EVALUATION OF URBAN FLOODS CONSIDERING CLIMATE CHANGE IMPACTS

EVALUATION OF URBAN FLOODS CONSIDERING CLIMATE CHANGE IMPACTS

Mohammad KaramouzProfessor, School of Civil Engineering, University of Tehran, Email: [email protected]

Ana HosseinpoorM.Sc. student, School of Civil Engineering, University of Tehran, Email:

[email protected]

Sara NazifPh.D. candidate, School of Civil Engineering, University of Tehran, Email:

[email protected]

Evaluation of urban floods considering climate change impacts 2

Introduction The urban floods could be very destructive in high population density

zone and in centers of economic and industrial activities.

The urbanization changes natural morphology of local rivers and often decreases their natural carrying capacity resulting from various activities and therefore intensifies the risk of urban floods.

The effects of climate change on hydrologic cycle have led to increased concerns about urban flood events especially in mega cities of the developing countries.

In this Study, the performance of drainage system of northern part of Tehran metropolitan area (capital of Iran) has been investigated


Case studyThe The northeastern part of Tehran

•51º-22´ and 51 º-30´ longitudes

•35º -42´ and 35º -53´ latitudes

•Includes

•Darband sub-basin(zone 1)

•Golabdare sub-basin(zone 2)

•Velenjak sub-basin(zone 3)

•Sadabad sub-basin(zone 4)

•Kashanak sub-basin(zone 5)

•Jamshidie sub-basin(zone 6)

•Zones 7- 19


zoneArea(hec)

12216.5

2639.8

3306.9

4803.16

5473

6188

71511.1

8424.5

9448.7

10242.9

11213.7

1240.46

13662.37

14734

15607.59

1622

17128.49

18913.68

19366.5


Types of channels •Natural

•Man made

average slope: About 21%. The percentage of impervious area: 85%

The altitude: Between 1290 m and 3900 m.

Total drainage area: About 110 km²

Case study


Data:

The rainfall data at Roodak station located in latitude 35º-51´ and longitude 51º-33´ with adequate recorded data which is out of the study area has been considered as a representative of the rainfall data of the study area.

The Ghasr station snow data has been used which is available between years 1975 and 1995.

Mehrabad (synoptic station ) is used as a source for wind and temperature

data which is about 10 Km away from the study area.

Case study


Three scenarios have been considered for studied area drainage system

modeling as follows: Scenario 1: The surface water collection system in about 10 years ago.

•The model for this scenario is developed to evaluate the effectiveness of development projects done in the recent years.

Scenario 2: The present situation of surface water collection system.

Scenario 3: The future plans for improving and development of case study drainage system have been modeled in this scenario.

Case study


Case study(a) Scenario 1

(Past)

(b) Scenario 2(Present)

)c (Scenario 3)Future(

Differences between considered scenarios

Added

abundant


Case study

Percentage of

impervious area (%)

Number of detention

ponds

Total Capacity of detention

ponds (m3)

Natural channel

Man made closed channel

Man made open

channel

Scenario 1 :814 7500 17720 16940 45690

Scenario 2 85 59450 16170 21095 30650

Scenario 3 :90713380 16170 29915 30650

The characteristics of the different scenarios for water drainage system


Case study

Governing Parameters in the scenarios:

The channel coverage and alignment

The number and placement of detention ponds

The land use and percentage of pervious area


Methodology For evaluation of climate change effects on urban floods a statistical

downscaling model (SDSM) developed by wilby et al. (2004) is used.

The drainage system of the study area has been simulated using StormNET model.

The critical rainfall that may result in probable floods in the region are those which satisfy the following inequality:

Where Where μμіі and and σσіі are the average and standard division of rainfall in season i of the rainfall series, are the average and standard division of rainfall in season i of the rainfall series,

respectively. Rxi is considered the extreme rainfall in season irespectively. Rxi is considered the extreme rainfall in season i

iiiRx *2


The effects of climate change on the magnitude and frequency of the extreme seasonal rainfalls in the future are evaluated.

The hydraulic model of the drainage system is calibrated.

The impact of rainfalls with different return periods is evaluated running the hydraulic simulation model for the three pre-defined scenarios in the study area.

Methodology


Methodology

Rainfall downscaling: SDSM carries five distinct tasks:

Screening of potential downscaling predictors

Assembly and calibration of model

Synthesis of ensembles of current weather data using observed predictors

Generation of ensembles of future weather data using GCM-derived predictor variables

Diagnostic testing/analysis of observed data and climate change scenarios


Methodology

Rainfall downscaling:

SDSM model has been used for long-lead rainfall prediction and downscaling for an individual site on a daily time–scale, by using GCM outputs.

During the downscaling with the SDSM, a multiple linear regression model is developed using selected large-scale predictors and the local rainfall.

Large-scale relevant predictors are selected using correlation analysis, partial correlation analysis and scatter plots, considering the sensitivity between the selected predictors and rainfall for the region


Methodology

Hydraulic modeling of urban drainage system :

The StormNET model developed by Boss International (2005) has been used for simulation of the urban drainage system.

StormNET is a link-node based model that performs hydrology, hydraulic, and water quality analysis of stormwater and wastewater drainage systems.


Methodology

Hydraulic modeling of urban drainage system :The StormNET model needs some data on:

Sub-basin (total area, pervious and impervious area, manning’s roughness,…) Detention pond (shape of detention pond, elevation,…) Flow diversion Snow pack Rainfall hyetograph Channel and pipe links (shape, manning’s roughness,…)


Methodology Hydraulic modeling of urban drainage system: It is important to consider the resulted runoff due to snow melting in

modeling of the drainage system.

The snow melt coefficient has been calculated as follows:

M is snowmelt runoff (mm), D is average number of degree day above zero (the snow melt base temperature has been determined 0°c for Tehran) and K is the snowmelt coefficient.

The following degree-day equation has been used to compute the melt rate:

Melt Rate = (Melt Coefficient) (Air Temperature - Base Temperature)

D

MK

D

MK

D

MK


Methodology

Hydraulic modeling of urban drainage system : The SANTA BARBARA method :Simulation of sub-basin runoff

The KIRPICH method: Estimation of the basins time of concentration.

The Manning's roughness for pervious area 0.015

The Manning's roughness for impervious area .0149.

The soil property group D

CN of the pervious area varies between 76 and 84.


Results> Downscaling Daily rainfall data at Roodak station has been transformed by the second

root function to better fit normal distribution The correlations between different combinations of available predictors and

daily rainfall have been calculated to find the most appropriate combination.

The combination of 3 predictors is selected 1) relative humidity at 850 hPa height, 2) near surface specific humidity3) near surface relative humidity.

The physical relation between the selected predictors and the rainfall of the study area has been implicitly considered by calculation of P-value between predictors and rainfall.

The model has been calibrated with rainfall data of 1977-1984 and validated for the remaining available data (1985-2000).


Results Errors of rainfall prediction include errors of mean and maximum daily rainfall

and wet spells, during the validation period for NCEP( National Center For Environmental Prediction) and HadCM3(Second Hadley Centre Coupled Ocean-Atmosphere GCM)

Signal source variable MAE )Mean Absolute Error(

RMSE )Root Mean Square

Error(

NCEP Mean rainfall (mm)

0.07 2.08

Maximum rainfall (mm)

0.11 5.09

Wet spell (hr) 3.4 .48

HadCm3 Mean rainfall (mm)

0.19 2.65

Maximum rainfall (mm)

0.12 3.4

Wet spell (hr) 4.1 .62


Results

0

5

10

15

20

25

30

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Rai

nfal

l (m

m)

Observed rainfall

Modeled rainfall using NCEP signals

•20 ensemble data of rainfall have been generated.


Results YearNumber of extreme rainfalls Average of extreme rainfalls

Spring Summer Fall Winter Spring Summer Fall Winter

2007121101329.03 17.8 27.06 46.56

201711281749.7 33.46 48.42 42.45

2027231121722.18 76.99 49.45 38.61

203720191126.25 15.4 32.87 62.65

20471508826.12 048.6 50.77

2057180161631.24 037.94 40.27

20671809937.39 035.69 38.66

2077180121232.85 056 45.15

20871919940.1 7155.9 46.44

2097150141442.85 047.41 59.44


Results The hydraulic model of the drainage system is calibrated with the observed

1995 flood hydrograph in Golabdareh (zone 2).

0

50

100

150

200

250

300

0 0.5 1 1.5 2 2.5time(hr)

Q(c

ms)

Simulated hydrograph

Observed hydrograph


Results The severity of wet and dry periods is increasing in the study area due to the

effects of climate change

The peak volumes of the floods are increasing

Increasing the peak volumes lead to considerable damaged and it is necessary

to revise the river training projects in the study area

Application of some new river training programs for dredging the channels and construction of new detention and retention ponds are needed.


Results

No obvious trends in the volume and peak of floods due to climate change effects could be obseved.

The man-made channels have changed the characteristics of the drainage system. The positive effects are not uniform.

In response to the river training projects in the last 10 years, the flood peaks and volumes have been increased considerably.

As the capacity of the system increases, the overflow volumes variations are mixed.


Conclusion In urban areas due to some special characteristics such as the population

concentration and limitations on the natural water systems, the effects of climate change are intensified.

One of the most important components of urban water cycle is urban runoff which is highly affected by climate change and urbanization.

In this study the effects of climate change on urban runoff in the northeast of Tehran is evaluated.

The downscaling model has been used to predict the future rainfall and then extreme rainfalls are identified.


Conclusion The characteristics of the extreme rainfall of future years including the

frequency and the magnitude are evaluated.

It seems that the severity of wet and dry periods is increasing in the study area due to effects of climate change. Further study is needed.

The identified extreme rainfalls are applied in a hydraulic simulation model considering the existing and the future expansion of the system.


Conclusion The results show that the peak volumes of the floods are increasing.

This may lead to considerable damaged and it is necessary to revise the current plans for river training projects of the drainage channels in the study area.

An integrated approach is needed to deal with the combined urban expansion and the climate change impacts.

Thank you for your Attention

For more information please contact:

Mohammad Karamouz: [email protected]

Tel: 0098-21-88555884

Fax:0098-21-88701507


Methodology Hydraulic modeling of urban drainage system : The flow diverted through a weir flow diversion is computed through the

following equation:

is the diverted flow (m3/s), is weir coefficient, is weir height. f is a coefficient that is computed as follows:

Qin is inflow to the flow diversion, Qmin is the minimum flow at which flow diversion begins and Qmax is the maximum capacity of the channel.

All of the weirs in the system are assumed to be rectangular.

)*( wwdiv HfCQ divQ

wC wH

minmax

min

QQ

QQf in


Results The extreme rainfall with different return periods are simulated by three

developed hydraulic models.

0

1000000

2000000

3000000

4000000

5000000

Vo

lum

e (m

^3)

140

160

180

200

220

240

260

Q (

cms)

Flood Volume

Overflow Volume

Flood Peak

0

1000000

2000000

3000000

4000000

5000000

2007

-20

2007

-200

2017

-100

2027

-50

2037

-20

2037

-200

2047

-100

2057

-50

2067

-20

2067

-200

2077

-100

2087

-50

2097

-20

2097

-200

Vo

lum

e (m

^3)

140

160

180

200

220

240

260

Q (

cms)

Flood Volume

Overflow Volume

Flood Peak

0

500000

1000000

1500000

2000000

2500000

3000000

3500000

4000000

4500000

2007

-20

2007

-200

2017

-100

2027

-50

2037

-20

2037

-200

2047

-100

2057

-50

2067

-20

2067

-200

2077

-100

2087

-50

2097

-20

2097

-200

Vo

lum

e (

M^

3)

140

160

180

200

220

240

260

280

300

320

Q (

cm

s)

Flood Volume

Overflow Volume

Flood Peak


Results

The average of long-term observed rainfall has been compared to the rainfall using NCEP signal and modeled rainfall using HADCM3

One of the predicted scenarios for future climate variation named HadCM3 (Second Hadley Centre Coupled Ocean-Atmosphere GCM) is used as the model input signal and rainfall is predicted

Date post:	08-Jan-2016
Category:	Documents
Upload:	steffi
View:	21 times
Download:	1 times

EVALUATION OF URBAN FLOODS CONSIDERING CLIMATE CHANGE IMPACTS

Documents