Distributed watershed modelling of a mountainous catchment

Distributed Watershed Modelling of a Mountainous Catchment

Submitted by Shaban Ali Gholami

Thesis Submitted in fulfilment of the requirement for the degree of Doctor of Philosophy

Department of Civil Engineering

INDIAN INSTITUTE OF TECHNOLOGY

NEW DELHI-110 016

MAY 1999

Certificate

This is to certify that the thesis titled "Disti・ibuted watci・shed Modclling of a Mountainous catchrnent" being submitted by Mr. SilABAN ALI GHOLAMI., 94RCE 022 to the Indian Institute of Tecbno1og),Delhi. India, for the award ofthe degree of DOCTOR OF PHILOSOPHY. is a record of bonafide research work carried out by him undei・my sしlpel・vision and guidance. The thesis work, in my opinion, has reached the standard, fulilliug the requirements of the regulation relating to the said degree. 、丁be reseai・ch report and the results presented in this thesis have not heen submitted. in part 01・tri full, to any other university or institute. foi・the award of any degi・ee oi・ diploma.

. G osa in Pro fesso r Depa責ment of Civil Engineering Indian Institute of Technology New Delhi-i 1 OOl6. INDIA え 、

J 奄・

ACKNOWLEDGEMENTS

I am immensely grateflul and deeply indebted to my supervisor Prof. A.K. Gosairi, Professor of Civil Engineering, Indian Institute of Technology (lIT) Delhi, for his keeii interest, constant encouragement supervision and guidance throughout the course of present investigations. His ever helping attitude, excellent leadership and dynamic personality has been or constant source of encouragement for me and also. It is appreciated to him which he has been prepared good facility for working easily in the GIS Laboratory. I am extremely thai

ffiil to Dr. (Mrs.) Sandhya Rao for valuable help and useful suggestions in the present stu

Sincere thanks are due to the soft conservation and watershed management research organization of Ministry of Jihad-e-Sazandegi; for providing the adequate data, topographic maps and Aireal photo and also for perusing my ffinancia and administrative works. I am thankifiti to the Authorities of Research and Training of Deputy of Jihad- e-Sazandegi Ministry under the government of Islaniic Republic of IRAN who awarded scholarship and ffinancially supported to do this research.

I also wish to express my sincerest thanks to all my friends in the GIS Laboratory for their affectionate assistance.

I would like to record my special heart full thank to my wife Mrs. Z.B. Taherian for her excellent cooperation and support during the entire period of this research and also my chiidrens who were a great companion for their mother.

I would like to express my grateflulness to the Minist

of Culture and High Education (MCHE) under the Islamic Republic of IRAb{ for supporting me to do this research.

New Delhi.

April, 1999 (SUABAN ALI GHOLAMI)

Abstract

Watershed management has been recognised as a widely accepted approach for optimal use of soil and water resources. Thei・efo i・e, "atet・shed-model hug incorporating hydrological processes takes a crucial role for propel- planning and development of these local resources. ln"ariably. one of the major diffficulties encountered in the use of hydi・ological models is the reqしliretllent ofcalibratiou ofthe model on the target arca. For the purpose of calibration.

most of the hydrological models i・equire a long sci・les of obser＼・ed rしii-ioft- data. These long series ofdata are usually not available on most of the small watersheds. Therefore. it is imperative that hydrological models should be formulatedl identiffied which can oveにome this requirement. One sLにh hydrological model, namely, SWAT (Soil and Water Assessment Tool) model has been identiffied and implemented ou Tehrau-Amameh. a mountainous, snow bound. 37.20 sq. kni watershed in Iran.

The study has provided an jiトsight into the\＼・orlくing of the continuous・ distributed water balance model. SWAT. The majo，・。可ecti×で ot the stし‘(.1き has been to assess the applicabiliiv o1- the ii1oclel( )n ungauged watersheds.

which bas become a necessity in vieu of the fact that it shall be impracticable to gauge the small wai'ersheds which are accepted as the viable planning and management units in all the countries around the globe.

It has been shown that the ability of the model to be used on ungauged watersheds is not very good since it is not always possible to deffine the model parameters through the observable characteristics. It has been depicted furthermore that through the use of a small series of obsei・×・ed flo'-"

data (about one year), the model performance has appi・eciablき enhanced・lt is therefore inferred that the chai・actel・of the××atei・shed, 'Ahidl-I is i・e fl ee ted thi・ough the observed flO\vS. is not attainable through anき・ other form of information. The structure of the S\\'AT model has proved to be very stable

bしIt lbl・tlle 5110\v collll)ollclll of tllc 1llodcl wllicll 11c^じ（1じ（1 11111)1_、oVelllellt 1111(1 11ad beell modified.

During the course of sin1しiIatioii analysis it lias also been observed that thc selection of an appropriate interval for anafysis is very important. Ii is nei necessary that if the siillulation results appear to he good at the montb!y interval then they shall also be good at the daily interval. However, the other way round may be true.

The capability of the model to simulate the sediii1ent｝・idld has also heen proven. It was very intel・esting to note that the impi・o＼・cillent in the×＼・ater yield simulation kept on bi・inging about the coi・1・esponding improvement i n the sediment yield. This pro"es the capabilit）・（)r the model for sediment y ield simulation beyond doubt.

The capability of SWAT model to generate various management scenarios has aEso been explored. The present catchment being predominanti）・ a rangeland, there were ＼・cry limited feasible altet・natives that 議で1・e considered. The impact assessment due io changes in rate of grazing.

duration of grazing. and crop rotation. as 'veli as changes due to utilization ofたrtilizer輿itrate and Phosphate). has been e＼・aluated OU the catchment. lt has been obser\led that the various alternatives considered have no1 made appreciable difference in ihe, outputs. Uo凧・C×・el・． it FflZIき． l)e said tllat this ability of the SWAT model shall go a long×＼av in cnhancini the acceptability of the model×＼・ith the end-users and shall make the model×で1き．

effective tool ofthe future.

1j 1

1 っj 11 11 3 一6 せ玉且 ■■且 19 24 28 釦 30 39 41

Table of Contents

Chapter 1 Introduction I

1. 1 Geirei-al

I

12 Objecti;'e of/lie Sti' 心 ²

1.3 Selectioji Oft/ic Simulation Model

1.4 The Su ゆ Area 6

1.5 i 如 del Desci-iption 7

I. 6 Model hip" 'vid Preprocessing 7

L 7 Seiisitivitj' Analysis

1_ 8 The Model Simulation 9

Chapter 2 Literature Rei'ew

2.1 Ge,,ei-a!

2.2 Ove;・view of J'VfitCrSlle( Sinuifation il勤des Deterministic Models

Conceptual 1'Todels Lumped iViodels Distributed Models Semi-Disti-ilnited Models

23 Litei-<'tui-e Review oft/ic即<fi-ologica! Piりcesses 2.3.1 tiiterceptioi'

2.3.2 Suo", indlt and accuniulatioii Energy Balance Approach

43 43

お 45 49 5() 3 一『一 7 9 3 4 4 4 っ1 0 一う 1さ 6 7 11 「J 『つ 4 QO つー ーう 6 「1 の0 一1, 一I', 一り ーさ ーり 一0 6 ノ0 6 っ1 一1 ツ1 「1 一1 一K OO n6 b6 Oつ Aフ 9 9 0ノ aノ

一 2 4 6 6 0 0 0 0 1 1 1 1 一 一

一

Air tenpeia加に ＿＿ 一

Radiation

Wind and Vapoしu・Pressure:

Degree-Day Method (Temperature Index)

Coillbiilation Energy-Temperature Index Sno'vmetl Model ( E-T mlldcx>、 Sno'v Accumulation and Runoff

2.3.3 Evapotranspiration Process (ET)、

Potential Evapotranspiration(PET)__ --- Reference Evapoti・anspiration(ETO） 一－一 Actual Evapoti・an spil・ation (A ET)

2.3.4 Inflitiation Process Rainfall Excess Method:

Index Models

The Soil Conservation Services (SCS) Method

Empirical lii丘ltration Models __

Approximate models ___

Physically Based Models

2・3・5 Depression and Pouding Sto'・age P'・ocess Water balance Method

Method based on Map Measurmeut________ _____-_

2.3.6 Runoff P' ocess Infiltration excess method

Green-Anipt method ----

Soil Conservation Sel-vice (SCS) method 2.3.7 Percolation Process

2.3.8 Late'・al Subsui・face Flow EPIC Model

SWAT model SPUR MODEL

2.3.9 G'・ound "ate'・Process

2.3.10 Ei・OSIOI1 alh1 Sedinicut p'・（)CCSS USLE Model

MUSLE Model ANSWERS Model

フI Gノ ーU I つ一 「j く」 I0 10 ll ll H H H

117 II 7

CREAMS Model S\VRRB Model EPIC Model WEPP Model SWAT Model SHESED Model PSIAC METHOD

C/i apter 3 The S/u

む

了．

・

3.2.1 . i Geology and Geomorpholog_｝・＿

3 .2. 1 .2 Topography 3.2. I .3 Soil Survey data 3.2. I .4 Land-use

3.2. t .5 Agricultural Management Practices 3.2.2 Hydrometeorological Data

3.2.2.2 Precipitation

-D．ユ．2.3 Tempe'・ature and Other Climatic Data 3,2.2.4 Hydrometric Data

3. 3 Th e Existence (111(1 A t'ai/abiliひ(?f Data 157

3.3.1 The Sources ofthe Collected Data 3.3.2 C0n11)UteI・－based Data Pi・ocessi ng 3.4 Analysis of Data

3.4.1 Precipitation 3.4

3.4.1.2

0 っ1 4 0 っー ho nV つー つ一 2 っー っ一 rJ tj 「j 一つ ーソ ーソ 11 1 1 1 1 1 1 1 1

Rainfall Data Snow fall

ワJ 00 一U nU ni つ一 1き 一b 6 6 6 6 1 1 11 1 - 1

3・4.2 Analysis of Temperatut・e

Chapter 4 Catchment Modelling

4. 1 I11f0(ItjCtjOfl 4.2 Model Selection

3 6 6 7 6 6 6 6 IA 1

1

11 曹1

167 I 69

69 73 76 77

179 4 5 一さ 6 6 6 7 8 1A つー っコ ーさ ーっ 8 5 8 5 5 5 8 8 9 9 9 9 9 1 1 電1 l l

l ll ll lt ll 11 11 1

198 4.4 11'o(k'! Coiiijmizeizts

4.4.1 I-lydi・010g》・

4.4.!.I sしIt・face IえLt ilO ff 4.4. I .2 Percolation

4.4.1.3 LateralSし山sut・face Flow 4.4.1.4 Gt・oしt tid-'へ，ater Flow 4.4. 1 .5 Evapotranspiration 4.4.!.6 Sno"IlieIt

4.4.1.8 Ponds 4.4.2 Weather・

4.4.2.】 Precipitation

4.4.2.2 Air Temperature and Solar Radiation 4.4.2.3 Wind Speed and Relative Huirtidit》・

4.4.3 Sediment Yield 4.4.4 Soil Temperature 4・4.5 Crop Gro'、・tu、

4.4.6 Nutrient

4・4.7 Agi・ic til tui・al Management 4.4.8 Routing Comnponent in Ch、annel

4.4.8.1 Channel FIo'v Routing

4.4.8.2 Channel Sediment Routing_

Ch apter 5 Preprocessingfor Su,n u/a/ion

5.1 Slvat Preprocessor

エ2 upu! Output Data Orgaizisatioii 5.2.2 Dynamic Data

5.3 Srntia/ Dala

5・3.1 Pi・oCcssii'g of Contours 5.3.2 \Vatershed Delineation

5・3.3 Generation ofbasin and Di・ai"age Netwoi・1< Characteristics

201 201

204 205 206 208 208 211

5.3.4 Watershed Delineation Cizaptei

6 Sensitivity Analysis 6.1 Introduction

6.2 Statistica' Criteria for Sensitivity Analysis 6.3 Sensitivity Analysis ofthe Model Parameters_

6.3. 1 Snow Threshold Coeffficient (TC) 6.3.2 Curve Number (CNII)

6.3.3 Snow Bound Accumulated Coeffficient (T-Laps) 6.3.4 Snow Melting Coe

cient (CCT)

6.3.5 Depth of Soil

6.3.6 Saturated Conductivity (SC)

6.3 .7 Alfa Factor for Groundwater (ABF) 6.3 .8 Revap Coeffficient (REVAPC) 6.3.9 Crack-flow (CRCO)

6.4 Discussion on Sensitivity of Model Chapter 7 Model Applicatoiz

7.1 Water Yield Simulation

71.1 Simulation of Ungauged Catchment 7. 1 .2 Improvement in Simulation

7. 1 .2. 1 Authentication of the Improvement in Simulation 7.1 .3 Simulation ofthe whole Catchrnent

7. 1 .4 Simulation Performance on the Remaining Period 7.2 Sediment Yield Simulation

7.2. 1 Simulation ofUngauged Catebment 7.2.2 Improvement in Simulation

7.2.3 Total Sediment Yield using PSIAC method

217 226 226 227

L)0

230 ,, -.,4 235 - - z.-,7 .. - L-,8 238

2n L.)9 ,' n L-)9

240 241 260 261 261 265

-, n-, L '

284

285

287

288

288

291

7.3 Scenario Generation 299

7.4 Result and Discussion

Chapter 8 CouI clusio,z -'-,7

Rej レ ^{en ces}

A 即 ^end な ^I

A 即 ^end な ^II

Bio-data げ ^{A uitlior}