64 <0.02 <0.01 64 3130 0.05 0.07 1390 3.9 7.8 01E0594/008 BPT2-45m 13-Nov-01 <2 127 960 512 403 31 127 100 <0.02 <0.01 100 666 0.01 0.05 713 -2.2 7.6 01E0594/009 BPT2-56m 13-Nov-01 <2 134 1100 514 390 34 127 100 <0.02 <0.01 100 693 0.01 0.05 748 -5.2 7.5 01E0594/010 BPT5-4m 15-Nov-01 <2 140 4400 1460 183 124 293 230 <0.02 110 340 2330 0.01 0.09 592 -4.3 8.2 01E0594/011 BPT5-10m 15-Nov-01 <2 243 9700 2810 214 260 668 100 <0.02 29 130 5570 0.06 0.08 1780 0 7.8
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered); N_total
– total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (infiltered) 3. FIA – Flow Injector Analysis
CSIRO Land and Water
78 Coast and Clean Seas Project No. WA9911
3.3.4 BENTHIC FLUX MEASUREMENTS
CLW’s heat-pulse BFM was deployed inside the Small Boat Harbour on only one
occasion. The corresponding time series of benthic flux measurements, which were
recorded at two-minute intervals, are depicted in Figure 40.
Measurements were made in shallow water within the shoreline zone of SGD that
was indicated by the porewater EC surveys—see BPT1 and BPT2 in Figure 39.
Similar to results obtained in Northern Harbour, both the direction and magnitude of
benthic flux were found to fluctuate in response to relatively small-scale variation in
sea level. There is clear evidence of a quasi-inverse relation between tide stage and
benthic flux but this is appears to be superimposed by fluctuations in benthic flux
due to waves and ocean swells.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 79 and Nutrient Discharge into Cockburn Sound
BPT2 (Small Boat Harbour) - 0.5m
0 2 4 6 8 10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
BPT3 (south of Small Boat Harbour) - 0.5m
0 2 4 6 8 10 12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
BPT1 (Small Boat Harbour) - 0.5m
0 2 4 6 8 10 12 14 16
18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
BPT4 (south of Small Boat Harbour) - 0.5m
0 2 4 6 8 10 12
14 16 18
0 1 2 3 4 5 6 7 8 9 10 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
BPT5 (south of Small Boat Harbour) - 0.5m
0 2 4 6 8 10
12 14 16 18
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Distance from Shoreline [m]
Bulk-ground Conductivity
[mS/cm]
Figure 39: Bulk-ground EC measurements; BP Refinery transects; Blue bars indicate locations where water samples were collected
CSIRO Land and Water
80 Coast and Clean Seas Project No. WA9911
Figure 40: Benthic flux measurements from BP Refinery on 12/5/01 from 8:56 AM to 11:10 AM; measurements are every two minutes
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 81 and Nutrient Discharge into Cockburn Sound
3.4 COCKBURN SOUND SUBMARINE POREWATER SURVEY
A survey of submarine porewater was undertaken to assess the nutrient status of
groundwater discharging into Cockburn Sound at the location of discharge. Between
September 2001 and June 2002, approximately 80 samples were collected along the
foreshore of Cockburn Sound between Woodman Point and Palm Beach in
Rockingham. Sampling locations are indicated in Figure 41.
Results of sample are tabulated in Table 14. See Table 8, Table 11 and Table 13 for
transect results from Northern Harbour, Challenger Beach and BP Refinery,
respectively.
3.4.1 IONIC CHEMISTRY
Based on ionic balances carried out on 18 samples, the dominant porewater
chemistry is that of seawater diluted by various portions of fresh groundwater.
Major ion ratios are depicted in Figure 42.
It was concluded that the portion of fresh groundwater contributing to a porewater
sample could be estimated based on the chloride concentration of the sample, as
compared to the chloride concentration of undiluted seawater.
The relation between porewater EC and chloride concentration is depicted in Figure
43.
3.4.2 CALCULATIONS OF GROUNDWATER NUTRIENT CONCENTRATIONS
Nutrient concentrations in groundwater were estimated from the measured nutrient
concentrations in porewater, allowing for dilution of groundwater by seawater. The
following relation was used
pg NDIN ×= (1)
sp
sg
CiCl
ClClDI
−−
= (2)
where:Ng = nutrient concentration in groundwater; Np = nutrient concentration in
porewater mixture; DI = dilution index; Clg = chloride concentration in groundwater
(500 mg.l-1); Cls = chloride concentration in seawater (19,000 mg.l-1); Clp = chloride
CSIRO Land and Water
82 Coast and Clean Seas Project No. WA9911
concentration in porewater mixture. The relation assumes that all nutrients in the
porewater mixture are derived from the groundwater component. Values of the
dilution index for the collected porewater samples are listed in Table 14.
To collect the porewater samples, an EC probe (Section 3.1.3) was used to locate the
freshest porewater at each sample location. Porewater salinity varied considerably at
each site and typically varied over spatial scales of several meters. Figure 44
indicates how the value of the dilution index increases as the chloride concentration
(i.e. seawater component) of the porewater mixture increases. The median value of
the dilution index for all porewater samples was approximately 2, which is
equivalent to a porewater mixture containing approximately 50% seawater and 50%
groundwater.
Figure 45 and Figure 46 depict total nitrogen and total phosphorus concentrations in
groundwater that were estimated from porewater nutrient concentrations using (1)
and (2) above.
3.4.3 SYNTHESIS OF RESULTS
High nutrient concentrations in submarine porewater—up to 130 mg.l-1 in samples
P11 to P25—were detected along the section of coastline south of James Point. The
corresponding estimates of total nitrogen concentrations in groundwater were up to
200 mg.l-1 and are consistent with results from previous groundwater investigations
(see Section 5.2), which reveal high levels of nutrient contamination in this area. A
seawater sample collected inside of the Small Boat Harbour at BP Refinery had a
total nitrogen concentration of greater than 6 mg.l-1, which is in the order of 30 times
greater than the ‘background’ level in Cockburn Sound (DEP, 1996). Nutrient
enrichment in the harbour is assumed to result from groundwater discharge and
limited flushing by seawater. Nevertheless, there no history of algae blooms
apparent in the harbour.
Extremely high porewater nitrogen concentrations were detected along a short 40-m
section of shoreline north of James Point and adjacent the HIsmelt facilities (samples
P33 and P58 to P60). Total nitrogen concentrations in groundwater of up to
740 mg.l-1 were estimated based on submarine porewater concentrations of up to
220 mg.l-1. There is also evidence of significantly elevated nitrate levels in the
aquifer inshore from this area of coast. Nitrogen concentrations of up to 60 mg.l-1
N_NO3 have been measured in groundwater bores that are monitored by HIsmelt
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 83 and Nutrient Discharge into Cockburn Sound
(Corporate Environmental Consultancy, 2002); however, neither ammonia nor total
nitrogen has been monitored. The source of nutrients for this contamination is
unknown.
Elevated nutrient concentrations were also detected in samples P32 and P39, which
had estimated total nitrogen concentrations in groundwater of 31 and 10 mg.l-1,
respectively.
Slightly elevated concentrations were detected along a section of shoreline north of
Challenger Beach, which corresponds approximately to the location of the Naval
Base caravan park (samples B4 to B6, and P40 to P42). The average of estimated
total nitrogen concentrations in groundwater was approximately 8 mg.l-1 in this area,
compared to 4 mg.l-1 in other areas outside of the above zones of contamination.
Elevated nutrient concentrations in groundwater might be the result of leaching from
septic tanks; however, the source of these nutrients has not been established.
Contrary to previous groundwater investigations, nutrient enriched groundwater was
not detected in the Northern Harbour porewater samples (P1 and P2). However,
elevated nitrogen (ammonium concentrations in the in the range up to 20 mg.L-1 and
one occurrence of 130 mg.L-1) concentrations do occur in groundwater adjacent to
Northern Harbour (Table 7). This result suggests that the Northern Harbour
groundwater probably also discharge further offshore and beyond the immediate
shoreline strip of seabed beneath which the porewater samples were collected (see
Sections 3.1.3 and 3.1.8 for further information about groundwater discharge into
Northern Harbour). As a general principle, the further inshore that a contaminant
enters groundwater, the deeper in the aquifer it is likely to be by the time it reaches
the coast because groundwater flow toward the coast is driven slowly downward by
recharge to the water table. Waste injection by Weston Bioproducts was into saline
groundwater at the base of the superficial aquifer. Remediation operations to recover
the Northern Harbour plumes (PPK, 2000) might also have reduced SGD rates and
nutrient discharge into Northern Harbour at the time of porewater sampling.
Along Rockingham foreshore (sample locations RP1 to RP10) total nitrogen
concentrations in porewater samples ranged from < 0.5 to 5 mg.l-1. The
corresponding estimates of total nitrogen concentration in groundwater were between
1 and 26 mg.l-1. The highest estimates were obtained at sample locations RP2 and
RP9; however, these elevated levels may be artifacts of high-salinity porewater. The
chloride concentrations of both samples were greater than 15,000 mg.l-1, which
CSIRO Land and Water
84 Coast and Clean Seas Project No. WA9911
indicates that the porewater mixture was approximately 80% seawater and 20%
groundwater. Dilution indexes were greater than 5 and the measured porewater
concentrations were scaled by this amount to estimate groundwater concentrations.
Based on the eight other porewater samples from this section of coast, the estimated
total nitrogen concentration in groundwater was between 1 and 8 mg.l-1. In
comparison, Appleyard (1994) found that groundwater in bores 125601 to 125609
(Figure 55, Section 5.2)—located in the same area—had total nitrogen
concentrations between 1 and 4 mg.l-1.
Table 14: Porewater chemistry
Name Easting (AMG84) Northing (AMG84) Date Cl DI EC @ 25 °C [mS.m-1] N_NH3 N_NO2 N_NO3 N_total P_total pH
P1 383143 6443680 11-Feb-02 13000 3.08 3770 0.17 <0.02 <0.01 0.50 0.07 7.6 Seawater 383143 6443680 11-Feb-02 19000 - 4850 <0.01 <0.02 <0.01 0.15 <0.01 8.1 P2 383506 6443456 11-Feb-02 8200 1.71 2420 1.00 <0.02 <0.01 1.70 0.08 7.5 P3 384076 6440315 11-Feb-02 15000 4.63 3740 0.37 <0.02 <0.01 0.76 0.06 7.6 B1 384086 6440191 21-Feb-02 11000 2.31 2930 0.17 <0.02 <0.01 0.41 0.07 7.8 B2 384112 6440083 21-Feb-02 9900 2.03 2750 0.11 <0.02 <0.01 0.40 0.06 7.5 B3 384172 6439840 21-Feb-02 5100 1.33 1830 0.09 0.02 0.23 0.58 0.09 7.5 P4 384233 6439666 11-Feb-02 5500 1.37 1630 <0.01 <0.02 0.51 1.50 0.04 7.4 B4 384243 6439556 21-Feb-02 14000 3.70 3610 0.41 <0.02 0.06 1.40 0.14 7.6 B5 384308 6439381 21-Feb-02 6900 1.91 1990 0.01 <0.02 5.70 5.70 0.03 7.4 B6 384343 6439276 21-Feb-02 9300 1.53 2770 0.01 <0.02 2.80 3.20 0.02 7.5 P41 384408 6439137 14-Feb-02 13000 3.08 3290 <0.01 <0.02 2.40 3.70 0.03 7.8 P40 384512 6438849 14-Feb-02 11000 2.31 2860 <0.01 0.05 1.20 1.90 0.06 7.7 P42 384549 6438642 14-Feb-02 17000 9.25 4110 <0.01 <0.02 0.67 1.10 0.04 7.6 P43 384556 6438398 14-Feb-02 3300 1.18 1100 0.02 <0.02 <0.01 0.14 0.08 8.1 P48 384549 6438219 14-Feb-02 7600 1.62 2210 0.26 <0.02 <0.01 0.56 0.07 7.8 P47 384554 6437967 14-Feb-02 11000 2.31 2970 0.04 <0.02 <0.01 0.19 0.05 7.7 P46 384514 6437550 14-Feb-02 20000 - 4630 <0.01 <0.02 0.14 0.42 0.02 7.7 P45 384471 6437235 14-Feb-02 21000 - 4670 0.45 <0.02 <0.01 0.74 0.01 7.6 P44 384438 6437102 14-Feb-02 14000 3.70 3630 0.04 <0.02 <0.01 0.24 0.08 7.7 P39 384338 6436713 14-Feb-02 10000 2.06 2770 <0.01 <0.02 1.80 5.10 0.02 7.8
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered); N_total – total nitrogen, persulphate
by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (unfiltered) 3. FIA – Flow Injector Analysis
Table 14: Porewater chemistry…continued
Name Easting
(AMG84) Northing (AMG84)
Date Cl DI EC @ 25 °C [mS.m-1]
N_NH3 N_NO2 N_NO3 N_total P_total pH
P38 384236 6436442 14-Feb-02 4400 1.27 1410 <0.01 <0.02 0.22 0.51 0.06 7.8 P37 384133 6436212 14-Feb-02 14000 3.70 3470 <0.01 <0.02 0.79 1.40 0.02 7.7 P36 383988 6435911 14-Feb-02 9900 2.03 2720 <0.01 <0.02 0.07 0.26 0.05 7.7 P35 383844 6435689 14-Feb-02 10000 2.06 2810 0.08 <0.02 <0.01 0.20 0.08 7.7 P34 383740 6435500 14-Feb-02 12000 2.64 3150 0.10 <0.02 <0.01 0.25 0.06 7.7 P63 383608 6435348 27-Mar-02 14000 3.70 3470 0.26 <0.02 <0.01 1.00 0.14 7.7 P62 383597 6435326 27-Mar-02 16000 6.17 3610 0.06 <0.02 <0.01 0.24 0.05 7.7 P61 383586 6435305 27-Mar-02 16000 6.17 3740 0.46 <0.02 <0.01 1.00 0.02 7.5 P60 383578 6435288 27-Mar-02 14000 3.70 3690 31 <0.02 <0.01 31.00 0.07 7.7 P33 383579 6435272 14-Feb-02 8800 1.81 2620 130.00 <0.02 <0.01 200.00 0.07 7.7 P59 383569 6435268 27-Mar-02 13000 3.08 3530 220 <0.02 <0.01 240.00 0.03 7.4 P58 383557 6435248 27-Mar-02 15000 2.31 3450 4.1 <0.02 <0.01 4.50 0.01 7.4 P57 383548 6435229 27-Mar-02 10000 2.06 2720 0.28 <0.02 <0.01 0.41 0.11 7.7 P56 383535 6435207 27-Mar-02 16000 6.17 4180 0.09 <0.02 <0.01 0.31 0.09 7.6 P55 383522 6435183 27-Mar-02 16000 6.17 3960 0.75 <0.02 <0.01 1.30 0.08 7.6 P54 383511 6435162 27-Mar-02 15000 4.63 3900 0.48 <0.02 <0.01 0.80 0.1 7.6 P53 383500 6435141 27-Mar-02 15000 4.63 3840 0.57 <0.02 <0.01 0.83 0.11 7.6 P52 383483 6435122 27-Mar-02 17000 9.25 3670 0.04 <0.02 <0.01 0.21 0.17 7.7 P49 383468 6435103 27-Mar-02 18000 18.50 4360 0.06 0.07 1.6 2.90 0.07 7.5 P50 383468 6435103 27-Mar-02 16000 6.17 3880 0.05 <0.02 <0.01 0.28 0.01 7.5 P32 383472 6435100 14-Feb-02 13000 3.08 3430 4.10 <0.02 <0.01 10.00 0.04 7.6 P31 383400 6434972 14-Feb-02 7900 1.67 2350 0.07 <0.02 0.09 0.43 0.08 7.9
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered);
N_total – total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (unfiltered)
3. FIA – Flow Injector Analysis
Table 14: Porewater chemistry…continued
Name Easting
(AMG84) Northing (AMG84)
Date Cl DI EC @ 25 °C [mS.m-1]
N_NH3 N_NO2 N_NO3 N_total P_total pH
P30 383150 6434659 14-Feb-02 12000 2.64 3490 0.12 <0.02 0.08 0.56 0.11 7.7 P29 383033 6434504 14-Feb-02 12000 2.64 3160 0.14 <0.02 0.21 0.55 0.09 7.6 P28 382880 6434181 14-Feb-02 9500 1.95 2690 0.27 <0.02 <0.01 0.63 0.06 7.6 P26 382793 6433973 14-Feb-02 8800 1.81 2460 0.87 <0.02 <0.01 1.20 0.06 7.6 P27 382727 6433778 14-Feb-02 21000 - 4780 0.25 <0.02 <0.01 0.67 0.13 7.7 P25 383004 6433464 13-Feb-02 3200 1.17 1070 1.60 <0.02 <0.01 2.50 0.12 7.5 P24 383068 6433230 13-Feb-02 1900 1.08 705 6.30 <0.02 <0.01 11.00 0.09 7.4 P23 383069 6433023 13-Feb-02 2200 1.10 707 59.00 <0.02 <0.01 86.00 0.03 7.3 P22 383128 6432873 13-Feb-02 1100 1.03 558 48.00 <0.02 <0.01 100.00 0.08 7.4 P18 383135 6432729 13-Feb-02 4400 1.27 1430 76.00 <0.02 0.01 120.00 0.14 7.8 P20 383154 6432688 13-Feb-02 7100 1.55 2200 34.00 <0.02 20.00 96.00 0.03 8.1 P17 383157 6432674 13-Feb-02 6800 1.52 2070 85.00 <0.02 7.50 130.00 0.05 7.7 Seawater 383157 6432674 13-Feb-02 19000 - 4620 1.10 <0.02 0.22 2.00 0.01 8.0 P21 383165 6432650 13-Feb-02 7000 1.54 2160 61.00 <0.02 11.00 120.00 0.07 7.7 P16 383172 6432604 13-Feb-02 4200 1.25 1420 29.00 <0.02 <0.01 120.00 0.13 7.4 P15 383177 6432540 13-Feb-02 2100 1.09 876 63.00 <0.02 <0.01 79.00 0.65 7.3 P14 383110 6432080 13-Feb-02 3800 1.22 1240 24.00 <0.02 <0.01 36.00 0.10 7.8 P13 383063 6431875 13-Feb-02 5300 1.35 1620 6.20 <0.02 0.08 9.30 0.13 7.8 P12 383018 6431724 13-Feb-02 10000 2.06 2860 6.60 <0.02 0.34 12.00 0.09 7.7 P11 382980 6431628 13-Feb-02 14000 3.70 3630 0.81 <0.02 <0.01 1.40 0.10 7.6 P10 382937 6431538 13-Feb-02 7400 1.59 2140 0.71 <0.02 <0.01 1.40 0.13 7.6 P9 382563 6430768 13-Feb-02 9600 1.97 2680 0.48 <0.02 <0.01 1.20 0.07 7.6
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered);
N_total – total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (unfiltered)
3. FIA – Flow Injector Analysis
Table 14: Porewater chemistry…continued
Name Easting
(AMG84) Northing (AMG84)
Date Cl DI EC @ 25 °C [mS.m-1]
N_NH3 N_NO2 N_NO3 N_total P_total pH
P8 382326 6430487 13-Feb-02 9500 1.95 2690 0.05 <0.02 <0.01 0.20 0.07 7.6 P7 382061 6430184 11-Feb-02 11000 2.31 3180 0.18 <0.02 <0.01 0.38 0.16 7.7 P6 381884 6429943 11-Feb-02 3700 1.21 1200 0.16 <0.02 <0.01 0.38 0.12 8.0 P5 381579 6429513 11-Feb-02 6900 1.53 2040 0.24 <0.02 <0.01 0.48 0.04 7.7 RP11 381186 6429081 11-Jun-02 8,600 1.78 2450 2.1 <0.02 <0.01 2.3 0.07 7.8 RP10 381004 6428916 11-Jun-02 8,600 2.31 3120 0.01 <0.02 0.15 0.33 0.07 7.8 RP9 380811 6428758 11-Jun-02 11,000 5.97 3360 2.3 <0.02 <0.01 2.3 0.08 7.7 RP8 380625 6428604 11-Jun-02 15,900 5.14 3310 1.0 <0.02 <0.01 1.1 0.06 7.8 RP7 380453 6428421 11-Jun-02 15,400 2.31 2870 3.3 <0.02 <0.01 3.5 0.08 7.7 RP6 380248 6428289 11-Jun-02 11,000 1.67 2220 0.15 <0.02 <0.01 0.34 0.06 7.5 RP4 380026 6428172 11-Jun-02 7,900 1.83 2450 0.58 <0.02 <0.01 0.69 0.08 7.9 RP3 379803 6428059 11-Jun-02 8,900 1.08 687 0.46 <0.02 <0.01 0.62 0.07 7.9 RP1 379305 6428043 11-Jun-02 1,900 2.31 3300 0.27 <0.02 <0.01 0.47 0.09 7.8 RP2 379554 6428022 11-Jun-02 11,000 5.14 2880 4.9 <0.02 <0.01 5.0 0.34 7.3
Notes: 1. All concentration are mg.l-1 2. N_NH3 – nitrogen ammonia fraction by FIA (filtered); N_NO2 – nitrogen nitrite fraction by FIA (filtered); N_NO3 – nitrogen nitrate fraction by FIA (filtered);
N_total – total nitrogen, persulphate by FIA (unfiltered); P_SR – phosphorus soluble reactive by FIA (unfiltered); P_total – total phosphorus, persulphate by FIA (unfiltered)
3. FIA – Flow Injector Analysis
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 89 and Nutrient Discharge into Cockburn Sound
������������� ������������������������������������ �
�� � �� � �� �������
��
��������
����������
��������������������
������������������������������������������������������
��
����������������������������
����
�� �� �� �� �� ������
����
������������
��������
�� ����
�����������������������
�������������
�� !��"�#$����
�%&
���'���&������(&
����%��)��'���
!��������*��������������$*��+����'���)���
!��������*�����
������
����
����
����
��
��
��
�� ���
��
�������
�� ���� ���
��
��������������������������������������
��� ���������������
���
������
���
�� ������
��� ���
��� ���
��
��
�
�����������
�����
��
�� ���
&��,������)������������ ���������������������$���%������'-��������%����'�(������-��������'��(���������'������������,��)�')���
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 41: Porewater sampling locations
CSIRO Land and Water
90 Coast and Clean Seas Project No. WA9911
Major Ion Ratios (November 2001 sampling)
-1,000
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
11,000
12,000
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 Chloride Concentration [mg.l -1 ]
Ionic Concentration
[mg.l -1 ]
Sodium Magnesium Potassium Sulphate Calcium Bicarbonate
Figure 42: Porewater ionic chemistry
Porewater Chloride-EC Relation
0
1,000
2,000
3,000
4,000
5,000
6,000
0 5,000 10,000 15,000 20,000 Chloride Concentration [mg.l -1 ]
Electrical Conductivity
@ 25°C [mS.m -1 ]
Nov 2001 sampling Feb 2002 sampling Mar 2002 sampling Jun 2002 sampling
Figure 43: Porewater chloride-EC relation
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 91 and Nutrient Discharge into Cockburn Sound
Porewater Dilution Index
0
5
10
15
20
25
30
35
40
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 Cl [mg.l -1 ]
DI
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Percent Seawater
Porewater sample Dillution index Percent seawater
Figure 44: Porewater dilution index
CSIRO Land and Water
92 Coast and Clean Seas Project No. WA9911
��
����
����
������������������������������
����
����
����������������������������������������
������
��
����������������������������
����
�� �� ����
��������
����
������������
��������
�� ����
�����������������������
�������������
��������������
����
����������������
�����������������
��������������������������������������������������
���������������������
�� � �� ! !� "���������
�#�����������$!$ �%����&������������'����( ��������)#*�&+
������'�����,����������(�������������������������,����������(��
-�!��
!�.�!���
!��.� ��� ��.�!����/�!����
&�������������'����( ���� ��� �����0���������)���� ����������0�������������)������+�����)��������+�,�����������
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 45: Groundwater nitrogen concentrations [mg.l-1] estimated from submarine porewater analyses
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 93 and Nutrient Discharge into Cockburn Sound
������������� ����������������������������������� �
�� � �� � �� �������
��
����
��
��
����������
��������������������
����
����
������������������������������������
����
��
��
����������
������
����������
����
�� �� ����
��������
����
������������
��������
�� ����
�����������������������
�������������
�� !��"�#$����
�%&
���'���&������(&
����%��)��'���
������
!��������*�����
!��������*��������������$*��+����'���)���
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
(����������',����&*�)*���-��.�/������'�$����(����&��,����&*�)*���
��0��� ��
��1�0������
����0�2����
���������������������$���%������'3���������$���%�(������3��������'��(���������'������������,��)�')���
������
Figure 46: Groundwater phosphorus concentrations [mg.l-1] estimated from submarine porewater analyses
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 95 and Nutrient Discharge into Cockburn Sound
4 QUANTIFYING SGD INTO COCKBURN SOUND
Submarine groundwater discharge can be quantified in two ways:
a). By inference from inshore water balance considerations
b). By direct and indirect field measurements of SGD.
Each approach can be employed independently of the other, with the advantage that
results can be crosschecked to reduce uncertainty in estimates. Field measurements
of SGD should be physically realistic and feasible when considered in context of the
inshore hydrogeology and aquifer water balance.
4.1 CONCEPTUAL MODELS OF SGD
Groundwater in shallow, coastal aquifers flows generally toward the ocean along
pathways of least resistance. If an aquifer is homogeneous, flow is mainly toward
the shoreline because it represents the shortest travel distance for groundwater to
discharge from the aquifer. In an inhomogeneous aquifer, groundwater will also
discharge offshore from the coast if either (a) low-conductivity sediment restricts
vertical flow toward the shoreline, or (b) high-conductivity zones, such as fractures
and solution features, focus SGD to other locations. Both of these possibilities occur
in Cockburn Sound.
A number of conceptual models of groundwater discharge into Cockburn Sound are
feasible and six possibilities are depicted in Figure 47. Other patterns of
groundwater discharge are also possible. In all cases, groundwater could discharge
further offshore and beyond the indicated zones of seawater intrusion due to
preferred flow through low-resistance pathways in the limestone. Potentially,
groundwater could discharge many hundreds of metres offshore by this mechanism.
Note, also, that there is potential for artesian groundwater from the Leederville
Aquifer to discharge into Cockburn Sound through overlying limestone. This
possibility is not depicted in Figure 47.
Absence of offshore hydrogeological data is the main limitation to identifying which
type of SGD is most likely to occur at different locations along Cockburn Sound
CSIRO Land and Water
96 Coast and Clean Seas Project No. WA9911
foreshore. In particular, the extent of basal clay is poorly described across most of
the study area and there is virtually no offshore hydrogeological data.
In Type I SGD, coastal sands overlie Tamala Limestone and the basal clay unit
between them is absent. Because there is no significant impediment to vertical flow
through the sands, groundwater discharges upward toward the sea and SGD is
focused along a relative narrow strip of the shoreline. The average hydraulic head
along the seabed is uniform and therefore hydraulic gradients in the aquifer are
steepest toward the closest point on the seabed, which is the shoreline.
In Type II, the shoreline zone of SGD is narrower because the coastal sands are
absent and high permeability limestone is in direct contact with the sea. This
situation is thought to occur between Northern Harbour and Challenger Beach where
limestone is exposed along the shoreline as a coastal cliff. It is emphasised, once
more, that groundwater probably discharges further offshore through secondary
porosity in the limestone.
Type III depicts the situation in which basal clay is present between the sand and
limestone but does not extend offshore. Groundwater flow toward the coast is ‘split’
vertically either side of the clay, though groundwater in the limestone can move
upward and discharge to the shoreline once it has passed beneath the clay. Note that
the dominant groundwater motion in the coastal sands inshore of the clay is vertical.
This is because it is relatively ‘easier’ for groundwater in the sands to drain vertically
downward into high-permeability limestone, travel laterally beneath the clay and
then discharge vertically upward to the ocean.
Basal clay extends a significant distance offshore in Type IV, which prevents
groundwater in the limestone discharging vertically at the shoreline. Two zones of
SGD occur—one along shoreline and the other further offshore—and saltwedges
form above and below the clay. The offshore zone of SGD may be poorly defined
dependent on the competency of the clay as an aquitard and the way it terminates
offshore (e.g. as a gradually thinning layer with poorly defined boundary or abruptly
with a relatively well-defined boundary). Inshore, water levels in the sand and
limestone are effectively ‘de-coupled’ by the aquitard, with tidal fluctuations
propagating more efficiently into high-permeability and semi-confined limestone.
Local, shallow groundwater mounds may develop in the sands above the clay in
response to local recharge.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 97 and Nutrient Discharge into Cockburn Sound
Type V depicts the case where the basal clay extends offshore and pinches out at the
base of a submarine limestone ridge or similar elevation feature in the top surface of
the limestone. The offshore zone of SGD may be more distinct—compared to
Type IV—because the limestone is in direct contact with the sea at the location of
groundwater discharge.
In Type VI, basal clay extends a sufficient distance offshore that groundwater in the
limestone is unable to flow beyond it. Groundwater ‘leaks’ up through the aquitard
and overlying sand to emerge as diffuse SGD.
Figure 47 presents a two-dimensional, vertical impression of SGD; however, the
coastal hydrogeology also affects horizontal patterns of groundwater flow. For
example, there is evidence of local mounding of shallow groundwater above basal
clay in the sand aquifer beneath BP Refinery (see Section 3.3.1). As a general
principle, it is expected that complex hydrogeology will be reflected by complex
patterns of shoreline and offshore SGD. To completely resolve these flow systems
would require systematic field investigations across large spatial and temporal scales.
Such detailed measurements may be impracticable to make in many situations.
CSIRO Land and Water
98 Coast and Clean Seas Project No. WA9911
Cockburn Sound
Tamala
Limestone
Cretaceous Sediment
Quaternary
Sand
Shoreline
SGD
Saltwater
Freshwater
Freshwater
Saltwater
Cockburn Sound
Tamala
Limestone
Cretaceous Sediment
Quaternary
Sand
Shoreline
SGD
SaltwaterFreshwater
Saltwater
Cockburn Sound
Tamala
Limestone
Cretaceous Sediment
Shoreline
SGD
Saltwater
Freshwater
Saltwater
TYPE I
TYPE II
TYPE III
Figure 47: Conceptual models of SGD into Cockburn Sound
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 99 and Nutrient Discharge into Cockburn Sound
Cockburn Sound
Tamala
Limestone
Cretaceous Sediment
Quaternary
Sand
Shoreline
SGD
Saltwater
Offshore
SGD
Freshwater
Freshwater
Saltwater
TYPE IV
Cockburn Sound
Tamala
Limestone
Cretaceous Sediment
Quaternary
Sand
Shoreline
SGD
Saltwater
Offshore
SGD
Freshwater
Freshwater
Saltwater
TYPE V
Saltwater
Cockburn SoundTamala
Limestone
Cretaceous Sediment
Quaternary
Sand
Shoreline
SGD
Saltwater
Diffuse offshore SGD
Freshwater
Freshwater
Saltwater
TYPE VI
Saltwater
Salt-
waterSalt-
water
Figure 47: Conceptual models of SGD into Cockburn Sound…continued
CSIRO Land and Water
100 Coast and Clean Seas Project No. WA9911
4.2 PREVIOUS ESTIMATES OF SGD
Table 15 summaries previous estimates of SGD into Cockburn Sound. While some
of the tabulated values were reported directly in the referenced literature, others were
implied by the reported values of hydraulic gradient, hydraulic conductivity and
aquifer thickness. In these cases, SGD in Table 15 was calculated by applying
Darcy’s law.
Table 15: Estimates of Submarine Groundwater Discharge into Cockburn Sound and related sites
Reference SGD Rate [m3.d-1.m-1]
Location and Method Assumed Values
Allen (1981) ~ 8 (superficial.) North Perth coastline, flownet
-
Bodard (1991) 2.3 - 5.2 (superficial) 0.1 - 0.6 (sand) 2.2 - 4.6 (limestone)
BP Refinery, flownet Sand: v = 15 - 73, B = 10, n = 0.3; Limestone: v = 270 - 562, B = 10, n = 0.3
WAWA (1993) 3.0 (superficial) Cockburn Groundwater Area, flownet water balance (1992 scenario)
Hamilton Hill: k = 100 - 200; Henderson: k = 100 - 200; Kwinana Beach: k = 500 - 800
Appleyard (1994) 0.7 - 2.5 (superficial) 0.1 - 0.5 (sand) 0.6 - 2.0 (limestone)
Cockburn Sound, flownet k = 10 - 50, i = 0.001 - 0.003
Davidson (1995) 2.5 (superficial) Jandakot groundwater mound, regional flownet
T > 4,000
HGM (1998) 3.0 - 8.0 (superficial) Northern Harbour Jervoise Bay, offshore CTD profiling
-
Nield (1999) 2.3 - 4.4 (superficial) Cockburn Groundwater Area, model calibration on hydraulic head (1999 scenario)
k = 900
PPK (2000) 5.0 (mean) 3.0 - 8.0 (seasonal)
Northern Harbour Jervoise Bay, flownet
T = 25,000, i = 0.0002
PPK (2000) 3.0 - 9.0 (superficial) Northern Harbour, harbour salinity measurements
-
Symbols and Units v: groundwater velocity [m.d-1] n: porosity [1] B: saturated thickness of aquifer unit [m] k: saturated hydraulic conductivity [m.d-1] i: hydraulic gradient [1] T = kB: aquifer transmissivity [m2.d-1]
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 101 and Nutrient Discharge into Cockburn Sound
4.2.1 FLOWNET CALCULATIONS
Bodard (1991) estimated that groundwater pore velocities in the area of the BP
Refinery varied in the range 15 to 73 m.yr-1 in coastal sands and 270 to 562 m.yr-1 in
Tamala Limestone. Based on average saturated thickness of sand equal to 10 m,
average thickness of limestone equal to 10 m, and porosity 0.3 for sand and
limestone, the estimated groundwater velocities equate to SGD of 0.1 to
0.6 m3.d-1.m-1 from the sand and 2.2 to 4.6 m3.d-1.m-1 from the limestone. The sum
of these give total SGD from the Superficial Formations of between 2.3 and
5.2 m3.d-1.m-1.
WAWA (1993) estimated that total SGD into Cockburn Sound in 1992 was
75,342 m3.d-1. A flownet approach was used and a hydraulic conductivity value of
200 m.d-1 was assumed for the Superficial Formations. Distributed uniformly over
25 kilometres of coast, this is equivalent to an average SGD of around 3.0 m3.d-1.m-1.
Appleyard (1994) calculated that SGD into Cockburn Sound was in the range 0.1 to
0.5 m3.d-1.m-1 from superficial sands and 0.6 to 2 m3.d-1.m-1 from the Tamala
Limestone. Hydraulic conductivity values between 10 and 50 m.d-1 were assumed
and applied with gradients between 0.001 and 0.003 to compute SGD.
Based on groundwater flownet calculations, Davidson (1995) estimated that total
SGD along the coastline of the Jandakot mound was approximately 66,450 m3.d-1,
which equates to average SGD along the shoreline of around 2.5 m3.d-1.m-1.
Transmissivity of the Superficial Formations was assumed greater than 4,000 m2.d-1.
More recently, PPK (2000) estimated that average SGD into Northern Harbour was
approximately 5 m3.d-1.m-1 based on average transmissivity 25,000 m2.d-1 and
average hydraulic gradient 0.0002. The estimated seasonal range of SGD in that
study was 3 to 8 m3.d-1.m-1.
Thus, previous estimates of SGD from Superficial Formations, based on flownet
analyses, have varied between 0.7 and 8 m3.d-1.m-1; a factor of approximately ten. A
number of the estimates arrived at a similar final value of SGD; however, they were
based on significantly different values of hydraulic conductivity and hydraulic
gradient, which vary between some estimates by an order of magnitude. These
differences highlight the inherent uncertainty in flownet analyses that are based on
water level data alone.
CSIRO Land and Water
102 Coast and Clean Seas Project No. WA9911
4.2.2 CTD PROFILING
Halpern Glick Mansel (1998) used results of conductivity-temperature-depth (CTD)
profiling in Jervoise Bay to estimate that SGD into Northern Harbour was
approximately 3-8 m3.d-1.m-1. Also using CTD data, DAL (2000) estimated that
SGD into Northern Harbour was 3-9 m3.d-1.m-1. In both studies, SGD was estimated
to be the quantity of fresh groundwater that was required to dilute seawater in the
harbour to the observed values of salinity.
Because these estimates were derived without reference to aquifer hydrogeology,
they provide important independent checks on the values of SGD calculated from
inshore flownet analyses. Nevertheless, seawater inside Northern Harbour is not
uniformly diluted by groundwater and therefore the above estimates of SGD involve
assumptions about mixing of groundwater and seawater in the harbour and the
harbour-flushing rate. These assumptions have significant influence on the estimated
value of SGD.
4.3 DIRECT MEASUREMENT OF SGD
Direct measurements of SGD by BFM indicated that shoreline SGD in Cockburn
Sound is strongly influenced by sea level dynamics and associated fluxes of seawater
into and out of the seabed. Based on the experimental results described in Sections
3.1.6 and 3.3.4, and the modelling results in Appendix C, it is concluded that the
fresh-groundwater component of total benthic flux measured by BFM cannot be
distinguish from the seawater component.
It is also clear from EC surveys of submarine porewater (Sections 3.1.3, 3.2.2 and
3.3.2) that benthic flux is distributed non-uniformly over the seabed and appears to
be relatively ‘patchy’ in distribution. Even if it were possible to measure the fresh-
groundwater component of benthic flux, an impractical amount of data would need
to be collected to enable point measurements of SGD to be interpolated and
integrated to yield a value of total SGD into Cockburn Sound.
Direct, point measurements of SGD using benthic flux meters are of limited
usefulness to develop aggregated estimates of total SGD. Benthic flux
measurements are strongly influenced by seawater exchange between the seabed
sediments and ocean in response to short time-scale sea level variations such as tidal
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 103 and Nutrient Discharge into Cockburn Sound
fluctuations. In addition, the field measurements indicate that the distribution of
SGD is patchy at spatial scales of less than ten metres. To properly describe the
distribution of SGD across the study area and aggregate volumetric estimates of SGD
would require an impracticable number of measurements. However, significant
insight into the flow and discharge mechanisms of the SGD flux was obtained from
high temporal resolution measurements by benthic flux meters, indicating at the
study site that tidal forcing is the dominant factor controlling the SGD process at
Cockburn Sound. Geochemical and isotope based estimates of SGD on the other
hand provide a more reliable regional scale estimate of the net SGD flux per unit
length of shoreline, and these measurements results are more comparable to
estimates of SGD based on onshore numerical modelling of groundwater fluxes and
groundwater balance. These figures are the most useful to coastal zone
environmental managers with responsibilities for protecting the near-shore marine
environment from land-based contaminant sources.
It is concluded that terrestrial SGD into Cockburn Sound cannot be reliably
estimated by direct measurement techniques employed in this study.
4.4 ESTIMATION OF SGD BY INSHORE WATER BALANCE
The alternative to estimating SGD by direct and indirect measurement techniques is
to infer SGD from the inshore aquifer water balance.
This section describes the application of a groundwater flow model to quantify the
aquifer water balance inshore from Cockburn Sound and estimate SGD into
Cockburn Sound. Estimates of groundwater discharge are obtained from the model
water budget of the modeled domain as the residual between groundwater inflow, net
recharge and pumping. Simulated values of SGD vary spatially along the length of
the coastline due to (a) variation in hydraulic connection between the coastal geology
and ocean, (b) spatial variability in groundwater recharge, aquifer transmissivity and
pumping, and (c) focusing of groundwater flow toward coastal embayments.
4.4.1 GROUNDWATER FLOW MODEL
A pre-existing groundwater flow model (Nield 1999) was used to estimate SGD.
The model was originally constructed in MODFLOW (McDonald and Harbaugh 1988).
CSIRO Land and Water
104 Coast and Clean Seas Project No. WA9911
In this study it was implemented in Visual MODFLOW v.2.8.2 (Waterloo Hydrologic
2000). A summary of the model structure is presented below:
• The finite difference grid (Figure 48) consists of 26,596 cells, which are each 100
× 100 m.
• Superficial Formations inshore from Cockburn Sound are represented as a single-
layer, unconfined aquifer with impermeable base. A single-layer model was
justified on the basis that Tamala Limestone is highly transmissive and most
groundwater flow in the Superficial Formations is conducted laterally through
secondary porosity in the limestone. It was assumed that groundwater movement
in overlying sands is predominantly downward (vertical) into the limestone in
response to local rainfall recharge. Contours corresponding to bottom elevations
of the Superficial Formations (Figure 49) vary from around –12 m Australian
Height Datum (AHD) at the southeast corner of the model domain to
approximately –30 m AHD in the vicinity of Northern Harbour.
• A line of general head cells (Figure 48) was used to represent the coastal
boundary and conductance values were assigned that varied according to which
geological unit contacts the ocean. Table 16 lists the conductance and external
head values, which were determined during calibration (see Section 4.4.3).
Where Tamala Limestone is exposed along the shoreline, the assigned
conductance value was higher than at locations where the aquifer discharges to
the ocean through overlying, relatively lower-permeability sand. External head
along the coast was set equal to mean sea level of -0.05 m AHD.
• Constant head cells (Figure 48) were specified along the inland boundary. Head
was set equal to 20 m AHD along the northern portion of boundary,
corresponding to the 20-m water table elevation contour. Along the southern
portion of boundary, head was set equal to invert levels along the Peel Main
Drain.
• The north and south boundaries of the model lie along assumed lines of regional
groundwater flow and were specified as no-flow boundaries.
• More than 700 pumping wells are simulated in the model (Figure 50). Unless
recorded pumping data was available—data was available for less than 1% of
wells—the abstraction rates from bores were set equal to 80% of the licensed
groundwater allocation limit. Therefore, the modelled groundwater abstraction
reflects the total allocation allowance rather than total abstraction and is possibly
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 105 and Nutrient Discharge into Cockburn Sound
an overestimate of pumping withdrawals. Thus, SGD estimate is possibly
underestimated
Table 16: Cockburn Groundwater Area model: conductance and external head values for the general head coastal boundary cells
Low-recharge Scenario High-recharge Scenario Conductivity Zone
Geological Unit Conductance
[m.d-1] External Head [m]
Conductance [m.d-1]
External Head [m]
1 Safety Bay Sand 500 -0.05 1000 -0.05 2 Tamala Limestone 1000 -0.05 2000 -0.05
4.4.2 GROUNDWATER RECHARGE
Recharge is the key water balance component that affects the model estimates of
SGD; however, there is significant uncertainty in quantifying recharge. Low
groundwater recharge implies less groundwater flow through the aquifer and low
SGD, while high recharge implies larger SGD. Two different scenarios were
simulated to provide estimates of SGD under assumptions of low and high average
groundwater recharge. Spatial distributions of recharge were assigned according to
land use and vegetation cover, including the categories: bushland, cleared or open
bushland, agriculture and horticulture, and industrial and urban. Each model cell
was assigned a land-use code and each land-use code assigned a groundwater
recharge rate. The resulting recharge distributions for low-recharge and high-
recharge scenarios are depicted in Figure 51 and Figure 52, respectively.
Recharge rates assigned to the different land use categories varied between 0% and
30% of mean annual rainfall in the low-recharge scenario, and between 0% and 40%
in the high-recharge scenario. The equivalent average recharge rates applied over the
entire model domain were 0.13 m.yr-1 (i.e. 14.7% of mean annual rainfall) and
0.24 m.yr-1 (i.e. 27.5% of mean annual rainfall), respectively.
4.4.3 CALIBRATION
Nield (1999) calibrated low-recharge and high-recharge scenarios by trial and error.
Hydraulic conductivity of each zone in Figure 48 was adjusted to obtain suitable
CSIRO Land and Water
106 Coast and Clean Seas Project No. WA9911
matches between the observed and modelled hydraulic heads. The calibrated values
of hydraulic conductivity are presented in Table 17 and model-simulated water table
contours are depicted in Figure 53. Conductance values of the coastal boundary cells
(Table 16) were also adjusted during calibration.
Table 17: Cockburn Groundwater Area model: calibrated hydraulic conductivity values
Hydraulic Conductivity [m.d-1] Conductivity Zone Low-recharge
Scenario High-recharge
Scenario
1 400 800 2 1660 3000 3 540 1000 4 500 900 5 25 30 6 6 8 7 34 34 8 20 20 9 120 120
10 25 35 11 10 14 12 30 30 13 60 60
4.4.4 RESULTS
SGD from each coastal boundary cell is depicted in Figure 53 for both low-recharge
and high-recharge scenarios. The modelled distributions of SGD along the coast
varied mainly in response to change in hydraulic connection between the aquifer and
ocean. SGD was largest along two sections of coastline, north and south of
Woodman Point, where Tamala Limestone directly contacts the ocean. Smaller
variations in SGD occurred in response to variability in recharge intensity inshore
from the coast, and variation in the density and distribution of pumping bores. In the
low-recharge scenario, SGD from the coastal boundary cells varied spatially between
1.4 and 4.6 m3.d-1.m-1. This increased to between 2.4 and 7.9 m3.d-1.m-1 in the high-
recharge scenario.
The model water budgets for both recharge scenarios are presented in Table 18 and
Table 19. They confirm that SGD and groundwater pumping (out flows) were
balanced by groundwater recharge and flow across the inland boundary (in flows).
Because all terms in the water balance are similar order of magnitude, uncertainty
and errors in estimations of groundwater inflow, recharge and pumping are all
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 107 and Nutrient Discharge into Cockburn Sound
manifest as similar order-of-magnitude error and uncertainty in SGD. Total SGD in
the low-recharge scenario was 48% of total groundwater out flow from the model
and groundwater abstraction from bores was approximately 46%. The remaining 6%
of out flow was minor discharges to the Peel Main Drain. In the high-recharge
scenario, SGD increased to 64% of total out flow and pumping reduced to 32%.
Based on these estimates, none of the water balance terms can be neglected as
insignificant.
Allowing for low and high estimates of recharge, and arbitrary ±20% errors in
pumping and groundwater flow across the inland boundary, mean SGD is estimated
to be in the range 2.5-4.8 ± 0.9 m3.d-1.m-1. Thus, total average SGD could be
between 1.6 and 5.7 m3.d-1.m-1, which is a factor of variation of approximately 3.5.
This order-of-magnitude uncertainty stems from uncertainty in the reliability and
accuracy of the available hydrologic and hydrogeologic data.
Table 18: Cockburn Groundwater Area model: model water balance for low-recharge scenario
Balance Term Flow In [Kl.d-1]
Flow Out [Kl.d-1]
Net Flow [Kl.d-1]
Inland boundary 56,982 -6,855 50,127 Groundwater recharge 65,288 0 65,288 Groundwater Abstraction 0 -58,245 -58,245 Artificial recharge 3,836 0 3,836 SGD 0 -61,007 -61,007
Total 126,106 -126,107 -1
Mean SGD = 61,007/24,000 = 2.5 m3.d-1.m-1 of shoreline
Table 19: Cockburn Groundwater Area model: model water balance for high-recharge scenario
Balance Term Flow In [Kl.d-1]
Flow Out [Kl.d-1]
Net Flow [Kl.d-1]
Inland boundary 55,186 -7,046 48,140 Groundwater recharge 121,750 0 121,750 Groundwater Abstraction 0 -58,245 -58,245 Artificial recharge 3,836 0 3,836 SGD 0 -115,480 -115,480
Total 180,772 -180,771 1
Mean SGD = 115,480/24,000 = 4.8 m3.d-1.m-1 of shoreline
CSIRO Land and Water
108 Coast and Clean Seas Project No. WA9911
4.4.5 CONCLUSIONS
Based on aquifer water-balance considerations, total average SGD across the
coastline adjacent to Cockburn Sound is estimated to be 2.5-4.8 ± 0.9 m3.d-1.m-1.
Most of this groundwater is believed to discharge to the ocean via preferred
pathways in Tamala Limestone. The estimated range in SGD, 2.5-4.8 m3.d-1.m-1,
corresponds to low and high estimates of average groundwater recharge. The error,
± 0.9 m3.d-1.m-1, is calculated based on arbitrary ±20% errors in pumping and
groundwater inflow across boundaries. The magnitude of this error could be larger.
In the modelling, differences in hydraulic connection between the aquifer and ocean
were observed to have the largest affect on the spatial variability of SGD.
Nevertheless, the modelled distributions of SGD cannot be corroborated against
independent data—since none exists—and are considered to have large uncertainty.
For example, the model can be successfully calibrated using conductance values
along the coast that result in significantly different spatial distributions of SGD,
compared to those depicted in Figure 53. The model provides regional-scale
estimates of SGD that do not take into account the detailed hydrogeology, and
associated flow paths and rates that control SGD at the local scale.
It is concluded that the estimated values of total average SGD are consistent with the
available regional hydrogeological and hydrologic data for the study area.
Nevertheless, results from local-scale measurements are affected by local-scale
processes and should not be constrained by regional-scale estimates derived from the
modelling. Focusing of flow through preferred pathways and tidal effects, for
example, may results in local SGD significantly larger or smaller than the regional-
scale estimates.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 109 and Nutrient Discharge into Cockburn Sound
1
3 4
9
7
13
13
56
2
2
1
1
4
1011 8
813
13
13
13
12
13
13
13
13
0 2,500 5,0001,250Meters
�
LegendHydraulic conductivity zoneModel gridGeneral head cellDrain cell (set to drain invert levels)Constant head cell (20 mAHD)Inactive cell
COCKBURNSOUND
Figure 48: Cockburn Groundwater Area flow model; finite difference grid, boundary conditions and hydraulic conductivity zones
CSIRO Land and Water
110 Coast and Clean Seas Project No. WA9911
-25
-24
-23
-22
-21
-20
-19
-28
-29
-30
-17
-15
-14
-13
-12
-17
-20
0 2,500 5,0001,250Meters
�
LegendBasement elevation contour [mAHD]General head cellDrain cell (set to drain invert levels)Constant head cell (20 mAHD)Inactive cell
COCKBURNSOUND
Figure 49: Cockburn Groundwater Area flow model; base of Superficial Formations
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 111 and Nutrient Discharge into Cockburn Sound
537
809
717
745
835
699
3247
1800
4540
1403
4186
1753
3206 1534
1206
16774110
1808
0 2,500 5,0001,250Meters
�
LegendBore abstraction [Kl/day]
0 - 8788 - 395396 - 835836 - 1,808
1,809 - 4,540
Artificial recharge [Kl/day]274411General head cellDrain cell (set to drain invert levels)Constant head cell (20 mAHD)Inactive cell
COCKBURNSOUND
Figure 50: Cockburn Groundwater Area flow model; groundwater abstraction and artificial recharge locations and rates
CSIRO Land and Water
112 Coast and Clean Seas Project No. WA9911
0 2,500 5,0001,250Meters
�
LegendLow Recharge ModelPercent rainfall recharge
0%10%20%25%30%General head cellDrain cell (set to drain invert levels)Constant head cell (20 mAHD)Inactive cell
COCKBURNSOUND
Figure 51: Cockburn Groundwater Area flow model; groundwater recharge rates for the low-recharge scenario, recharge is expressed as a percentage of mean annual rainfall of 0.87 m/yr
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 113 and Nutrient Discharge into Cockburn Sound
0 2,500 5,0001,250Meters
�
LegendHigh Recharge ModelPercent rainfall recharge
0%22%32%40%General head cellDrain cell (set to drain invert levels)Constant head cell (20 mAHD)Inactive cell
COCKBURNSOUND
Figure 52: Cockburn Groundwater Area flow model; groundwater recharge rates for the high-recharge scenario, recharge is expressed as a percentage of mean annual rainfall of 0.87 m/yr
CSIRO Land and Water
114 Coast and Clean Seas Project No. WA9911
0
2
3
1
4
5
6
78
9
0.5
1011
1213
14 15
16
17
18 19
20
0 2,000 4,0001,000Meters
�
LegendLow RechargeHigh RechargeGeneral head cellDrainConstant headInactive cell
COCKBURNSOUND
0246810
SGD [Kl/d/m]
Figure 53: Cockburn Groundwater Area flow model; calibrated water table elevations and model computed SGD for the low-recharge and high-recharge scenarios
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 115 and Nutrient Discharge into Cockburn Sound
4.5 ESTIMATION OF SGD BASED ON SALTWATER-FRESHWATER RELATIONS
Figure 54 depicts saltwater interface profiles that were calculated using the abrupt
interface model of Glover (1959). Glover’s solution expresses the vertical depth, z
[L], to the saltwater interface as a function of the horizontal distance, x [L], inshore
from the coast
2
2 2⎟⎠⎞
⎜⎝⎛+=
αα K
Q
K
Qxz (3)
The solution also incorporates groundwater discharge, Q [L2/T], to the seawater
boundary and assumes that the aquifer is isotropic with hydraulic conductivity K and
infinitely thick. The term α is a linear expansion coefficient that relates salinity to
density and has a constant a value of 0.00225 [-] for a freshwater aquifer in contact
with the sea. The saltwedge geometry therefore depends on the value of Q/K, which
is the ratio of SGD to aquifer hydraulic conductivity. By applying Darcy’s Law, Q =
KBi, it can be seen that Q/K = Bi, where B is the saturated thickness of aquifer and i
is the hydraulic gradient.
In the aquifer adjacent to Cockburn Sound the saltwedge extends up to 2,000 m
inshore, indicating that the value of Q/K is relatively small; a value of around .0025
is predicted from Figure 54. For K in the range 500-2,000 m.d-1, Glover’s abrupt
interface solution predicts SGD (Q) in the range 1.25-5 m3.d-1.m-1.
CSIRO Land and Water
116 Coast and Clean Seas Project No. WA9911
Figure 54: Abrupt saltwater interface profiles from the solution of Glover (1959)
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 117 and Nutrient Discharge into Cockburn Sound
4.6 SUMMARY
Conceptual models of SGD into Cockburn Sound based on the available regional-
scale hydrogeological information indicate that groundwater is likely to discharge
both along the shoreline and offshore from the coast. SGD along the shoreline has
been detected by field measurements (e.g. Section 3.1.3) and it is probable that
groundwater also discharges tens to hundreds of metres offshore via high-
conductivity zones in Tamala Limestone (e.g. Section 3.1.8). Accurate patterns of
SGD at local scales of interest cannot be determined because they depend upon
detailed, local-scale hydrogeology, which is not described across the study area. In
particular, there is virtually no offshore hydrogeological data in Cockburn Sound.
Direct measurements of SGD using benthic flux meters are not useful for
aggregating estimates of total SGD. Benthic flux measurements are strongly
influenced by seawater exchange between the coastal sediments and ocean in
response to sea level variation. In addition, field measurements indicate that the
distribution of SGD is patchy at spatial scales of less than ten metres. To properly
describe the distribution of SGD across the study area and aggregate volumetric
estimates of SGD would require an impracticable number of measurements.
However, BFM’s that measure SGD fluxes over short time scales of seconds to
minutes are very useful for understanding the SGD flux transfer mechanisms at the
sediment-water interface.
Based on groundwater flow modelling and aquifer water-balance considerations,
total average SGD across the coastline adjacent to Cockburn Sound is estimated to
be 2.5-4.8 ± 0.9 m3.d-1.m-1. The range, 2.5-4.8 m3.d-1.m-1, is derived from low and
high estimates of average groundwater recharge; the error, ± 0.9 m3.d-1.m-1, is
calculated based on arbitrary ±20% errors in pumping and groundwater inflow across
boundaries. Calibrated values of hydraulic conductivity from the flow modelling,
and the simulated rates of SGD, imply a saltwater-freshwater interface position that
is consistent with field observations.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 119 and Nutrient Discharge into Cockburn Sound
5 QUANTIFYING NUTRIENT DISCHARGE INTO COCKBURN
SOUND
5.1 MECHANISMS FOR NUTRIENT DISCHARGE
Nutrients in the inshore aquifer can enter Cockburn Sound via two mechanisms:
• SGD of terrestrial groundwater, and
• Aquifer ‘flushing’ by seawater
The first case represents simple advection by groundwater flow of dissolved, mobile
nutrients that enter the aquifer between the coast and top of the regional flow system.
The un-utilised mobile fraction will eventually discharge into Cockburn Sound along
with the regional groundwater flow. Nutrient transport pathways and travel times
along these pathways—from source to sea—are determined by the spatial and
temporal patterns of groundwater recharge and flow. Nutrients are mobilized by
local recharge and transported toward the coast by regional groundwater flow, which
is dominated by channelised flows through high-conductivity pathways. Nutrients
that are mobilized into more active, or faster flowing parts of the regional flow
system will reach Cockburn Sound more quickly than nutrient that are mobilized into
relative inactive parts of the flow system. Thus, transport times are non-trivial to
calculate since SGD estimates based on the inshore water balance do not account for
preferred flow paths through the Tamala Limestone. In addition, the scale of aquifer
heterogeneity and its relation to the scale of contaminant sources inshore from
Cockburn Sound has not been investigated and is not known.
The second mechanism for groundwater nutrient discharge into Cockburn Sound has
not been investigated. It is probable that seawater movement into and out of coastal
sediments plays a role in transporting nutrients from subsurface to submarine
environments. Seawater entering the aquifer during periods of rising sea level can
contact and mix with nutrient enriched groundwater. Some of this water will drain
back to the ocean during a subsequent period of falling sea level and transport
nutrients with it. The longer the period of sea level variation, the further into the
aquifer this effect might propagate. The effect may be important in locations where
the aquifer beneath a contaminant source is regularly flushed by saltwater flows.
CSIRO Land and Water
120 Coast and Clean Seas Project No. WA9911
5.2 NITROGEN CONCENTRATION IN GROUNDWATER
Table 20 and Figure 55 present summaries of groundwater nutrient data inshore from
Cockburn Sound. Nitrogen is the main nutrient of concern and has been monitored
more intensely than phosphorus. Measured phosphorus concentrations in
groundwater are typically several orders of magnitude less than nitrogen. Significant
point source emissions of phosphorus in groundwater have not been detected to date.
Figure 55 is a summary of last-recorded, total nitrogen concentrations in
approximately 200 bores across the study area. This information was collated from
several sources, which are listed on Figure 55, and spans the ten-year period from
1991 to 2000. Bores that were sampled by Appleyard (1994) to estimate nutrient
discharge into Cockburn Sound are labeled using the notation from that study.
Because the data are not contemporaneous, they provide a mostly qualitative
impression of past and recent distributions of nitrogen in groundwater. Based on the
calculated SGD rates, groundwater is expected to travel hundreds of metres through
the aquifer each year and, therefore, quantitative analysis of nutrient data that was
collected in different years is not appropriate.
Two prominent and well-known areas of nitrogen contamination of groundwater are
indicated in Figure 55: (a) inshore from Northern Harbour, and (b) south of James
Point. PPK (1999) investigated two nutrient-enriched groundwater plumes
discharging into Northern Harbour that were found to emanate from the sludge
drying beds at Woodman Point Wastewater Treatment Plant, and from a wastewater
injection bore operated under license by Weston Bioproducts. Both operations were
recently decommissioned and remediation options for the plumes were investigated
by PPK (2000). Recovery of the plumes by groundwater pumping commenced in
December 2000 and is currently ongoing.
Nutrient enrichment of groundwater south of James Point is the direct result of
contamination by industry. CSPB produces nitrogen and phosphate fertilizers, and
ammonium sulphate is produced by WMC as a by-product of nickel refining.
Other potential sources of nitrogen in groundwater discharging into Cockburn Sound
include fertilizers applied for market gardening, horticulture and domestic lawns, and
leachate from septic tanks in residential areas.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 121 and Nutrient Discharge into Cockburn Sound
Table 20: Summary of nutrient concentrations in groundwater in Cockburn Sound area and related sites
Reference Nutrient Concentration Location and Method
Bodard (1991) NO3 = 1.2, 2.5 (sand) NO3 = 10, 20 (limestone)
BP Refinery; representative values Bore screens: 2-14 mBGL sand 16-35 mBGL limestone
Davidson (1995) NO3 > 1 Swan Coastal Plain; groundwater samples from bores in the superficial aquifer
Appleyard (1994) N-NO3 = <0.2 – 160 (sand) N-NH3 = <0.02 - 394 (sand) N-total = 0.9 - 394 (sand) P-total = <0.01 - 0.74 (sand) N-NO3 = <0.2 – 1.6 (limestone) N-NH4 = 0.2 - 383 (limestone) N-total = 1 - 383 (limestone) P-total = 0.02 - 0.47 (limestone)
Cockburn Sound; groundwater samples from 39 bores in the superficial aquifer
PPK (1999) N-total = 8 Inshore from Northern Harbour beneath market garden area
PPK (1999) N-NO3 = 0.5 - 25 (WPMC5) N-NO3 = 26 (JBMB1) N-total = 0.5 – 25 (15 mean)
Woodman Point Waste Water Treatment Plant plume into Northern Harbour, bores WPMC5 and JBMB1
PPK (1999) N-NH3 = 22 - 252 (OB7) N-NH3 = 21 - 62 (JBMB7) N-total = 22 – 279 (128 mean)
Weston Bioproducts plume into Northern Harbour, bores OB7 and JBMB7
PPK (2000) N-NO3 = <0.05 – 64.5 N-NH4 = <0.1 – 210 N-DIN = 0.6 – 210
Inshore from Northern Harbour; groundwater samples from 28 bores
This Report N-NO3 = <0.0 – 21.0 N-NH4 = <0.1 – 130
Inshore from Northern Harbour; groundwater samples from 22 bores (see Table 7.
All concentrations are mg.l-1
CSIRO Land and Water
122 Coast and Clean Seas Project No. WA9911
�
��� � ��� � ��� �������
�������������������������������������������� ����� ����!���""#�$$��������������!�%�&�����������""'��(���� ����� ��!�)��*+,��(�����(���� ����� ��!�)��-$��(�����(���� ����� ��!�)��,�-$��(���� ����� ��!�)��.�/���(���� ��������!�)��,0*��(����
����
����
��
��
������������
��������
��������
������������������
��������������
����
������������
����������������������������
����������������
��������
����
����������������������������������������
����������������������������������������������������������������
��
��
������������������
��������
��������
��
��
��
����
����������
��
����
��
��
����������
�������������� ����
��
��
����������
����������
��
��
��
����������
��
��
������������
����������������
��
��
����
��
����
������������
��������
�� ����
�����������������������
�������������
*�,1��2�+&����
,�-$
*��!���$���**.$
*���-�� ��!���
��������������
��������
������������
��������
���
��� ����
���
������ ����������������������� ������� � ���������������������������������� ��� �
�����������
���
����
.���������!/���1������&����-���3����
��4�����
���4��������4������5������
��������������������&��-������!0��������-����!�.������0��������!��.���������!������������/�� �! ���
������
������
������
������
������
������
������
������� �������
������� �������
������� �������
������� �������
������� �������
Figure 55: Last-recorded, total nitrogen concentrations in groundwater [mg.l-1]; data spans the ten-year period from 1991 to 2000; groundwater bores sampled by Appleyard (1994) are labeled using notation from that study
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 123 and Nutrient Discharge into Cockburn Sound
5.3 SIGNIFICANCE OF MEASURED CONCENTRATIONS
It is difficult to assess the significance of groundwater nutrient concentrations in
isolation from an assessment of the marine, near-shore nutrient budget. The
ecological and environmental implication of nutrient-enriched groundwater
discharge into Cockburn Sound depends not only on the nutrient concentrations in
groundwater but also on the rate of groundwater discharge (i.e. the nutrient mass
flux), the relative sizes of other nutrient inputs, the sizes of the extant nutrient
‘pools’ in the receiving environment, and the dynamics of nutrient cycling. Such an
assessment is beyond the scope of this report.
As a guide, Table 21 presents default trigger values of nutrients and pH for slightly
disturbed marine ecosystems in south-west Australia. The values are taken from the
“Australian and New Zealand Guidelines for Fresh and Marine Water Quality”
(ANZECC, 2000). Default trigger values represent default concentrations that
should trigger further investigation to determine whether there is a genuine risk to
the ecosystem. Note that trigger concentrations apply to the receiving waters; they
are not discharge water-quality guidelines.
Table 21: Default trigger values for physical and chemical stressors for south-west Australia for slightly disturbed inshore marine ecosystems (ANZECC, 2000); inshore waters are defined as coastal lagoons (excluding estuaries) and embayments and waters less than 20-m depth
Water Quality Measure Default Trigger Value
Total nitrogen (N-total) 0.23 mg.l-1 Oxides of nitrogen (N-NO3, N-NO2) 0.005 mg.l-1 Ammonium (N-NH4) 0.005 mg.l-1 Total phosphorus (P-total) 0.02 mg.l-1 pH 8.0-8.4
1. This table is adapted from Table 3.3.6 of ANECC (2000) 2. Guideline trigger values represent bioavailable concentrations or
unacceptable levels of contamination that trigger the incorporation of additional information or further investigation to determine whether a real risk to the ecosystem exists.
The Water and Rivers commission’s Water Quality Protection Note for discharge of
groundwater from aquifer dewatering operations (WRC, 1999) provides a table of
water-quality protection criteria for receiving water bodies. The relevant water
quality indicators and criterion are reproduced in Table 22. For the same reasons as
discussed above, the criterion in Table 22 relate to acceptable levels of perturbation
CSIRO Land and Water
124 Coast and Clean Seas Project No. WA9911
of background conditions in the receiving water body rather than absolute
concentrations in the discharge water.
Table 22: Receiving water quality protection criteria for discharge of groundwater from aquifer dewatering (WRC, 1999)
Indicator Criterion
Nutrients
Water discharges should not add nutrients (e.g. phosphorus, nitrogen) in quantities sufficient to cause excessive or nuisance algal growth in the receiving water. The water discharged should not cause the seasonal background nutrient levels in the receiving water body to increase by more than 10%.
pH Water discharged should not cause the receiving water’s seasonal background pH to vary by more than ± 0.2 units.
1. This table is adapted from Table 2 of WRC (1999)
Perhaps a more useful approach is to compare the measured nutrient concentrations
in submarine porewater to typical nutrient concentrations in shallow groundwater at
other locations on the Swan Coastal Plain. Nitrogen is usually present in
groundwater as nitrate (NO3) and levels in urbanized areas and intensive horticultural
areas can be relatively high due to leachate from septic tanks and nitrogen fertilizers.
Nitrate levels within urbanized and cultivated areas vary depending on the type of
land use but commonly exceed 20 mg.l-1 (Davidson, 1995). Elsewhere, nitrate
concentrations in shallow groundwater are typically less than 1 mg.l-1.
Background ammonia concentrations in groundwater are normally small (less than
nitrate) because the ammonium is oxidized to nitrate before reaching the water table.
In contrast, nutrient contaminated porewater samples from Cockburn Sound contain
high concentrations of ammonia (10’s-100’s of milligrams per litre) and relatively
small nitrate concentrations. This is indicative of groundwater that has been directly
polluted by ammonia-based contaminants.
As a general guide for the metropolitan area, a total nitrogen concentration in
groundwater greater than 5 mg.l-1 is indicative of nutrient enrichment due to human
activities. Depending on the rate of discharge, and the characteristics of the receiving
water body, discharge of nutrient enriched groundwater may, or may not, have the
potential to impact on the aquatic ecosystem.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 125 and Nutrient Discharge into Cockburn Sound
5.4 PREVIOUS ESTIMATES OF NUTRIENT DISCHARGE
Estimates of nitrogen discharge into Cockburn Sound by SGD during the period
1978 to 2000 have varied approximately in the range 490 to 1240 kg.d-1 (180 to
450 tonnes.yr-1). Previous estimates are listed in Table 23. Most estimates prior to
2000 were in the order of 900 kg.d-1 (330 tonnes.yr-1) and the latest estimate by D.A.
Lord (2000) was approximately 600 kg.d-1 (220 tonnes.yr-1).
Appleyard (1994) estimated that the phosphorus load into Cockburn Sound at that
time was approximately 6 kg.d-1 (2 tonnes.yr-1), which is several orders of magnitude
less than for nitrogen.
Despite the considerable scope for errors in the above estimates, uncertainty is not
usually addressed. Both the calculation of SGD rates (e.g. Section 4.2) and spatial
interpolation of groundwater nutrient data can introduce large errors in estimated
values of nutrient mass flux. For example, a simple doubling of hydraulic
conductivity has the effect of doubling both the SGD rate and associated nutrient
load. Further discussion on this topic is presented in Section 5.7.
CSIRO Land and Water
126 Coast and Clean Seas Project No. WA9911
Table 23: Estimates of nutrient discharge into Cockburn Sound via SGD
Reference Nutrient Discharge Location and Method Assumed Values
Appleyard (1990) in Appleyard (1994)
N-flux = 547 kg.d-1 P-flux = 11 kg.d-1
Coastline between Cape Peron and Fremantle
-
Mackie Martin (1992) in Appleyard (1994)
N-flux = 959 kg.d-1 As above -
Appleyard (1994) N-flux = 909 kg.d-1 (superficial) N-flux = 159 kg.d-1 (sand) N-flux = 751 kg.d-1 (limestone) N-flux = 0.0001-0.1 kg.d-1.m-1 (sand) N-flux = 0.0005-0.7 kg.d-1.m-1 (limestone) P-flux = 5.5 kg.d-1 (superficial) P-flux = 2.7 kg.d-1 (sand) P-flux = 2.7 kg.d-1 (limestone)
Cockburn Sound, flownet coupled with groundwater nutrient concentrations from 39 bores
k = 10-50, i = 0.001-0.003
DEP (1996) N-flux = 940 kg.d-1 (superficial)
Cockburn Sound -
PPK (2000) N-flux as DIN = 74 kg.d-1 (superficial)
Weston Bioproducts plume into Northern Harbour
PPK (2000) N-flux as DIN 34 kg.d-1 (superficial)
Woodman Point Waste Water Treatment Plant plume into Northern Harbour
PPK (2000) N-flux as DIN = 50 kg.d-1 (superficial)
Regional (market garden) inputs into Northern Harbour; excluding plumes
Measured nutrient concentrations in groundwater from 28 bores
D.A. Lord (2001) N-flux 1978 = 493 kg.d-1 (superficial) N-flux 1990 = 1243 kg.d-1 (superficial) N-flux 2000 = 600 kg.d-1 (superficial)
Cockburn Sound -
Symbols and Units k: saturated hydraulic conductivity [m.d-1] i: hydraulic gradient [1]
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 127 and Nutrient Discharge into Cockburn Sound
5.5 ESTIMATION OF NUTRIENT DISCHARGE BASED ON SGD RATES FROM THE
INSHORE WATER BALANCE MODEL AND MEASURED NUTRIENT
CONCENTRATIONS IN SUBMARINE POREWATER
This sections presents estimates of nutrient mass flux into Cockburn Sound based on
SGD rates estimated from the inshore water balance model (Section 4.4) and
groundwater nutrient concentrations derived from the submarine porewater survey
along Cockburn Sound foreshore (Section 3.4).
5.5.1 METHOD
An initial task was to assign groundwater nutrient concentrations to each 100-m
section of coastline corresponding to the 100-m-wide coastal boundary cells of the
model. This was carried out for the section of coastline from Woodman Point to the
southern boundary of the model (i.e. for 137 boundary cells). The estimates of
groundwater nutrient concentrations developed in Section 3.4 were spatially
interpolated. An average value (arithmetic mean) was assigned where more than one
porewater sample had been collected within the section of coastline spanned by a
particular boundary cell. For cells in which no porewater sample had been collected,
a groundwater concentration was assigned subjectively based on the nearest
estimated concentrations. Note that the porewater sampling density was increased in
areas of known groundwater nutrient contamination (Figure 41). Note, also, that
some sections of the coast could not be accessed for porewater sampling (Figure 41).
The total nutrient mass flux from each 100-m section of coastline was calculated by
multiplying the model-simulated values of SGD [L3.T-1] by the assigned
groundwater nutrient concentrations in each coastal boundary cell
gQ NSGDN ×= (4)
where: NQ is the nutrient mass flux in groundwater [M.T-1] and Ng is the nutrient
concentration in groundwater [M.L-3]. Therefore, the total nutrient load NT [M] from
the modelled-section of coastline during time period ∆t is
∑=
∆=n
iQiT NtN
1
(5)
where n is the number of coastal cells.
CSIRO Land and Water
128 Coast and Clean Seas Project No. WA9911
5.5.2 RESULTS
Figure 56 and Figure 57 depict model-simulated SGD for the low-recharge and high-
recharge scenarios, respectively. Figure 58 and Figure 59 show the estimated
nitrogen loads into Cockburn Sound for these estimates of SGD, and Figure 60 and
Figure 61 depict the corresponding phosphorus loads. Solid bars indicate nutrient
loads estimated from measured porewater concentrations and hatched bars indicate
interpolated values. Total nitrogen discharge into Cockburn Sound was estimated to
be 394 kg.d-1 (144 tonnes.yr-1) for the low-recharge scenario and 836 kg.d-1
(305 tonnes.yr-1) for the high-recharge scenario. Because total SGD for the high-
recharge scenario was approximately twice as large as total SGD for the low-
recharge scenario, the estimated total nitrogen load was also approximately doubled.
A break down of nitrogen loads along nominated sections of the coastline is
presented in Table 24. As identified by Appleyard (1994), a large portion of the total
nutrient load (approximately 75% in the current estimate) is discharged from the
coastline south of James Point. Approximately 10% of estimated nitrogen discharge
was from the short section of shoreline north of James Point and around 7% was
discharged from the area adjacent to the Naval Base caravan Park.
Table 24: Nitrogen loads for nominated sections of coastline
Nitrogen Load
Scenario Northern Harbour
[tonnes.yr-1]
Naval Base [tonnes.yr-1]
James Point North
[tonnes.yr-1]
James Point South
[tonnes.yr-1]
Total [tonnes.yr-1]
Low-recharge 6 (4%) 10 (7%) 15 (10%) 106 (73%) 144 High-recharge 10 (3%) 20 (7%) 34 (11%) 224 (74%) 305
1. Northern Harbour: 1.7 km length (6443650-6441950 mN AMG84) 2. Naval Base: 2.0 km length (6439950-6437950 mN AMG84) 3. James Point North: 1.0 km length (6435450-6434450 mN AMG84) 4. James Point South: 3.0 km length (6433950-6430950 mN AMG84)
Nutrient enriched porewater was not detected in Northern Harbour, with this section
of coastline contributing only 6-10 tonnes.yr-1 (4%) of the total nitrogen load. This
compares to estimates of around 70 tonnes.yr-1 (30%) in Appleyard (1994) and
approximately 58 tonnes.yr-1 in PPK (2000). Thus, the Northern Harbour plumes
probably discharge further offshore, beyond the shoreline strip in which the
porewater samples were collected.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 129 and Nutrient Discharge into Cockburn Sound
Figure 56: Model simulated SGD for low-recharge scenario
CSIRO Land and Water
130 Coast and Clean Seas Project No. WA9911
Figure 57: Model simulated SGD for high-recharge scenario
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 131 and Nutrient Discharge into Cockburn Sound
Figure 58: Estimated nitrogen load into Cockburn Sound based on model simulated SGD (low-recharge scenario) and groundwater nutrient concentrations estimates from porewater sampling; solid bars indicate measured concentrations; hatched bars indicate interpolated concentrations
CSIRO Land and Water
132 Coast and Clean Seas Project No. WA9911
Figure 59: Estimated nitrogen load into Cockburn Sound based on model simulated SGD (high-recharge scenario) and groundwater nutrient concentrations estimates from porewater sampling; solid bars indicate measured concentrations; hatched bars indicate interpolated concentrations
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 133 and Nutrient Discharge into Cockburn Sound
Figure 60: Estimated phosphorus load into Cockburn Sound based on model simulated SGD (low-recharge scenario) and groundwater nutrient concentrations estimates from porewater sampling; solid bars indicate measured concentrations; hatched bars indicate interpolated concentrations
CSIRO Land and Water
134 Coast and Clean Seas Project No. WA9911
Figure 61: Estimated phosphorus load into Cockburn Sound based on model simulated SGD (high-recharge scenario) and groundwater nutrient concentrations estimates from porewater sampling; solid bars indicate measured concentrations; hatched bars indicate interpolated concentrations
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 135 and Nutrient Discharge into Cockburn Sound
5.6 ESTIMATION OF NUTRIENT DISCHARGE BASED ON SGD RATES FROM THE
INSHORE WATER BALANCE MODEL AND GROUNDWATER NUTRIENT
CONCENTRATIONS MEASURED BY APPLEYARD (1994)
Appleyard (1994) measured contemporaneous nutrient concentrations in 39 bores
along Cockburn Sound foreshore (Figure 55). Thirty-one of these bores were along
the section of coastline considered above; 17 in the Safety Bay Sand and 14 in the
Tamala Limestone. Here, nutrient discharge into Cockburn Sound is estimated based
on SGD rates from the inshore water balance model (Section 4.4) and the
groundwater nutrient concentrations measured by Appleyard (1994).
5.6.1 METHOD
The following methodology was used:
1. Nutrient discharges from Safety Bay Sand and Tamala Limestone were estimated
separately and then summed to calculate the total nutrient load.
2. It was assumed that 85% of total SGD was from Tamala Limestone and 15%
from Safety Bay Sand. This rough estimate is subject to a large uncertainty.
There are several difficulties with this approach, which are discussed in more
detail in Section 5.7.
3. The nutrient concentrations measured in groundwater bores were assigned
uniformly throughout the aquifer within the bores’ discharge widths (length of
coastline); these were chosen to match those used by Appleyard (1994).
Groundwater concentrations in Safety Bay Sand were based on the 17 shallow
bores installed into sand and concentrations in Tamala Limestone were based on
the 14 bores installed into limestone.
4. Nutrient discharge from each 100-m section of coast was calculated as follows
)85.0()15.0( TLSBSQ NSGDNSGDN ××+××= (6)
where NSBS and NTL are the assigned groundwater nutrient concentrations [M.L-3]
in Safety Bay Sand and Tamala Limestone, respectively.
CSIRO Land and Water
136 Coast and Clean Seas Project No. WA9911
5.6.2 RESULTS
Figure 62 and Figure 63 depict the estimated nitrogen loads into Cockburn Sound
and Figure 64 and Figure 65 show the corresponding phosphorus loads. The red bar
on each plot near to Northern Harbour indicates nitrogen and phosphorus loads that
were calculated in a previous study by Mackie-Martin PPK (1993). These values
were adopted and used in the analysis by Appleyard (1994) and are included here for
consistency. They are estimates of nutrient discharge into Northern Harbour from
Weston Bioproduct’s (formerly NB Love Starches) wastewater injection operation.
Total nitrogen discharge into Cockburn Sound was estimated to be 868 kg.d-1
(317 tonnes.yr-1) for the low-recharge scenario and 1,603 kg.d-1 (585 tonnes.yr-1) for
the high-recharge scenario. A break down of nitrogen loads along the nominated
sections of the coastline, as above, is presented in Table 25. Approximately 75-80%
of the total nitrogen load was estimated to discharge from the area of contaminated
aquifer located south of James Point; however, the absolute values of nutrient loads
were approximately double those calculated based on nutrient concentrations in
porewater.
Insignificant discharges of nutrients occurred north of James Point North and in the
area adjacent to the Naval Base caravan Park because elevated nitrogen levels in
these areas were not detected in the groundwater bore that were sampled.
Table 25: Nitrogen loads for nominated sections of coastline
Nitrogen Load
Scenario Northern Harbour
[tonnes.yr-1]
Naval Base [tonnes.yr-1]
James Point North
[tonnes.yr-1]
James Point South
[tonnes.yr-1]
Total [tonnes.yr-1]
Low-recharge 76 (24%) 1 (<1%) 6 (1%) 234 (74%) 314 (99%) High-recharge 84 (14%) 2 (<1%) 6 (1%) 486 (83%) 578 (99%)
1. Northern Harbour: 1.7 km length (6443650-6441950 mN AMG84) 2. Naval Base: 2.0 km length (6439950-6437950 mN AMG84) 3. James Point North: 1.0 km length (6435450-6434450 mN AMG84) 4. James Point South: 3.0 km length (6433950-6430950 mN AMG84)
Between 76 and 84 tonnes.yr-1 (14 to 24%) of the total nitrogen load was discharged
into Northern Harbour. Of this, 67 tonnes.yr-1 came from the adopted estimate of
nutrient discharge from Weston Bioproducts’ wastewater injection plume (Mackie-
Martin PPK, 1993).
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 137 and Nutrient Discharge into Cockburn Sound
Figure 62: Estimated nitrogen load into Cockburn Sound based on model simulated SGD (low-recharge scenario) and groundwater nutrient concentrations measured by Appleyard (1994)
CSIRO Land and Water
138 Coast and Clean Seas Project No. WA9911
Figure 63: Estimated nitrogen load into Cockburn Sound based on model simulated SGD (high-recharge scenario) and groundwater nutrient concentrations measured by Appleyard (1994)
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 139 and Nutrient Discharge into Cockburn Sound
Figure 64: Estimated phosphorus load into Cockburn Sound based on model simulated SGD (low-recharge scenario) and groundwater nutrient concentrations measured by Appleyard (1994)
CSIRO Land and Water
140 Coast and Clean Seas Project No. WA9911
Figure 65: Estimated phosphorus load into Cockburn Sound based on model simulated SGD (high-recharge scenario) and groundwater nutrient concentrations measured by Appleyard (1994)
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 141 and Nutrient Discharge into Cockburn Sound
5.7 SYNTHESIS OF RESULTS
Table 26, and Figure 66 and Figure 67 present summaries of estimated total nitrogen
discharge into Cockburn Sound via SGD. Previous estimates are labeled according
to date (i.e. the year) of the estimate, and estimates from the present study and
Appleyard (1994) are labeled “a” to “h” according to which combination of SGD and
groundwater nitrogen concentrations was used to evaluate the mass flux.
To facilitate the comparison of results, the mean rate of groundwater discharge,
SGDmean, for each estimate was calculated as
L
QSGDmean = (7)
where Q is the total groundwater discharge [L3.T-1] along a stretch of coastline with
length L. The mean nitrogen concentration in groundwater, Nmean, was then
calculated according to
Q
NN Q
mean = (8)
where NQ is the corresponding estimate of total nitrogen discharge in groundwater
[M.T-1] along the same stretch of coastline. These calculations were performed to
highlight differences between estimates that were not obvious if only the final values
of nutrient loads were considered.
Estimates labeled “a” to “d” are those calculated in Sections 5.5 and 5.6 above. They
are based on the SGD values derived from the inshore water-balance model
(Section 4.4) and groundwater nitrogen concentrations determined from the
submarine porewater survey (Section 3.4). Estimates “e” and “f” are from Appleyard
(1994), and “g” and “h” are worst-case and best-case estimates, which provide limits
to the analysis.
Figure 67 is a “bubble plot” that depicts the magnitudes of total nitrogen loads
according to the mean values of SGD and groundwater nitrogen concentration used
to calculate them. It illustrates the range of uncertainty in both quantities and
emphasises the fact that similar estimates of nutrient load have been derived from
dissimilar estimates of SGD and nutrient concentrations. Compare estimates “b” and
CSIRO Land and Water
142 Coast and Clean Seas Project No. WA9911
“c” for example. In estimate “b”, SGDmean ≈ 6 m3.d-1.m-1, Nmean ≈ 11 mg.l-1 and NQ =
305 tonnes.yr-1. In estimate “c”, mean discharge is approximately half, SGDmean ≈
3 m3.d-1.m-1, but mean concentration is approximately doubled, Nmean ≈ 23 mg.l-1,
such that NQ = 317 tonnes.yr-1. Thus, the estimates are approximately equivalent but
for different reasons.
Overall, the considered estimates of SGDmean and Nmean both vary by a factor of
approximately 3. This results in a potential factor of 9 (i.e single order of
magnitude) variation in NQ, as represented by estimates “g” and “h”. This type of
simple sensitivity analysis has not been considered in previous studies. In some
instances, the groundwater flow rates and concentrations that were used to estimate
nutrient mass flux were not reported.
180
454
350 343
219
144
305 317
585
230
430
965
108
0
200
400
600
800
1000
1200
1978 1990 1992 1996 2000 a b c d e f g h Estimate
N Q [tonnes.yr -1 ]
Previous Estimates This Study and Appleyard (1994)
Figure 66: Estimates of total nitrogen discharge into Cockburn Sound via groundwater
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 143 and Nutrient Discharge into Cockburn Sound
Table 26: Estimates of total nitrogen discharge into Cockburn Sound via groundwater, and equivalent average values of SGD and nitrogen concentrations in groundwater
Label NQ
[tonnes.yr-1] L [km]
SGDmean [m3.d-1.m-1]
Nmean [mg.l-1]
SGD Estimate Nitrogen Concentration
Estimate
a 144 13.7 2.8 10.3 This study – low recharge This study - porewater b 305 13.7 5.5 11.1 This study – high recharge This study - porewater c 317 13.7 2.8 22.6 This study – low recharge Appleyard (1994) d 585 13.7 5.5 21.3 This study – high recharge Appleyard (1994) e 230 16.0 2.1 18.8 Appleyard (1994) Appleyard (1994) f 430 16.0 2.1 35.1 Appleyard (1994) Appleyard (1994) g 965 13.7 5.5 35.1 This study – high recharge Appleyard (1994) h 108 13.7 2.1 10.3 Appleyard (1994) This study - porewater
1978 180 - - - D.A. Lord (2000) D.A. Lord (2000) 1990 454 - - - D.A. Lord (2000) D.A. Lord (2000) 1992 350 - - - Mackie Martin (1992) Mackie Martin (1992) 1996 343 - - - DEP (1996) DEP (1996) 2000 219 - - - D.A. Lord (2000) D.A. Lord (2000)
Min. 2.1 10.3 Max. 5.5 35.1
1. Yellow-highlighted values were used to calculate blue-highlighted values
144 305
317 585
230
430
108
965
0
5
10
15
20
25
30
35
40
45
1 2 3 4 5 6 7 SGD mean [m 3 .d -1 .m -1 ]
N mean [mg.l -1 ]
a b c d e f g h
Total Nitrogen Discharge
[tonnes.yr -1 ]
Figure 67: Estimates of total nitrogen discharge into Cockburn Sound via groundwater with data plotted according to equivalent average values of SGDs and nitrogen concentrations in groundwater
CSIRO Land and Water
144 Coast and Clean Seas Project No. WA9911
5.7.1 UNCERTAINTY
Uncertainty in the above estimates of nutrient mass flux stem from the uncertainties
in estimating SGD and determining correct spatial distributions of nutrient
concentrations in the aquifer.
To estimate SGD, relatively accurate determinations of hydraulic gradient can be
made; however, the hydraulic conductivity of sediments is uncertain and varies
typically over several orders of magnitude. This introduces a same order of
magnitude uncertainty in SGD and nutrient flux estimates. By taking into
consideration the inshore groundwater balance, and ensuring that SGD rates are
consistent with realistic estimates of recharge and abstraction from the aquifer, the
level of uncertainty can be minimised. This approach was used in the present study;
however, the SGD rates that were calculated based on low and high estimates of
groundwater recharge still varied by a factor of at least 2.
Additional uncertainty was introduced during the process of deciding which value(s)
of nutrient concentration were most representative of nutrient distributions in
groundwater. For example, results from Northern Harbour highlight a potential
inadequacy of the submarine porewater survey. Because porewater samples were
collected within 10 m of the shoreline, the groundwater component of these samples
was most likely dominated by shallow groundwater from the upper part of the
aquifer (see conceptual models of SGD in Section 4.1). Nutrient concentrations in
shallow groundwater may or may not be representative of nutrient concentrations in
deeper groundwater discharging further from the shore. The nutrient plume from
Woodman Point Wastewater Treatment Plant, which is detected in groundwater
monitoring bores, was not detected in the porewater samples collected along the
shoreline of Northern Harbour.
In general, unless a porewater survey is extended offshore and over the full width of
SGD—assuming this is known—then there is potential that a single porewater
sample collected immediately along the shoreline could misrepresent the average
porewater concentration at that location. To resolve this uncertainty would require a
precise knowledge of where groundwater discharges so that porewater samples could
be collected at these locations. Nevertheless, patterns of groundwater discharge are
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 145 and Nutrient Discharge into Cockburn Sound
complex and strongly influenced by the coastal hydrogeology, which is poorly
defined at the scale at which porewater measurements are made.
Despite these limitations, the porewater survey had several advantages over previous
attempts to estimate nutrient mass flux based on concentrations measured in inshore
groundwater bores. Porewater samples were collected within 0.5 m of the seabed
and it is reasonable to assume that measured nutrient concentrations are an accurate
reflection of the concentrations in groundwater discharging at the measurement
locations. In addition, porewater samples were collected at more locations along the
coastline—compared to previous sampling of groundwater bores—and the density of
sampling was increased at locations where significant nutrient contamination was
detected; for example, north and south of James Point where porewater samples were
collected at 20-m intervals. Where possible, porewater samples were also collected
in offshore transects.
The main difficulty in estimating nutrient mass flux from nutrient concentrations
measured in groundwater bores is the interpolation within the aquifer of this data.
The spatial resolution of measured concentrations is typically poor. Deciding how
representative of the surrounding aquifer these measurements are, and how the data
should be interpolated spatially, is non-trivial and largely subjective. This is
compounded by the fact that it is rarely obvious which value of SGD should be
coupled with the measured or interpolated nutrient concentration at a particular
location in the aquifer.
For example, the available hydrogeological data indicates that most groundwater in
the Safety Bay Sand drains vertically into high-permeability Tamala Limestone and
then flows toward the coast via preferential flow paths. In this situation, it is unclear
whether nutrient concentrations measured in shallow groundwater bores that are
screened within the Safety Bay Sand should be coupled with flow rates in the sand or
limestone. This issue has not been addresses in previous investigations, which
simply assume that all groundwater in the Safety Bay Sand flows laterally through
the sand and discharges at the coast (i.e. vertical draining of groundwater into the
underlying Limestone is ignored). Because the contrast in hydraulic conductivity
between the limestone and overlying sands is so large, absence of vertical drainage
into the limestone is highly unlikely except where the sand and limestone are
separated by clay aquitard that prevents vertical flow.
CSIRO Land and Water
146 Coast and Clean Seas Project No. WA9911
To improve our understanding of groundwater nutrient transport into Cockburn
Sound, annual SGD nutrient loads computed in this study and previous studies
should be compared with estimates of mass loadings at the source locations (e.g.
based on loss of product or leaching rates). In each case, the following question
should be considered: Is the estimated nutrient mass loading at the source location
consistent with associated estimates of nutrient mass discharge to Cockburn Sound
over time? For example, if the estimate of nutrient mass discharge for a particular
nutrient plume, over a twenty-year period, was found to be significantly greater than
the estimate of nutrient mass loading at the source, then this would indicate that the
SGD rates used to calculate the mass discharge loadings were possibly over
estimated. Alternatively, the groundwater nutrient concentrations may have been
underestimated. This highlights one advantage of the porewater survey, which
detects nutrient concentrations directly below the seabed. Porewater sampling
affords a reasonably high level of confidence in estimates of nutrient concentrations
at the location of SGD; therefore, any inconsistencies between nutrient mass loadings
at source and discharge locations are more likely to be the results of errors in the
estimated SGD rates.
Spatial variability in groundwater recharge and flow, control the transport pathways
and travel times for nutrients to move from source areas inshore from Cockburn
Sound to the ocean. Nutrient contamination along the foreshore strip of Cockburn
Sound is mobilized into the groundwater system by local groundwater recharge. The
aquifer beneath these areas is then ‘flushed’ by groundwater flow from further
inshore as it moves toward Cockburn Sound. The presence of preferred flow paths
mean that some areas of the aquifer are likely to contain less active—slower
flowing—areas of groundwater compared with other parts of aquifer containing
preferred flow paths. Nutrients that are mobilized into relatively inactive areas of the
flow system are likely to take longer to reach Cockburn Sound. In addition, these
areas of aquifer are likely to contain nutrients for longer periods following
contamination events since the flushing time will be longer. The reverse case applies
in areas where the groundwater flow system is relatively active.
5.7.2 COMPOSITE ESTIMATE
A composite estimate of nitrogen discharge into Cockburn Sound, based on the
following assumptions and logic, is presented in Table 27:
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 147 and Nutrient Discharge into Cockburn Sound
• Compared to SGD rates derived from the inshore water-balance model, the
groundwater discharge rates used by Appleyard (1994) now appear to be
relatively small. Therefore, the model-derived values of SGD were used to
estimate nutrient mass flux.
• It is likely that there have been significant changes in concentrations of nutrients
in groundwater since the study by Appleyard in 1994, and an updated set of
nutrient concentrations in groundwater bores along the Cockburn Sound coastline
is not available. Therefore, nutrient concentrations in groundwater that were
derived from the submarine porewater survey were used here. An advantage is
that previously undetected discharges of nutrient-enriched groundwater from
north of James Point and adjacent to the Naval Base caravan park are included.
A disadvantage is the absence of comparable data sets with which to evaluate the
uncertainty in these measurements.
• Groundwater nutrient concentrations derived from the porewater survey appear to
underestimate the true concentrations in groundwater inshore from Northern
Harbour. Therefore, nitrogen mass flux into Northern Harbour computed by
PPK (2000, Table 3.2, Scheme 4) is used here. The estimate by PPK allows for a
51% reduction in nitrogen loading to the harbour due to plume recovery
operations.
The resultant estimate of total nitrogen load to Cockburn Sound was 234
±88 tonnes.yr-1 along the length of coastline considered in this study.
Approximately 71% was contributed from south of James Point, 12% from Northern
Harbour, 11% from north of James Point and 6% from the shoreline adjacent to the
Naval Base caravan Park.
Table 27: Composite estimate of total nitrogen discharge into Cockburn Sound via groundwater
Northern Harbour1
[tonnes.yr-1]
Naval Base2 [tonnes.yr-1]
James Point North2
[tonnes.yr-1]
James Point South2
[tonnes.yr-1]
Total [tonnes.yr-1]
29±14 (12%)
15±5 (6%)
25±10 (11%)
165±59 (71%)
234±88
1 PPK (2000, Table 3.2) 2 This study, estimates “a” and “b”
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 149 and Nutrient Discharge into Cockburn Sound
REFERENCES
Appleyard, S. J., 1994, The flux of nitrogen and phosphorus from groundwater into Cockburn Sound, Perth Metropolitan Region, Geological Survey of Western Australia (unpublished), Hydrogeology Report 1994/39.
Bodard, J. M., 1991a, Assessment of the BP refinery site, Kwinana, Western Australia: Volume I Hydrological framework and site work recommendations, BHP Research, Melbourne Laboratories.
Bodard, J. M., 1991b, Assessment of the BP refinery site, Kwinana, Western Australia: Volume II Hydrological framework and site work recommendations, Figures and attachments, BHP Research, Melbourne Laboratories.
Corporate Environmental Consultancy, 2002, Hismelt (Operations) Pty Limited Commercial Hismelt Plant Kwinana Western Australia Public Environmental Review, April 2002.
D.A. Lord, 2001, Cockburn Sound Management Council, The state of Cockburn Sound: A pressure-state-response report, prepared by D.A. Lord & Associates Pty Ltd in association with PPK Environment and Infrastructure Pty Ltd, Report No. 01/187/1, June 2001.
Davidson, W. A., 1995, Hydrogeology and groundwater resources of the Perth Region Western Australia, Geological Survey of Western Australia, Bulletin 142.
Davis, G. B., Johnston, C. D. and Patterson, B. M., 1994, Deep drilling data, multiport installation specifications, groundwater chemistry and soil gas composition for the first bioremediation field trial at the Kwiana refinery site, Report to BP (Kwinana), Report No. 94/34, September 1994.
DEP, 1996, Southern Metropolitan coastal waters study (1991-1994) final report, Department of Environmental Protection Western Australia, Report 17, November 1996.
Department of Land Administration, 5m and 1m Topographic data (supplied by WRC).
Department of Minerals and Energy, Perth Metropolitan Region 1:50,000 Environmental Geology Series Fremantle, Part Sheets 2033 I and 2033 IV.
Department of Minerals and Energy, Perth Metropolitan Region 1:50,000 Environmental Geology Series Rockingham, Part Sheets 2033 II and 2033 III.
Department of Minerals and Energy, 1:50000 Environmental Geology digital data, (supplied by Lorraine)
CSIRO Land and Water
150 Coast and Clean Seas Project No. WA9911
Gerbaz, S., 1999, Investigation of saltwater intrusion at BP Refinery (Kwinana). A thesis presented for Degree of Bachelor of Engineering, The University of Western Austalia, October 1999.
Glover, R. E., 1959, The pattern of fresh-water flow in a coastal aquifer. Journal of Geophysical Resaerch 10: 1039-1043.
HGM (Halpern Glick Mansel), 1998, Investigation of water quality in the Jervoise Bay Northern Harbour; December 1997 – March 1998, report to Department of Commerce and Trade.
Hirschberg, K-J., 1990, Lake Coogee Area assessment of monitoring data 1981-1987, Geological Survey of Western Australia, Hydrogeology Report 1989/38 (unpublished?).
Krupa, S., Gefvert, C., Brock, J., Grall, C. and Krupa, A., 2000, Field report for the International Oceanographic Committee, Submarine groundwater intercomparison experiment, Cockburn Sound, Western Australia, November 25 - December 6, 2000 (unpublished).
Lee, D. R., 1977, A device for measuring seepage flux into lakes and estuaries, Limnology and Oceanography, 22, 140-147.
Nield Consulting, 1999, Modelling of the Superficial Aquifer in the Cockburn Groundwater Area (Draft), unpublished report to Water and Rivers Commission and Kwinana Industries Council.
Mackie-Martin PPK, 1993, NB Love Starches Consultative Environmental Review: Proposal to increase quantity of wastewater disposed of by shallow bore injection.
Passmore, J. R., 1970, Shallow coastal aquifers in the Rockingham district Western Australia, Water Research Foundation of Australia, Bulletin No. 18.
Patterson, B.M.; Grassi, M.E.; Davis, G.B.; Robertson, B. and McKinley, A.J., 2002, The use of polymer mats in series for sequential reactive barrier remediation of ammonium-contaminated groundwater: laboratory column evaluation, Environmental Science and Technology, 36, 3439-3445.
Patterson, B.; Smith, A.; Davis, G.; Robertson, B.; Hick, W.; Grassi M., Herne, D. and Turner J., 2002, Quantifying submarine groundwater discharge and demonstrating innovative clean-up to protect Cockburn Sound from nutrient discharge (Project No. WA9911), final report to Coast and Clean Seas National Heritage Trust, Centre for Groundwater Studies and CSIRO Land and Water, November 2002.
Paulsen, R.J., C.F. Smith, and T.-f. Wong, 1997, Development and evaluation of an ultrasonic groundwater seepage meter, in Geology of Long Island and Metropolitan New York, 88-97.
Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 151 and Nutrient Discharge into Cockburn Sound
PPK Environment and Infrastructure, 1999, Remediation strategy for nitrogen-rich groundwater at Jervoise Bay, report prepared for Department of Commerce and Trade, 98K148A:PR2:4454:RevB, April 1999.
PPK Environment and Infrastructure, 2000, Jervoise Bay Project groundwater recovery plan, report prepared for Department of Commerce and Trade, 98L084A:PR2:5426, April 1999.
Rust PPK Environment and Infrastructure, 1996. Kwinana Works Groundwater Monitoring Manual, report prepared for Westfarmers CSBP Ltd., 98H065A:PR2:0639:RevA, May 1996.
Searle, D. J., Semeniuk, V. and Woods, P. J., 1988, Geomorphology, stratigraphy and Holocene history of the Rockingham-Becher Plain, South-western Australia, Journal of the Royal Society of Western Australia, 70(4), 89-109.
Smith, A.J. and Hick, W.P., 2001, Hydrogeology and aquifer tidal propagation in Cockburn Sound, Western Australia, CSIRO Land and Water Technical Report 06/01, February 2001.
Stieglitz, T., 2001a, Seismic Survey in Northern Harbour & Cockburn Sound December 2000, unpublished report to SCOR/LOICZ Working Group 112.
Stieglitz, T., 2001b, Ground conductivity transects in Northern Harbour, Cockburn Sound, December 2000, unpublished report to SCOR/LOICZ Working Group 112.
Stieglitz, T., Ridd, P. V. and Hollins, S., 2000, A Small Sensor for Detecting Animal Burrows and Monitoring Burrow Water Conductivity. Wetlands Ecology and Management, 8, 1-7.
Walker, N., 1994, Tidal influences on coastal groundwater levels. A thesis for the Degree of Bachelor of Engineering, The University of Western Australia, October 1994.
WAWA, 1993, Cockburn Groundwater Area Management Plan, June 1993, produced by Groundwater and Environment Branch, Water Authority of Western Australia, ISBN-0-7309-5260-6.
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 153 and Nutrient Discharge into Cockburn Sound
APPENDIX A: AQUIFER TIDAL PROPAGATION AT NORTHERN
HARBOUR
CSIRO Land and Water APPENDIX A
154 Coast and Clean Seas Project No. WA9911
THE TIDAL METHOD
The tidal method (Ferris, 1951; Carr and Van Der Kamp, 1969; Townley, 1995) is a
simple technique for estimating aquifer diffusivity—the ratio of aquifer
transmissivity to the aquifer storage coefficient—based upon the response of the
aquifer to tidal forcing at a boundary. Similar to conventional pump test techniques,
in which the piezometric head response of an aquifer to artificial pumping is used as
a basis for estimating aquifer hydraulic properties, the tidal method takes advantage
of natural tidal forcing and obtains an estimate of diffusivity from the attenuation of
a tidal signal as it propagates into an aquifer. The tidal signal can be expressed as a
linear combination of sinusoidal terms or tidal constituents (Forman and Henry,
1989, p.109) that are differentially attenuated as they travel though the aquifer; the
degree and rate of attenuation depends upon the aquifer hydraulic properties. There
are two possible mechanism for tidal propagation: (1) flow of tidal water into and out
of an unconfined aquifer, (2) compression and expansion of a confined aquifer due to
the load of the incoming and outgoing tides. In an aquifer that has tidal forcing at a
lateral boundary, the attenuation of the tidal signal is normally described by the two
quantities tidal efficiency and lag.
Tidal efficiency, TE, is a normalised measure that relates the amplitude of head
fluctuations in the aquifer to the amplitude of fluctuations at the tidal boundary. In
general,
ak
akk F
RTE = (9)
where Fa is the amplitude of forcing at the tidal boundary, Ra is the amplitude of the
response at a point in the aquifer and k denotes the tidal constituent or frequency.
The tidal efficiency has a value between zero and one.
Lag is an inverse measure of the velocity of propagation of a tidal constituent as it
moves through the aquifer
kkk FRlag θθ −= (10)
where Fθ is the phase of tidal forcing and Rθ is the phase of the tidal response at a
point in the aquifer. Note that the slower the speed of propagation, the larger the lag.
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 155 and Nutrient Discharge into Cockburn Sound
Ferris (1951) presents analytical expressions that describe the tidal efficiency and lag
in a one-dimensional, semi-infinite and homogeneous aquifer with uniform
transmissivity (see also Carr and Van Der Kamp, 1969). Inverse solutions are
obtained for 1D-transient groundwater flow with a tidal or harmonic boundary
condition at x = 0.
t
h
T
S
x
h
∂∂
∂∂ =
2
2
(11)
( )θω HtHth a −= cos),0( (12)
Here, Ha is the amplitude of head fluctuations at the tidal boundary [L] and Hθ is the
phase measured in radians. Since (11) and (12) are both linear, the groundwater flow
problem they describe can be decomposed into a steady-state flow problem and a
harmonic flow problem that consists of one or more frequencies. The harmonic
solution is rearranged to yield the following expressions for tidal efficiency and lag
⎟⎟⎠
⎞⎜⎜⎝
⎛−=
kk TP
SxTE
exp
π (13)
T
SPxlag k
k 4π= (14)
The coordinate x represents distance from the tidal boundary, S is the aquifer storage
coefficient [1], T is aquifer transmissivity [L2T-1]—which is assumed uniform, Pk is
the period [T] of the tidal constituent k and ωk = 2π/Pk is the corresponding angular
frequency [T-1]. Since a particular tidal signal may contain several dominant
frequencies, e.g. semi-diurnal and diurnal constituents, several independent
estimations of diffusivity are possible from a single tidal signal. In general, longer-
period constituents will propagate further into the aquifer and yield estimates of
diffusivity that apply to larger regions of the aquifer.
From (13) and (14), efficiency-based and lag-based expressions for aquifer
diffusivity are straightforward to derive
( ) kk PTE
x
S
T2
2
ln
π= (15)
CSIRO Land and Water APPENDIX A
156 Coast and Clean Seas Project No. WA9911
( )2
2
4 k
k
lag
Px
S
T
π= (16)
Equating (15) and (16) gives the non-dimensional relationship
kk
k TEP
lagln
2
1⎟⎠⎞
⎜⎝⎛−=
π (17)
where the left-hand term is a normalised lag. It follows that, on a semi-log plot, the
normalised lag and tidal efficiency are related by the straight line with slope equal to
–1/2π. The tidal efficiency decreases exponentially with increasing distance from the
tidal boundary, whereas the lag increases linearly.
FOURIER ANALYSIS
The term Fourier analysis refers to any data analysis technique that describes or
measures fluctuations in a time series by comparing them with sinusoids. Well
known methods include filtering, least squares regression on sinusoids and harmonic
analysis (Bloomfield, 1976; Chatfield, 1975; James, 1995). Fourier techniques are
commonly used to detect frequency components that are hidden within a 'noisy'
signal.
Harmonic analyses fits uniformly sampled time series data to a set of harmonic
frequencies; that is, a set of frequencies that are integer multiples of one another.
The fitted harmonics are normally in the frequency range Nνν ≤≤0 , where
tPN
N δν
2
11 == (18)
is the Nyquist frequency [T-1], PN is the Nyquist period [T] and δt is the data
sampling interval [T].
The Nyquist condition simply states that the smallest period that can be reasonably
fitted to the data is equal to twice the sampling interval. This corresponds to fitting,
at most, one local maxima or minima between each pair of data point in the time
series. From a practical point of view, the Nyquist conditions tells us that it is
unreasonable to fit, say, a semi-diurnal frequency (12 hour period) to data that has
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 157 and Nutrient Discharge into Cockburn Sound
only been sampled 24 hourly. The smallest period and highest frequency that should
be fitted to diurnal data are 48 hours and 0.5 cycles per day, respectively.
DISCRETE FAST FOURIER TRANSFORM (FFT)
The discrete FFT is a computationally efficient algorithm that enables regularly
sampled time series data to be fitted by a set of harmonic frequencies (Chatfield,
1975; Bloomfield, 1976; James, 1995). For a time series of n data values, the FFT
fits n/2 frequencies over the frequency range NN vvnv ≤≤/2 . The FFT of the time
series is the set of n/2 complex valued Fourier coefficients Φ(vj), where the moduli
|Φ(vj)| are equal to the amplitudes of the fitted frequencies and arguments the
arg(Φ(vj)) are equal to the phases.
The Inverse Fast Fourier Transform (IFFT) is a reverse operation that converts a set
of Fourier coefficients Φ(vj) back to the original time-series data y(ti). This
relationship is denoted symbolically as (James, 1995)
)()( ji vty Φ⇔ (19)
A measure of the energy or power at each FFT frequency is provided by the Power
Spectral Density (PSD) function
)( )()( *jjj vvvS ΦΦ= (20)
Here, Φ*(vj) denotes the complex conjugates of the Fourier coefficients. A large
value of S(vj) indicates more energy at frequency vj. This means that the dominant
frequencies in a time series are represented as larger spikes on a PSD plot.
FREMANTLE TIDE DATA
The tide station at Fremantle—maintained by the Department of Transport, Western
Australian—is the nearest tidal monitoring location to Cockburn Sound. Historic
tide data is available upon request to the Maritime Division at a minimum
continuous sampling interval of five minutes. Fremantle tide data (Figure 68) used
in this report are sampled at 10-minute intervals over the period 01/09/2000 to
16/01/2001.
CSIRO Land and Water APPENDIX A
158 Coast and Clean Seas Project No. WA9911
Fremantle has a mixed tide that is characterised by diurnal (24-hour) and semi-
diurnal (12-hour) constituents. The K1† (period = 23.9345 hours) and O1‡ (period =
25.8193) diurnal constituents are the dominant frequencies (Department of
Transport, unpublished harmonic analysis). The semi-diurnal frequencies vanish for
part of each month, which makes them unsuited for use with the tidal method.
FREMANTLE TIDE DATA
Tide data was logged continuously at 10-minute intervals for the duration of the
SGD intercomparison experiment. Water levels were measured using a capacitive
water level probe (Dataflow Systems) and model 392 data recorder (Dataflow
Systems) suspended within a 50-milimetre PVC stilling well and mounted off a sea
wall at the intercomparison site. The same equipment was used to record water
levels—also at 10-minute intervals—in each of the five monitoring bores. Logger
internal clocks were synchronized using the same laptop computer so that the tide
and water levels in all bores were recorded at the same ten-minute intervals.
The tide data collected during the intercomparison is presented in Figure A2.
Because the tidal logger failed on 12 December and the length of tidal record
obtained to that point in time was too short, the K1 and O1 tidal constituents could
not be separated in a harmonic analysis. Department of Transport tide data for
Fremantle (Figure 69) was used to ‘repair’ the Cockburn Sound data set and extend
the length of the time series to match that obtained for the groundwater monitoring
bores.
Harmonic Analysis
All data sets presented in this report are standardised to a ten-minute sampling
interval. The corresponding 72 cycles per day Nyquist frequency is more than
adequate to detect the expected diurnal or 1 cycle per day tidal constituents. The
number of harmonics fitted in each analysis depends upon the number of data values
in the time series being considered. Since the discrete FFT fits n/2 harmonics to n
data points, a larger number of data points result in more frequencies and a higher
resolution PSD plot. When a standardised sampling interval is used, the total length
of the time series determines the number of data points and Fourier frequencies. All
† K1 is the lunisolar diurnal constituent; with O1 it expresses the effect of the Moon's declination ‡ O1 is the lunar diurnal constituent
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 159 and Nutrient Discharge into Cockburn Sound
data sets used in this report contained 4628 data points, corresponding to the period
from 25/11/2000 @ 17:10 to 27/12/2000 @ 20:20. This means that 2314
frequencies are fitted in the harmonic analyses presented.
Foreman and Henry (1989, Table A1) specify that the minimum total length of time
series required for separation of the K1 and O1 tidal constituents in a harmonic
analysis are 24 hours and 328 hours (13.67 days), respectively. The data sets used in
this report are approximately 32 days long, which satisfies the above crtieria.
Harmonic analyses are performed using Mathematica 3.0 (Wolfram Research, 1988-
1996) by first detrending the time series—that is, subtracting the mean data value
from each data value—and then implementing the intrinsic FFT function
“Fourier[list]”. The PSD function is then calculated as nvvvS jjj /)( )()( *ΦΦ= .
Figure 70 to Figure 75 depict water levels and the PSD functions for six of the seven
intercomparison monitoring bores; NH1a, NH2a, NH2b, NH3, NH4 and NH5. A
continuous record was not obtained in NH1b due to problems with the data logger.
The ‘spectral signatures' of these data sets are consistent with the tide data and the
K1 and O1 tidal constituents are clearly discernable. Peaks on the PSD plots at
lower frequencies ( 25.0≤jv ) correspond to the passage of weather systems,
associated fluctuations in barometric pressure and 'piling up' of water at the shoreline
due to prevailing onshore winds. These weather effects are non-stationary, however,
and are complicated by the fact that barometric pressure fluctuations affect the
aquifer storage directly through atmospheric loading, as well as indirectly through
tidal loading. They are therefore not suited for use with the tidal method.
CSIRO Land and Water APPENDIX A
160 Coast and Clean Seas Project No. WA9911
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) Fremantle Tide Data
20 40 60 80 100 120
Frequency (cycles/day)
Va
lue
(b) Fremantle Power Spectral Density
0.5 1 1.5 2 2.5 3
0
20
40
60
80
100
120
O1
K1
Figure 68: Tidal data and Power Spectral Density for Fremantle for the period 1/9/2000 to 16/1/2001; sampling interval is 10 minutes
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 161 and Nutrient Discharge into Cockburn Sound
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) Cockburn Sound Tide Data
5 10 15 20 25 30 35
Loggerfailure
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(b) Cockburn Sound ‘Repaired’ Tide Data
5 10 15 20 25 30 35
Figure 69: Cockburn Sound tidal data and ‘repaired’ Cockburn Sound tidal data for the period 25/11/2000 to 3/1/2001; sample interval is 10 minutes
CSIRO Land and Water APPENDIX A
162 Coast and Clean Seas Project No. WA9911
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH1a Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH1a Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0.5
1.0
1.5
2.0
2.5
3.0
0.0
Figure 70: Hydrograph and Power Spectral Density for groundwater monitoring bore NH1a for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 163 and Nutrient Discharge into Cockburn Sound
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH2a Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH2a Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0.0
0.5
1.0
1.5
2.0
2.5
Figure 71: Hydrograph and Power Spectral Density for groundwater monitoring bore NH2a for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
CSIRO Land and Water APPENDIX A
164 Coast and Clean Seas Project No. WA9911
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH2b Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH2b Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0.0
0.5
1.0
1.5
2.0
2.5
Figure 72: Hydrograph and Power Spectral Density for groundwater monitoring bore NH2b for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 165 and Nutrient Discharge into Cockburn Sound
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH3 Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH3 Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0
1
2
3
4
5
6
Figure 73: Hydrograph and Power Spectral Density for groundwater monitoring bore NH3 for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
CSIRO Land and Water APPENDIX A
166 Coast and Clean Seas Project No. WA9911
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH4 Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH4 Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0
2
4
6
8
10
12
Figure 74: Hydrograph and Power Spectral Density for groundwater monitoring bore NH4 for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 167 and Nutrient Discharge into Cockburn Sound
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
Time (days)
Ele
vation
(mA
HD
)
(a) NH5 Tide Data
0 5 10 15 20 25 30
Frequency (cycles/day)
Va
lue
(b) NH5 Power Spectral Density
0.5 1 1.5 2 2.5 3
O1
K1
0
5
10
15
20
Figure 75: Hydrograph and Power Spectral Density for groundwater monitoring bore NH5 for the period 25/11/2000 to 27/12/2000; sampling interval is 10 minutes
CSIRO Land and Water APPENDIX A
168 Coast and Clean Seas Project No. WA9911
AQUIFER TIDAL PROPAGATION
At each monitoring bore, applying the tidal method to the K1 and O1 tidal
constituents gives four estimates of aquifer diffusivity; each constituent yields an
efficiency-based and lag-based estimate of diffusivity. Results for the
Intercomparison monitoring bores are summarised in Tables A2 and A3.
Tidal efficiency and lag are calculated from FFT’s of the ocean tide data and
groundwater level responses in the monitoring bores. At each bore
ocean1
bore11 )(
)(
K
K
KTEνν
ΦΦ
= ocean1
bore11 )(
)(
O
O
OTEνν
ΦΦ
= (21)
and
⎟⎟⎠
⎞⎜⎜⎝
⎛ΦΦ
=ocean1
boren11 )(
)(arg
K
KKlag
νν
⎟⎟⎠
⎞⎜⎜⎝
⎛ΦΦ
=ocean1
bore11 )(
)(arg
O
OOlag
νν
(22)
Once tidal efficiency and lag are determined, diffusivity is calculated using equations
(15) and (16).
The validity of applying the ‘Ferris model’ to the Cockburn Sound data set is tested
by plotting normalised lag verses tidal efficiency on a semi-log graph (Figure 76).
The data are compared with each other and the theoretical relationship from equation
(17). Most notably, all the bores diverge from the theoretical relationship between
tidal efficiency and normalised lag that is predicted by the Ferris model, and there is
a clear separation of the deep and shallow bores.
Shallow Bores
There is evidence of a logarithmic relationship between tidal efficiency and
normalised lag in the shallow bores but the slope is clearly different to that of the
Ferris model (-1/2π). In general, the observed attenuation of tidal amplitude is
greater than the model-predicted attenuation, for the same speed of propagation
through the aquifer. Correspondingly, the observed speed of propagation is faster
than the model-predicted speed of propagation, for the same attenuation of
amplitude. This difference is most likely a consequence of the modelling
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 169 and Nutrient Discharge into Cockburn Sound
assumptions and results in estimates of lag-based diffusivity that are consistently
larger than efficiency-based estimates.
Despite this deficiency of the model, the estimates of diffusivity determined from the
shallow bores fall within the range 3,600 to 11,800 m2.d-1, which is within a factor of
3.3. Assuming that the saturated depth of the sand units overlying the Tamala
Limestone is around 17 metres (based on the drilling data) and assuming a specific
yield of approximately 0.25, the estimated range in hydraulic conductivity of the
sands is 53 to 174 m.d-1, which is characteristic of medium to coarse sand (Fetter,
1994; Bouwer, 1978).
Deep Bores
Figure 76 indicates that the relationship between tidal efficiency and normalised lag
in the deep bores is significantly different compared to both the shallow bores and
the Ferris model. Again, this is likely to be a consequence of the model assumptions,
which do not account for the high-transmissivity Tamala Limestone underlying the
sand. The assumptions of 1-D flow and a homogeneous aquifer, in particular, are
poor representations of the true field conditions.
During the SCOR/LOICZ SGD Intercomparison, it was observed that some of the
submarine discharge being measured by benthic flux meters on the seabed had
seawater salinity. This is consistent with the fact that the ocean level was falling
during the experiment and indicates that salt water, previously forced into the aquifer
during a period of higher sea level, was draining back out of the aquifer. The period
of the intercomparison corresponds to the first ten days in Figure 70 to Figure 75 in
which the trend of falling sea level is clearly depicted. It is believed that there is very
active exchange of seawater between the ocean and Tamala Limestone, which is
manifest as ‘tidal pumping’.
The current hypothesis is that during periods of rising sea level, saltwater flows into
the Tamala Limestone, which has the effect of displacing both the saltwater interface
and free surface in the overlying sands upward. Downward displacement of the free
surface and saline interface occurs during periods of falling sea level. In effect, rapid
propagation of tidal oscillations through the limestone extends the ocean boundary
inshore and underneath of the freshwater flow system. The fact that the saltwater
wedge extends more than two kilometres inland from the coast is an indication of
how transmissive the limestone is (Figure 16). In addition, groundwater levels
CSIRO Land and Water APPENDIX A
170 Coast and Clean Seas Project No. WA9911
several hundred metres inland from the coast are strongly correlated with monthly
average sea level (PPK, 2000).
Because the deep bores are screened just above the limestone, this affects their tidal
responses. Compared with the Ferris model, the attenuation of tidal amplitude in the
deep bores is much larger than expected for the same speed of propagation.
Therefore, the lag-based estimates of diffusivity, which are in the range 8,400 to
28,100 m2.d-1, are much larger than the efficiency-based estimates, which are in the
range 51,800 to 441,400 m2.d-1. For the reasons discussed, none of these values is
considered realistic and it is concluded that the Ferris model is not suited to the
particular field conditions.
Table 28: Summary of tidal analysis for K1 tidal constituent
Bore ID Distance to Shore [m]
TEK1 Efficiency-based
Diffusivity [m2.d-1]
LagK1 Lag-based Diffusivity [m2.d-1]
NH1a 179 0.152 27,300 0.094 296,900 NH2a 93 0.171 8,400 0.117 51,800 NH2b 96 0.076 4,200 0.293 8,800 NH3 45 0.301 4,200 0.146 7,800 NH4 28 0.451 3,600 0.090 7,700 NH5 18 0.611 4,000 0.064 6,500
Table 29: Summary of tidal analysis for O1 tidal constituent
Bore ID Distance to Shore [m]
TEO1 Efficiency-based
Diffusivity [m2.d-1]
LagO1
Lag-based Diffusivity
[m2.d-1]
NH1a 179 0.147 28,100 0.075 441,400 NH2a 93 0.174 9,100 0.044 347,600 NH2b 96 0.075 4,400 0.246 11,700 NH3 45 0.321 5,000 0.115 11,800 NH4 28 0.493 4,900 0.081 9,000 NH5 18 0.639 5,200 0.050 10,200
APPENDIX A Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 171 and Nutrient Discharge into Cockburn Sound
Figure 76: Non-dimensional plot of normalised lag verses tidal efficiency for groundwater monitoring bores NH1 to NH5
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.001 0.01 0.1 1
Tidal Efficiency
Nor
mal
ised
Lag
Ferris
K1
O1
ShallowBores
DeepBores
APPENDIX B Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 173 and Nutrient Discharge into Cockburn Sound
APPENDIX B: BENTHIC FLUX METERS
CSIRO Land and Water APPENDIX B
174 Coast and Clean Seas Project No. WA9911
Three types of benthic flux meter (BFM) are referred to in this study:
Lee-type
Ultrasonic
Heat-pulse
A brief description of their design and operation is presented below.
LEE-TYPE
Lee-type meters (Lee, 1977) consist of an open drum with a flexible plastic bag
connected by a short length of conduit. With the open end of the drum pushed
securely into the sediment and the meter fully submerged, the plastic bag is checked
periodically and the time-integrated benthic flux calculated from the change in
volume of water in the plastic bag. An increase of water volume indicates a mean
flux of porewater out of sediment into the water column during the measurement
period; a decrease indicates a mean flux of seawater into sediment. Benthic flux, in
units L.T-1, was calculated by diving the change in volume [L3.T-1] by the ‘footprint’
area of the collection drum [L2].
ULTRASONIC
The ultrasonic BFM used during the SGD intercomparison was developed in the
USA (Paulsen et al., 1997). Flow velocity is measured based on the difference in the
speed of propagation between ultrasonic waves traveling in the direction of flow and
against the direction of flow. In operation, a collection device is pressed into the
seabed and the flow amplified by passing it through a relatively small diameter tube
in which the flow velocity is measured using ultrasonics. A distinct advantage of
this instrument is its speed. Essentially instantaneous measurements of SGD can be
made because the time required for ultrasonic waves to travel the length of the
analyser tube is very short.
APPENDIX B Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 175 and Nutrient Discharge into Cockburn Sound
HEAT-PULSE
The heat-pulse BFM referred to in this study was developed by CSIRO Land and
Water. The operating principle is based on using heat as a tracer of flow (Figure 77).
Flow through the collection device on the seabed is amplified through a relatively
small diameter tube that contains a small heater and electronic temperature sensors
(thermistors). The heater is positioned mid-way in the thermistor array and is
controlled electronically. To make a measurement, the heater is turned on for 0.2
seconds and the direction and velocity of flow in the analyser tube estimated based
on the direction and speed of travel of the heated water, which is detected by the
thermistor array. Temperature at each thermistor is recorded continuously at
0.2-second intervals. The objective is to heat the water to a temperature that is
detectable by the thermistors but not to the point that natural convection perturbs the
flow.
Figure 78 depicts example heat-pulse data measured during instrument calibration.
Each curve corresponds to the temperature time series at the labeled thermistor
(Figure 77). Thermistors were equally spaced at 10 mm intervals in the instrument
used in this study. A temperature count of 50 was approximately equivalent to 1
degree Celsius. In Figure 78, the heater was ‘fired’ at 5 seconds and the heat pulse
moved down the analyser tube in the direction of thermistors T4, T5 and T6. In
operation, this would indicate a flux of porewater from sediment into the water
column. The apparent flow velocity in the analyser tube is calculated by dividing the
distance between two thermistors by the time taken for the heat pulse to travel
between them. Two estimates can be made using the inner pair of thermistors, T2-
T3 or T4-T5, and the outer pair, T1-T3 or T4-T6, dependent on the direction of flow.
The travel time between thermistors is calculated as either the time difference
between peak temperatures at the thermistors or the time difference between the
arrivals of the temperature front at thermistors. The latter method is preferred if
there are flow reversals in the tube since this causes multiple temperature peaks that
are difficult to distinguish.
The true flow rate is obtained from calibration curves that relate the apparent flow
velocity in the analyser tube to the true flow rate determined by electronic mass
balance. Calibration curves for the inner and outer pairs of thermistors are presented
in Figure 79 and Figure 80. The correspondence between the apparent and true flow
rates is not one-to-one because the transport of heat in the analyser tube is not purely
advective; heat is also transported by conduction, which is a diffusive process. In
CSIRO Land and Water APPENDIX B
176 Coast and Clean Seas Project No. WA9911
general, the calibration model is less precise and more uncertain for higher flow rates
since travel times between thermistors are smaller and this number is the divisor in
the calculation of apparent flow rate. The minimum flow rate that was practical to
detect in the laboratory was approximately 1 mm.s-1. Below this rate, the heat
transport was achieved mainly by conduction (i.e. the Péclet number was large). At
lower flow rates, the heat pulse traveled away from the heater at approximately the
same rate in both directions.
T1T2 T3 T4
T6T5
+ve
-ve
Thermistors
and flow
direction
Heater
Figure 77: Heat-pulse BFM
APPENDIX B Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 177 and Nutrient Discharge into Cockburn Sound
BFM Data - LaboratorySteady Flow 19/2/01, 12:50:00 PM
2140
2160
2180
2200
2220
2240
2260
2280
2300
0 10 20 30 40 50 60 70
Time [s]
Tem
p [
cou
nts
]
T1 T2 T3 T4 T5 T6
Figure 78: Example heat-pulse data from laboratory calibration; flow rate = 2.8 mm.s-1
Calibration on Inner Thermistor Pair
y = 0.965x + 2.1076
R2 = 0.99
0
1
2
3
4
5
6
7
8
9
10
11
12
0 1 2 3 4 5 6 7 8 9 10 11 12
Mass Balance Flux [mm/s]
BF
M F
lux
[mm
/s]
1st Thermistor Pair Linear regression
Figure 79: Calibration curve for the heat-pulse BFM using inner thermistor pairs T2-T3 and T4-T5
CSIRO Land and Water APPENDIX B
178 Coast and Clean Seas Project No. WA9911
Calibration on Outer Thermistor Pair
y = 1.2647x + 0.9252
R2 = 0.9961
0
1
2
3
4
5
6
7
8
9
10
11
12
0 1 2 3 4 5 6 7 8 9 10 11 12
Flux from Mass Balance [mm/s]
Flu
x fr
om
BF
M A
nal
yser
[m
m/s
]
2nd Thermistor Pair Linear regression
Figure 80: Calibration curve for the heat-pulse BFM using outer thermistor pairs T1-T3 and T4-T6
APPENDIX C Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 179 and Nutrient Discharge into Cockburn Sound
APPENDIX C: MODELS OF TIDE AND WAVE AFFECTS ON BENTHIC
FLUX
CSIRO Land and Water APPENDIX C
180 Coast and Clean Seas Project No. WA9911
EFFECT OF SEA LEVEL DYNAMICS
As groundwater in an unconfined aquifer moves toward and eventually discharges at
the coast, the direction of groundwater flow changes from predominantly horizontal
flow in the aquifer inshore from the coastline to predominantly vertical flow directly
beneath the seabed. This change from horizontal to vertical flow generally takes
place near to the shoreline or area of SGD.
Harmonic solutions of 1D, unsteady groundwater flow (e.g. Ferris, 1951 and
Townley, 1995) are commonly applied to simulate aquifer tidal responses that are
due to lateral propagation of ocean water-level fluctuations into an aquifer. Tidal
components are typically diurnal (1-day period) and can induce water level responses
and lateral oscillations in groundwater flow at significant distances inshore from the
beach if the aquifer is sufficiently permeable. In such cases, the aquifer is simulated
as a horizontal flow system and vertical flow components near to the beach are
ignored.
The same harmonic solutions also can be applied to simulate vertical oscillations in
benthic flux across the seabed. These are induced by tidal and longer-period sea
level fluctuations but also arise due to higher-frequency fluctuation in sea level such
as swell and waves. Coastal swell and waves typical have periods in the order of
seconds to ten-of-seconds. The affects of these high-frequency fluctuations in sea
level do not propagate far into the aquifer because the boundary condition is
perturbed only for a short period. Predominantly small-scale vertical oscillations are
induced in groundwater flow directly beneath the seabed. Although the volume of
water that is exchanged across the seabed is relatively small compared to tidal and
longer-period exchanges, the instantaneous fluxes caused by waves and swell can
still be significant because the period of exchange is small (i.e. flux = volume/time).
Two analytical solutions (Ferris, 1951; Townley, 1995) are suitable for quantitatively
exploring the affects of tides, waves and swell on benthic flux. Both solutions
assume a 1D homogeneous aquifer with a harmonic head (sea level) boundary
condition at the ocean. The Ferris solution is for a so-called semi-infinite aquifer
that is infinitely deep and extends an infinite distance inshore. The Townley solution
allows for a finite depth and length aquifer with impermeable base.
APPENDIX C Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 181 and Nutrient Discharge into Cockburn Sound
The purpose of the following analyses is to demonstrate that the theoretical
magnitudes of tide-induced and wave-induced benthic fluxes are significant in
comparison with the estimates rates of fresh terrestrial SGD into Cockburn Sound.
TIDE-FORCED MODEL
The tide-pumping model explores lateral propagation of tidal flows into and out of
the aquifer. It applies the Townley solution for 1D-transient groundwater flow in an
aquifer with a tidal or harmonic boundary condition at x = 0. For each tidal
constituent the governing flow equations are
t
h
T
S
x
h
∂∂=
∂∂
2
2
(23)
)cos(),0( tHth a ω= (24)
0),( =
∂∂−
x
tLhT (25)
where h is hydraulic head [L], S is the aquifer storage coefficient [1], T is
transmissivity [L2/T], L is aquifer length [L], x is distance inshore from the tidal
boundary, t is time, Ha is the amplitude [L] of sea level fluctuation at the tidal
boundary, ω = 2π/P is the angular frequency of sea level fluctuation [T-1] and P is the
period [T]. The solution, expressed by Townley using complex variables, is
α
α
cosh
)(cosh
)(⎟⎠⎞
⎜⎝⎛ −
= L
xL
Hxh ap (26)
TP
SLi
2
2πα = (27)
where hp is a complex head, α is complex and 1−=i . The amplitude of
fluctuation in head is equal to |hp| and the phase is equal to arg(hp). Note that in (26)
the group of terms to the right of Ha is an efficiency function that is equal to 1 at x =
0 where the amplitude of groundwater level fluctuations is equal to the amplitude of
sea level fluctuations. The value of this function decays with increasing distance
CSIRO Land and Water APPENDIX C
182 Coast and Clean Seas Project No. WA9911
away from the shoreline and the rate of decay is determined by the value of the non-
dimensional ratio L2S/TP.
The issue of interest is the groundwater flux across the coastal boundary at x = 0
αα tanhL
THQ ap = (28)
where Qp is a complex flux [L2/T]. Because the inshore boundary of the aquifer is a
long way from the coast, equation (28) can be simplified by recognising that
tanh α ≈ 1 when L is large
P
STHQ a
a
2
2π= (29)
In other words, the amplitude of tidal flux, Qa, across the coastal boundary is a
function of the ratio Ha2ST/P. This relation can be applied to typical parameter
values for the coastal aquifer adjacent to Cockburn Sound, which are listed below
0.05 < Ha < 0.25 [m]
0.1 < S < 0.3 [1]
200 < T < 2000 [m2.d-1]
P ≈ 1 [d]
0.2 < Ha2ST/P < 150 [m.d-1]
Figure 81 is a plot of Qa for these ranges and it is clear that the simulated tidal
groundwater flows into and out of the aquifer are the same order of magnitude as the
estimated values of terrestrial SGD into Cockburn Sound. The approximate range of
values in Figure 81 is 1 < Qa < 30 m3.d-1.m-1. It should be expected that benthic flux
meters that are capable of detecting the terrestrial component of SGD would also
detect tidal fluxes.
WAVE-FORCED MODEL
The wave-pumping model simulates vertical fluctuations in groundwater flow
beneath the seabed in response to relatively high-frequency oscillations in sea level
that do not propagate significant distances laterally into the aquifer. For a particular
wave period the governing equations for confined groundwater flow are
APPENDIX C Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 183 and Nutrient Discharge into Cockburn Sound
t
h
K
S
z
h o
∂∂=
∂∂
2
2
(30)
)cos(),0( tHth a ω= (31)
0),( =
∂∂−
z
tDhK (32)
where So is the aquifer specific storativity [L-1], K is vertical hydraulic conductivity
[L/T], D is the thickness of aquifer beneath the seabed [L] and z is distance beneath
the seabed. The solution (Townley, 1995) is
β
β
cosh
)(cosh
)(⎟⎠⎞
⎜⎝⎛ −
= D
xD
Hzh ap (33)
KP
SDi o
2
2πβ = (34)
In this case, the aquifer tidal efficiency is determined by the value of the non-
dimensional ratio D2So/KP. Benthic flux across the seabed at z = 0 is
ββ tanhD
KHq ap = (35)
where qp has units [L/T]. The amplitude of flux is shown below to be a function of
two ratios, Ha2SoK/P and D2So/KP, but the solution is insensitive to values of
D2So/KP that are greater than 2
⎟⎟⎠
⎞⎜⎜⎝
⎛++
−=γγγ
γγπ2cos2exp24exp1
2cos2exp412
2
P
KSHq oa
a (36)
KP
SD o2
πγ = (37)
CSIRO Land and Water APPENDIX C
184 Coast and Clean Seas Project No. WA9911
Typical parameter values for Cockburn Sound are
0.1 < Ha < 0.5 [m]
10-3 < So < 10-5 [m-1]
10 < K < 100 [m.d-1]
10 < P < 15 [s]
25 < D < 35 [m]
0.01 < Ha2SoK/P < 230 [m2.d-2]
1 < SoD2/KP < 1000 [m2.d-2]
Values of qa are plotted in Figure 82, where the approximate range is
1 < qa < 40 m.d-1. Note that these values will be larger when integrated over an area
of seabed and it is clear they are the same order of magnitude and potentially larger
than tidal groundwater flows and estimates of the terrestrial component of SGD into
Cockburn Sound. This result predicts that there can be significant fluctuations of
benthic flux into the out of the seabed over time scale of tens of seconds.
H TS2
/Pa
Qa[m
/d/m
]3
0 20 40 60 80 100 120 140
5
10
15
20
25
30
Figure 81: Tide-pumping model; amplitudes of tidal groundwater fluxes across the coastal boundary calculated for typical values of tidal range and aquifer properties
APPENDIX C Technical Report - January 2003
Quantifying Submarine Groundwater Discharge 185 and Nutrient Discharge into Cockburn Sound
H KS P2
o/a
Qa[m
/d]
0 50 100 150 200
5
10
15
20
25
30
35
40
Figure 82: Wave-pumping model; amplitudes of wave-induced groundwater fluxes across the seabed calculated for typical values of wave height and aquifer properties