Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Can Hydrologic Complexity Simplify Field Level Modeling?
Zachary M. EastonBiological Systems Engineering, Virginia Tech,
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Models
If land use is your only variable, management options are extremely limited
% Forest
% Agriculture
Total P
(mg/L)
100%
100%0%
0%
0.00
0.30
How is management evaluated?
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
• As a result, we have sometimes dogmatically* developed nonpoint source pollution control practices based on specific land uses and ignored the interaction between land management and physical, landscape scale processes.
*dog.ma …, n. … 2. a specific tenet of doctrine authoritatively put forth: the dogma of the Assumption. ‐The Random House Dictionary
X
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group Easton et al., 2008. J. Hydrol.
The HRU concept• Many models define HRUs
as the coincidence of soil type and landuse• Hydrological/chemical
properties are defined at the HRU
• So runoff/P loss is the same here (lowland pasture)
• As here (upland pasture)
Soils Landuse
HRUs
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Models Need to:•Consider spatially distributed properties other than land use and soils
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Models Need to:•Consider the spatial distribution of hydrological patterns
•Account for locations of features (e.g., potential pollutant sources) relative to hydrologically active zones
“Pollutant Sources”Ditch ‐ Stream
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Models Need to:•Consider the spatial distribution of hydrological and chemical patterns
•Account for locations of features (e.g., potential pollutant sources) relative to hydrologically active zones
•Minimize (or eliminate) model calibration
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Hillside
Flood Plain(valley bottoms)
Stream
Baseflow(groundwater)
Shallow bedrock, fragipan, or other restrictive layer
HillsideDrainage
HydrologyScientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Rain
Infiltration
“Runoff”
CN‐method based on this theory (Horton 1933, 1940)Common Perception of Runoff
Infiltration Excessa.k.a. Hortonian Flow
Scientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
March
April
August
September
October
November
1 10 100
20
0
30
50
40
70
60
80
100
90
10
Return Period (yr)
% o
f Are
a G
ener
atin
g H
orto
nian
Flo
wIs Infiltration Excess Runoff
Common?
78%
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Rain
Subsurfacewaterrises
Some areas saturate to the
surface
Saturation Excess RunoffScientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Rain
Rain on saturated areas becomes overland flow
Upland interflow may exfiltrate
Dunne and Black. 1970. Water Resour. Res.
Saturation Excess RunoffScientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Variable Source Areas
Most Watershed Models were not intended to capture this complexity•Soil and Water Assessment Tool (SWAT)•General Watershed Loading Function (GWLF)•Agricultural Nonpoint Source Pollution Model (AGNPS)•Hydrologic Simulation Program Fortran (HSPF)
Scientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
tanln
a
High :
Low :
Topographic Index
33.10
3.52
Topographic Index
10=f(S)
=f(S)
=f(S)=f(S)
=f(S)=f(S)
=f(S) =f(S)f(S)
=f(S)
Easton et al. 2008. J. HydrolEaston et al. 2011. Hydrol. Proc
Wetness IndexClasses
Wetness Index Classes
10921
S
∑
=
i Local StorageS = Watershed Storage
Scientific Background
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
LanduseTI
HRUs
• VSA concept defines HRUs as the coincidence of topographic index and landuse
• So runoff/P loss is now not the same here (lowland pasture)
• As here (upland pasture)• Better Assumption?
Soils
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Variable Source Area HydrologyUSDA WE38, PA
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Topography to explain soil depth and fractions
Using measured pedon data from ARS long term watershed
TI Class (Dry to Wet) TI Class (Dry to Wet)
San
d or
Cla
y (%
)
Soi
l Dep
th (c
m)
Collick et al. 2015. Hydrol Proc
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
ln atan
High :
Low :
Topographic Index
33.10
3.52
Soil genesis as explained byTopography!
Easton et al. 2008 J. HydrolCollick et al. 2015. Hydrol ProcFuka et al. 2015. Hydrol Proc
Wetness IndexClasses
10
1
Zi = local soil depth
Z2= 80 cm
Z6=100 cm
Z10= 130 cm
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group Collick et al., 2015. Hydrol Proc
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Hortonian Dominated Systems Riesel, TX
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
2 4 6 8 10
800
1000
1200
1400
1600
TI Class
Dep
th to
C H
or.(m
m)
R^2=0.28y=50.80 x + 1025.52
(a)
2 4 6 8 105
10
15
20
25
TI ClassM
eas.
Ksa
t(mm
/hr)
R^2=0.19y=−0.88 x + 17.00
(b)
2 4 6 8 1025
30
35
40
45
TI Class
Fiel
d C
apac
ity a
t 33k
pa(%
)
R^2=0.79y=−2.17 x + 46.00
(c)
2 4 6 8 10
1.65
1.70
1.75
1.80
1.85
TI Class
Bulk
Den
sity
(g/c
m^3
)
R^2=0.17y=−0.01 x + 1.75
(d)
2 4 6 8 10
0.48
0.50
0.52
0.54
0.56
TI Class
Poro
sity
(%)
R^2=0.50y=0.01 x + 0.47
(e)
2 4 6 8 100.06
0.07
0.08
0.09
0.10
0.11
0.12
0.13
TI ClassAW
C(v
olum
e)
R^2=0.71y=−0.01 x + 0.13
(f)
Soil Depth
AWC
Field CapacityKsat
PorosityBulk Density
TI Class
Soi
l Dep
th (c
m)
Ksa
t(m
m/h
r)
Fiel
d ca
paci
ty (3
3kpa
%)
Bul
k D
ensi
ty (g
cm
-3)
Por
osity
(%)
AWC
(cm
3cm
-3)
Fuka et al. 2015 Hydrol Proc
TI Class TI Class
TI Class TI Class TI Class
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
ln atan
High :
Low :
Topographic Index
33.10
3.52
Soil moisture as explained byTopography!
Wetness IndexClasses
AWC2= 0.12AWC6=0.09
AWC10= 0.06
Easton et al. 2008 J. HydrolCollick et al. 2015. Hydrol ProcFuka et al. 2015. Hydrol Proc
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
TI adjusted FAO Soils ( ) vs base SSURGO Soils ( )
●
●
● ●
●
●
●
●
800 1000 1200 1400 1600 1800 2000
800
1000
1200
1400
1600
1800
2000
Depth to C Hor.(mm)
Est
. ZM
X
R^2=0.28y=0.33 x + 538.88
(a)Soil Depth
● ●●●●
●
●
●
0 10 20 30 40 50 600
10
20
30
40
50
60
Meas. Ksat(mm/hr)
Est
. Ksa
t(mm
/hr) R^2=0.59
y=2.07 x + −15.38
(b)Ksat
Soi
l Bas
ed S
oil D
epth
(mm
)
Soi
l Bas
ed K
sat(
mm
/hr)
Pit Soil Depth (mm) Pit Ksat (mm/hr)
How does SSURGO perform?
Fuka et al. 2015 Hydrol Proc
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Jan 20 Feb 10
Y=125.0ha
−4
−2
0
2
4
Jan 20 Feb 10
W1=71.2ha
−4
−2
0
2
4
Jan 20 Feb 10
Y2=53.4ha
−4
−2
0
2
4
Jan 20 Feb 10
W6=17.1ha
−4
−2
0
2
4
Jan 20 Feb 10
Y10=8.5ha
Jan 20 Feb 10
Y6=8.5ha
−4
−2
0
2
4
Jan 20 Feb 10
Y8=8.4ha
−4
−2
0
2
4
Jan 20 Feb 10
W10=8.0ha
−4
−2
0
2
4
Jan 20 Feb 10
Y13=4.6ha
−4
−2
0
2
4
Jan 20 Feb 10
W13=4.6ha
W12=4.0ha
−4
−2
0
2
4
Y14=2.3ha
−4
−2
0
2
4
SW17=1.2ha
−4
−2
0
2
4
SW12=1.2ha
−4
−2
0
2
4
−−−
Distr.SSURGOMeas.
PS=0.4ha
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
0.00
0.00
0.00
0.00
0.01
0.10
1.00
2/26/04 2/26/06 2/26/08 2/26/10 2/26/12
Solu
ble
P lo
ss, m
g L-
1
ObservedStd P routinesNew P routines
0.00
0.00
0.00
0.01
0.10
1.00
2/26/04 3/30/07 3/3/08 10/5/09 3/25/10 12/23/1
Solu
ble
P lo
ss, m
g L-
1
ObservedStd P routinesNew P routines
Commercial fertilizer Corn-Hay
Commercial fertilizer and poultry litter Corn-Hay
Collick et al. 2015 JEQ
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
No manure application
0.0
1.0
2.0
3.0
4.0
1/1/2010 2/20/2010 4/11/2010
P lo
ss, m
g L-
1
ObservedStd P routinesNew P routines
Poultry litter/manure application in January (2 tons per ac)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1/1/2010 2/20/2010 4/11/2010
P lo
ss, m
g L-
1
ObservedStd P routinesNew P routines
0
10
20
30
40
50 Prec
ipita
tion,
mm
Collick et al. 2015 JEQ
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Discussion• TI initialization exhibited strong correlations with field level measurements • Resulted in improved predictions of field level contaminant transport, particularly for P
• SSURGO initialization captured soil properties and thus the hydrologic response poorly • Sometimes under predicting and sometimes over predicting, in part due to the overestimated soil depth and AWC
• These results indicate that adjusting model parameters based on topography can result in more accurate P estimation at the field scale
Invent the FutureBiological Systems EngineeringVirginiaTech
Invent the Future
Watershed Science and Engineering Group
Closing Thought"Many... modelers seem to follow... the example of Pigmalion, the sculptor... who carved a statue so beautiful that he fell deeply in love with his own creation... It is feared... hydrologists fall in love with the models they create. In hydrology... the proliferation of models has not been matched by the development of criteria for [evaluating] their effectiveness..." (Dooge 1986, Water Resour. Res. 22‐page S49)