Watershed Models

WATERSHED

MODELS

A CRC title, part of the Taylor & Francis imprint, a member of the Taylor & Francis Group, the academic division of T&F Informa plc.

EDITED BY

VIJAY P. SINGH

DONALD K. FREVERT

WATERSHED

MODELS

Published in 2006 by CRC Press

Taylor & Francis Group

6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742

© 2006 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group No claim to original U.S. Government works

Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1

International Standard Book Number-10: 0-8493-3609-0 (Hardcover) International Standard Book Number-13: 978-0-8493-3609-6 (Hardcover) Library of Congress Card Number 2005044000

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use.

No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Singh, V. P. (Vijay P.)

Watershed models / Vijay Singh, Donald Frevert. p. cm.

Includes bibliographical references (p.). ISBN 0-8493-3609-0 (alk.paper)

1. Watersheds—Mathematical models. 2. Watershed Management—Mathematical models, 3. Hydrologic models. I. Frevert, Donald K. II. Title,

GB980.S626 2005

551.48 – dc22 2005044000

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

Taylor & Francis Group is the Academic Division of T&F Informa plc.

Dedicated to watershed modelers around

the globe.

vii

Contents

Preface xi

Acknowledgments xix

Contributors xxi

Section 1: Preliminaries

1.

Introduction 3

Vijay P. Singh and Donald K. Frevert

2.

History and Evolution of Watershed Modeling

21

Derived from the Stanford Watershed Model (SWM)

Anthony S. Donigian, Jr. and John Imhoff

3.

Regional Calibration of Watershed Models

47

Richard M. Vogel

Section 2: Large Watershed Models

4.

Large Scale Hybrid Watershed Modeling

75

Mustafa M. Aral and Orhan Gunduz

5.

Simulation of Water and Energy Budgets Using a

97

Macroscale Hydrological Model for the Upper

Mississippi River Basin

Rajagopalan Srinivasan and Venkat Lakshmi

Section 3: Streamflow Models

6.

Gridded Surface/Subsurface Hydrologic Analysis

131

(GSSHA) Model: A Model for Simulating Diverse

Streamflow-Producing Processes

Charles W. Downer, Fred L. Ogden, Justin Neidzialek, and Siqing Liu

7.

USGS Modular Modeling System (MMS) –

159

Precipitation-Runoff Modeling System (PRMS)

George H. Leavesley, Steve L. Markstrom, and Roland J. Viger

8.

The Xin’anjiang Model on Digital Basin Platform

179

viii

Section 4: Streamflow and Water Quality Models

9.

A First Principle, Physics-Based Watershed Model:

211

WASH123D

Gour-Tsyh Yeh, Guobiao Huang,

Hwai-Ping Cheng, Fan Zhang, Hsin-Chi Lin, Earl Edris, and David Richards

10.

Flexible Integrated Watershed Modeling with

245

MIKE

SHE

Douglas N. Graham and Michael B. Butts

11.

Better Assessment Science Integrating Point

273

and Nonpoint Sources (BASINS)

Paul B. Duda, Jack L. Kittle Jr.,

Anthony S. Donigian, and Russell Kinerson

12.

MEDIFIS: A Physically Based, Spatially-

291

Distributed Runoff and Erosion Model for Extreme

Rainfall

Events

Joao P. Nunes, G. Nuno Vieira and J. Seixas

13.

BAYMOD: Modeling Irrigated Catchments Using the 315

Streamflow

Integral

Approach

Luke D. Connell, M. Gilfedder, and Russell Mein

Section 5: Urban Watershed Models

14.

EPA Storm Water Management Model, SWMM5

339

Wayne C. Huber, Lewis A. Rossman, and Robert E. Dickinson

15.

IDEAL: Integrated Design and Evaluation

361

Assessment of Loadings Model

Bill J. Barfield, J.C. Hayes, S.L. Harp, K.F. Holbrook, and J. Gillespie

16.

SEDIMOT III Model

381

Bill J. Barfield, J.C. Hayes, E. Stevens, S.L. Harp, and A. Fogle

Section 6: Agricultural Watershed Models

17.

The SPAW Model for Agricultural Field and Pond

401

Hydrologic

Simulation

ix

18.

The APEX Model

437

Jimmy R. Williams and R.C. Izaurralde

19.

GAMES: The Guelph Model for Evaluating the

483

Effects of Agricultural Management Systems on

Erosion

and

Sedimentation

R.P. Rudra

Section 7: Planning and Management Models

20.

Use of Distributed Models for Watershed

503

Management: Case Studies

M. Arabi, Rao S. Govindaraju, M. Sophocleous, and J.K. Koelliker

21.

RiverWare 527

Edith A. Zagona, Tim Magee,

H. Morgan Goranflo, Terrance Fulp, Donald K. Frevert, and Jerry L. Cotter

22.

A Parsimonious Watershed Model

549

James F. Limbrunner, Richard M. Vogel, and Steven C. Chapra

23.

MODSIM: River Basin Management Decision

569

Support

System

John W. Labadie

24.

Water Rights Analysis Package (WRAP) Modeling

593

System

Ralph A.. Wurbs

25.

Hydrological River Basin Environment Assessment

613

Model

(Hydro-BEAM)

Toshiharu Kojiri

26.

The State of Colorado’s Stream Simulation Model

627

(StateMod)

Ray Alvarado and Ray Bennett

xi

PREFACE

There are dozens of mathematical models of watershed

hydrology available today in the world, but these are not found in

one place. In 1995, Vijay Singh edited a book entitled Computer

Models of Watershed Hydrology which contained 26 popular

computer models from around the globe. In 2002 Singh and

Donald Frevert edited two books entitled Mathematical Models of

Large Watershed Hydrology and Mathematical Models of Small

Watershed Hydrology and Applications. Clearly, there were

several worthy models that were not included in these books. The

motivation for putting together this book stemmed from the desire

to provide, under one cover, a comprehensive account of some of

those popular mathematical models of watershed hydrology that

were not included earlier. The objective of this volume is to

include a variety of models that span a range of characteristics,

such as representativeness, comprehensiveness, broad-based

applicability, and use of modern tools. Based on these

considerations, 24 models were selected for inclusion in this

volume. It is hoped that these models fulfill the intended objective.

Because there is a large number of models available these days,

any number of combinations of the models could satisfy the

intended objective equally well. Therefore, in any model selection,

personal bias is unavoidable, and the models included here may

reflect our personal bias. This model selection in no way implies

an endorsement of the models included or a rejection of those not

included. We personally have used and like very much some of the

models not included here.

The subject matter of this volume is divided into seven

sections encompassing 26 chapters. The first section contains three

chapters. Beginning with introductory remarks on watershed

modeling in Chapter 1, a history and evolution of watershed

modeling derived from the Stanford Watershed Model (SWM) is

presented in Chapter 2. It traces the refinement of SWM and its

coupling with the Agricultural Runoff Management Model (ARM)

and Nonpoint Source Pollutant Loading Model (NPS) into

Hydrological Simulation Program-Fortran (HSPF), and goes on to

describe successive enhancements of HSPF up to the most recent

xii

Release No. 12 in 2001. It then discusses the software tools, such

as Interactive hydrologic analyses and data management (ANNIE),

Watershed Data Management (WDM), GENeration and analysis of

model simulation SCeNarios (GenScn), and Expert system for

calibration of HSPF (HSPEXP) products, developed by the U.S.

Geological Survey (USGS), which have greatly enhanced

watershed modeling in general and HSPF in particular. Integration

and enhancement of the strongest features of HSPF and these

USGS software products culminated in the Environmetal

Protection Agency’s (EPA) Better Assessment Science Integrating

Point and Nonpoint Sources (BASINS) modeling system.

Furthermore, HSPF has been integrated into the U.S. Army Corps

of Engineer’s Watershed Modeling System (WMS). The chapter

concludes that HSPF/BASINS serves as a focal point for

cooperation and integration of watershed modeling and model

support activities between the USGS and the EPA, and provides an

opportunity for the use of common tools and methodologies by

federal agencies. Chapter 3 reviews numerous approaches for the

regionalization of watershed models, and introduces a new

methodology for the regionalization of watershed models. The

approach involves concurrent calibration of a watershed model to

many sites in a region.

Large-scale watershed modeling constitutes the subject of

Section 2, comprising two chapters. Chapter 4 discusses a hybrid

surface/subsurface flow and transport model that blends the

powerful distributed parameter models with relatively simple

lumped parameter models. This hybrid formulation decreases the

computational requirements and at the same time provides a

representative description of the watershed flow processes. Chapter

5 presents a macroscale hydrological model to understand the

water and energy balance for the Upper Mississippi River basin. It

shows that the spatial and temporal variability of droughts and

floods can be analyzed using the model-simulated soil moisture

regimes.

The subject of Section 3 is streamflow models and

encompasses three chapters. Chapter 6 presents a physics-based,

distributed parameter Gridded Surface and Subsurface Hydrologic

Analysis model (GSSHA) simulating diverse

streamflow-producing mechanisms. This model is a successor of the

xiii

dimensional, physically based Hortanian model CASC2D and has

been coupled to the Department of Defense Watershed Modeling

System (WMS). Chapter 7 presents the Modular Modeling System

(MMS) of the Precipitation-Runoff Modeling System (PRMS)

developed by the USGS. MMS is an integrated system of computer

software to support the development and integration of a wide

variety of hydrologic and ecosystem models, whereas PRMS is a

physics-based hydrologic model. The integrated system includes

spatial data analysis using geographical information system (GIS),

statistical and graphical analysis tools, and a variety of parameter

estimations, sensitivity analysis, and optimization capabilities.

Chapter 8 presents the latest version of the Xin’anjiang model on a

digital platform. The discussion includes input requirements,

procedures and assumptions used, output capabilities, and two case

studies illustrating its application.

Section 4 deals with streamflow and water quality models

encompassing five chapters. Chapter 9 focuses on a Numerical

Model of Water Flow and Contaminant and Sediment Transport in

WAterSHed Systems of 1-D Stream-River Network, 2-D Overland

Regime, and 3-D Subsurface Media (WASH123D). It addresses

particular features of WASH123D in the treatment of interactions

among media interfaces, the inclusion of various types of control

structures and pumps, the formulation of reaction-based water

quality simulations, and the implementation of optional

hydrodynamics in river network and overland regime. The design

capability and demonstrative examples ranging from seconds to

years in temporal scale and from meters to hundreds of kilometers

in spatial scales are presented.

The physically based, distributed, integrated hydrological

and water quality modeling system (MIKE SHE) is the subject of

Chapter 10. It is a modular modeling system that allows

mix-and-match of simple, lumped-parameter or water balance methods with

advanced, physically based, finite-difference methods for each of

the hydrologic processes. MIKE SHE is directly linked to both, a

1-dimensional river flow modeling module (MIKE 11) and a

module for modeling of urban sewer systems (MOUSE). MIKE 11

is used in applications ranging from simple routing of surface

water to fully dynamic channel flow with dynamic flow control

structures. MOUSE is used with MIKE SHE to study the

xiv

interaction between urban infrastructure networks and surface/

subsurface hydrology, in either steady-state or fully dynamic

applications.

Chapter 11 discusses BASINS developed by the EPA’s

Office of Water to facilitate examination of environmental

information, to support analysis of environmental systems, and to

provide a framework for examining management alternatives. It is

a multipurpose environmental analysis system for use by regional,

state, and local agencies in performing watershed- and

water-quality-based studies. BASINS is designed to be flexible,

supporting analyses at a variety of scales using tools that range

from simple to sophisticated.

Chapter 12 presents the process-based, spatially-distributed

hydrological and soil erosion model (MEFIDIS) runoff and erosion

model. This model simulates a single extreme event, taking a

raster-based approach to spatial distribution, and is dynamic in

time. Model performance and robustness are analyzed using a large

number of measured events.

Chapter 13 presents an alternative to spatial distribution for

physically based process modeling of catchment flow and solute

transport. The approach is based on an integral relation for

catchment streamflow which describes the accumulation through

the stream network of the hillslope contributions. The model is

tested through application to an irrigated catchment.

Section 5 deals with urban watershed models and is

comprised of three chapters. Chapter 14 presents the EPA’s Storm

Water Management Model (SWMM) widely applied in the United

States, Canada, and around the world since 1971 for analysis of

complex hydrologic, hydraulic, and water quality problems

associ-ated with urban drainage. It provides a brief description of the

history of the program leading to the development of the most

current version: SWMM5. The SWMM5 graphical user interface is

described as well as algorithms used to simulate rainfall-runoff,

conveyance (system hydraulics), water quality, and treatment

processes. Information is also presented about parameter input and

estimation, output options, sensitivity and optimization, and user

resources.

Chapter 15 presents the model IDEAL developed to predict

runoff and pollutant loadings from urbanized watershed and to

xv

evaluate the impact of stormwater structural BMPs on the loading.

Chapter 16 presents a lumped model SEDIMOT III developed to

predict runoff and sediment loadings from watersheds in transition

from undisturbed to disturbed conditions and to evaluate the

impact of stormwater and sediment structural BMPs on the

loading. The model makes calculations for a single storm based on

user inputs of precipitation and watershed characteristics.

Section 6 deals with agricultural watershed models

comprising three chapters. Chapter 17 discusses the

Soil-Plant-Air-Water (SPAW) computer model which simulates the daily

hydrology of agricultural fields and ponds including wetlands,

lagoons, and reservoirs. Data input and file selection are by

graphical screens. The program and descriptions include theory,

data requirements, example files and applications, and operational

details. Chapter 18 deals with the Agricultural Policy/

Environmental eXtender (APEX) model developed for evaluating

various land management strategies considering sustainability,

erosion (wind, sheet, and channel), economics, water supply and

quality, soil quality, plant competition, weather, and pests. The

individual field simulation component of APEX is taken from the

Environmental Policy Integrated Climate (EPIC) model. The

APEX model extends the EPIC capabilities to whole farms and

small watersheds with the addition of components for routing

water, sediment, nutrients, and pesticides across complex

landscapes and channel systems. The APEX components (subarea,

routing, reservoir, groundwater, grazing, manure management, and

feedlot dust emission and distribution) are described.

GAMES, the Guelph Model for Evaluating the Effects of

Agricultural Management Systems on Erosion and Sedimentation,

constitutes the subject matter of Chapter 19. It was developed as a

screening tool for watershed management. The application of this

modeling concept to small agricultural watersheds can be used to

identify areas of excessive soil loss and sources of sediments,

selection of remedial strategies, and developing cost-effective

monitoring programs. The model concept has been extended to

include phosphorus in a sister version of the GAMES model called

the GAMESP.

Section 7 deals with applications of hydrologic models to

watershed planning and management and is comprised of seven

xvi

chapters. Chapter 20 deals with case studies using distributed

watershed models for watershed management. Experiences

primarily with the Soil and Water Assessment Tool (SWAT)

model over several watersheds in Kansas and Indiana are

described. Chapter 21 discusses a general river and reservoir

modeling tool, called RIVERWARE, developed by the U.S.

Bureau of Reclamation (USBR) for operational scheduling and

forecasting, basin planning, policy evaluation, and other analysis

and decision processes. Solution procedures include data-driven

simulation, rule-based simulation using user-specified logical

policy statements, and linear goal programming optimization, and

can include water ownership accounting. RIVERWARE is used by

the Tennessee Valley Authority (TVA), USBR, U.S. Army Corps

of Engineers, and other agencies to facilitate decisions ranging

from operational schedules to Environmental Impact Statement

(EIS) analysis.

Chapter 22 presents a parsimonious daily water balance

model in that it has only four adjustable parameters. It combines

empiricism and mechanistically based elements, following in the

tradition of the U.S. Soil Conservation Service (now the Natural

Resource Conservation Service) Curve Number (CN) Method. The

model simulates daily variations in evapotranspiration, soil

moisture, saturated groundwater, groundwater outflow, snow

accumulation, snowmelt and streamflow. Chapter 23 deals with a

Generalized River Basin Management Decision Support System

(MODSIM) which is used for problems ranging from short-term

water scheduling including flow routing to long-term operational

planning for helping to resolve conflicts between urban,

agricultural, and environmental concerns. A powerful graphical

user interface connects MODSIM with the various data base

management components, and a state-of-the-art network flow

solver assures that water is allocated according to physical,

hydrological, and institutional/legal/administrative aspects of river

basin management. MODSIM is coded in the new MS.NET

framework, which allows customized vb.net scripts to be prepared

by users and embedded into MODSIM without recompilation of

the MODSIM code. This allows interfacing of MODSIM with

complex operational rules, water quality models, artificial neural

networks for stream-aquifer response modeling, and geographic

xvii

information systems. Several case studies are presented

documenting ongoing use of MODSIM in the United States and

abroad.

Chapter 24 presents a Water Rights Analysis Package

(WRAP) model which simulates management of the water

resources of a river basin or multiple-basin region under a

priority-based water allocation system. The model facilitates assessment of

hydrologic and institutional water availability/reliability for water

supply, environmental instream flow, and hydropower

require-ments with specified reservoir storage and conveyance facilities,

operating practices, and institutional arrangements for managing

water resources. The WRAP model is a component of the Texas

Water Availability Modeling (WAM) System, which is routinely

applied in local, regional, and statewide planning and

administration of the statewide water rights permit system.

Chapter 25 discusses a Hydrological River Basin

Environment Assessment Model (Hydro-BEAM) which is a

mesh-typed multilayer runoff model for performing the environment

assessment with GIS technology representing the temporal and

spatial distributions. The last chapter, 26, deals with the State of

Colorado’s Stream Simulation Model (StateMod), which is a

monthly or daily water allocation and accounting model capable of

making comparative analyses for the assessment of various historic

and future water management policies in a river basin. It is

designed for application to any river basin.

This book will be of interest to those who are engaged in

the practice of hydrology, civil engineering, agricultural

engineering, environmental science, forest and range science,

climatology, or watershed science. Professors who are engaged in

graduate instruction and research as well as graduate students in

these areas will find this book to be useful. This book will be of

special appeal to hydrologic modelers and model users.

V. P. Singh

Baton Rouge, Louisiana

D. K. Frevert

xix

ACKNOWLEDGMENTS

The editors express their deep appreciation to all the

contributors who enthusiastically embraced the idea of this book

and who spent much time, effort, and resources to write their

contributions. The book reflects their collective effort and

dedication. There are thousands of people who have long been

engaged in developing and applying mathematical models of

watershed hydrology. Without their efforts we would not have the

models that we take for granted these days. We all owe a debt of

gratitude to all these people for their unselfish work and devotion

to advance the science and practice of hydrologic modeling. This

book is dedicated to all of these people. Mr. Hemant Chowdhary of

the Department of Civil and Environmental Engineering at

Louisiana State University played a key role in bringing the book

to a presentable form. Without his support, it would have been

difficult to complete the book in time. His help is gratefully

acknowledged. Finally, the editors’ families are acknowledged for

their unwavering support of this project.

xxi

Contributors

M. Arabi, Ph.D.

School of Civil Engineering Purdue University

West Lafayette, Indiana

Mustafa M. Aral, Ph.D.

School of Civil & Environmental Eng. Georgia Institute of Technology Atlanta, Georgia

Ray Alvarado

Colorado Water Conservation Board Denver, Colorado

Bill J. Barfield, Ph.D.

Biosystems & Agricultural Eng. Oklahoma State University Stillwater, Oklahoma

Ray Bennett

Colorado Division of Water Resource Denver, Colorado

Michael B. Butts

DHI Water & Environment Hørsholm, Denmark

Steven C. Chapra, Ph.D.

Dept. of Civil & Environmental Eng. Tufts University

Medford, Massachusetts

Hwai-Ping Cheng, Ph.D.

Engineering Research and Dev. Center U.S. Army Corps of Engineers Vicksburg, Mississippi

Luke D. Connell, Ph.D.

CSIRO Petroleum Resources Victoria, Australia

Jerry L. Cotter, Ph.D.

Water Management Division U.S. Army Corps of Engineers Ft. Worth District, Texas

Robert E. Dickinson

Camp Dresser & McKee Tampa, Florida

Anthony S. Donigian, Jr.

AQUA TERRA Consultants Mountain View, California

Charles W. Downer, Ph.D.

So. Florida Natural Resources Center Homestead, Florida

Paul B. Duda

AQUA TERRA Consultants Decatur, Georgia

Earl Edris, Ph.D.

Engineering Research and Dev. Center U.S. Army Corps of Engineers Vicksburg, Mississippi

A. Fogle

Kentucky Geological Survey Lexington, Kentucky

Donald K. Frevert, Ph.D.

U.S. Bureau of Reclamation Denver, Colorado

Terrance Fulp, Ph.D.

U.S. Bureau of Reclamation Boulder City, Nevada

J. Gillespie

Soil and Water Conservation District Greenville, South Carolina

xxii

Contributors

M. Gilfedder, Ph.D.

CSIRO Land and Water Queensland, Austrailia

H. Morgan Goranflo, Ph.D.

River Operations

Tennessee Valley Authority Knoxville, Tennessee

Rao S. Govindaraju, Ph.D.

School of Civil Engineering Purdue University

West Lafayette, Indiana

Douglas N. Graham

DHI Water & Environment Hørsholm, Denmark

Orhan Gunduz, Ph.D.

Dept. of Environmental Eng. Dokuz Eylul University Izmir, Turkey

S.L. Harp, Ph.D.

Biosystems & Agricultural Eng. Oklahoma State University Stillwater, Oklahoma

J.C. Hayes, Ph.D.

Agricultural and Biological Eng. Clemson University

Clemson, South Carolina

K.F. Holbrook

Woolpert

Charlotte, North Carolina

Guobiao Huang, Ph.D.

Dept. of Civil & Environmental Eng. Pennsylvania State University University Park, Pennsylvania

Wayne C. Huber, Ph.D.

Department of Civil Engineering Oregon State University

Corvallis, Oregon

John Imhoff

AQUA TERRA Consultants Ouray, Colorado

R. C. Izaurralde, Ph.D.

Joint Global Change Research Inst. College Park, Maryland

Russell Kinerson, Ph.D.

U. S. Environ. Protection Agency Washington, D.C.

Jack L. Kittle, Jr.

AQUA TERRA Consultants Decatur, Georgia

J.K. Koelliker, Ph.D.

Biological & Agricultural Eng. Kansas State University, Manhattan, Kansas

Toshiharu Kojiri, Ph.D.

Water Resources Research Center DPRI, Kyoto University

Kyoto, Japan

John W. Labadie, Ph.D.

Department of Civil Engineering Colorado State University Fort Collins, Colorado

Venkat Lakshmi, Ph.D.

Department of Geological Sciences University of South Carolina Columbia, South Carolina

xxiii

Contributors

George H. Leavesley, Ph.D.

USGS, Water Resources Division Denver, Colorado

James F. Limbrunner, Ph.D.

Department of Civil & Environ. Eng. Tufts University

Medford, Massachusetts

Hsin-Chi-Lin, Ph.D.

Engineering Research and Dev. Center U.S. Army Corps of Engineers Vicksburg, Mississippi

Siquing Liu

So. Florida Natural Resources Center Homestead, Florida

Tim Magee, Ph.D.

University of Colorado Boulder, Colorado

Steve L. Markstrom, Ph.D.

USGS, Water Resources Division Denver, Colorado

Russell Mein, Ph.D.

Department of Civil Engineering Monash University

Victoria, Australia

Justin Neidzialek

So. Florida Natural Resources Center Homestead , Florida

João Pedro Nunes, Ph.D.

Faculty of Science and Technology New University of Lisbon

Caparica, Portugal

Fred L. Ogden, Ph.D.

Dept. of Civil & Environmental Eng. University of Connecticut

Storrs, Connecticut

T. Pagano, Ph.D.

Natural Resources Conserv. Services U.S. Department of Agriculture Portland, Oregon

Liliang Ren, Ph.D.

College of Water Resources and Env. Hohai University

Nanjing, P.R. China

David Richards, Ph.D.

Engineering Research and Dev. Center U.S. Army Corps of Engineers Vicksburg, Mississippi

Lewis A. Rossman, Ph.D.

Water Supply and Water Res. Div. U.S. Environ. Protection Agency Cincinnati, Ohio

R. P. Rudra, Ph.D.

Water Resources Engineering University of Guelph

Ontario, Canada

Keith E. Saxton, Ph.D.

(Retired) U.S. Dept. of Agriculture Pullman, Washington

J. Seixas, Ph.D.

New University of Lisbon Caparica, Portugal

xxiv

Contributors

Vijay P. Singh, Ph.D., D.Sc.

Dept. of Civil & Environmental Eng. Louisiana State University

Baton Rouge, Louisiana

Rajagopalan Srinivasan

Department of Geological Sciences University of South Carolina Columbia, South Carolina

M. Sophocleous, Ph.D.

Kansas Geological Survey Lawrence, Kansas

E. Stevens, Ph.D.

Biosystems & Agricultural Eng. Oklahoma State University Stillwater, Oklahoma

G. Nuno Vieira, Ph.D.

Faculty of Sciences and Technology New University of Lisbon

Caparica, Portugal

Roland J. Viger, Ph.D.

USGS, Water Resources Division Denver, Colorado

Richard M. Vogel, Ph.D.

Dept. of Civil & Environmental Eng. Tufts University,

Medford, Massachusetts

Jimmy R. Williams, Ph.D.

Texas Agricultural Experiment Station

Temple, Texas

Patrick H. Willey, Ph.D.

Natural Resources Conserv. Services U.S. Department of Agriculture Portland, Oregon

Ralph A. Wurbs, Ph.D.

Department of Civil Engineering Texas A&M University

College Station, Texas

Gour-Tsyh Yeh, Ph.D.

Dept. of Civil & Environmental Eng. University of Central Florida Orlando, Florida

Fei Yuan, Ph.D.

College of Water Resources and Env. Hohai University Nanjing, P.R. China

Edith A. Zagona, Ph.D.

Fan Zhang, Ph.D.

Dept. of Civil & Environmental Eng. University of Central Florida Orlando, Florida

3

CHAPTER 1

Introduction

Vijay P. Singh and Donald K. Frevert

Watershed models simulate natural processes of the flow of water, sediment, chemicals, nutrients, and microbial organisms within watersheds, as well as quantify the impact of human activities on these processes. Simulation of these processes plays a fundamental role in addressing a range of water resources, environmental, and social problems. The current generation of watershed models is quite diverse and varies significantly in sophistication and data and computational requirements. Newly emerging technologies are being increasingly integrated into watershed models. This chapter introduces some of these technologies as well as the theme of the book.

1. APPLICATION OF WATERSHED MODELS

Today it is difficult to think of an environmental or a water resources problem whose solution does not involve application of some kind of a watershed model. Indeed watershed models have become a main tool in addressing a wide spectrum of environmental and water resources problems, including water resources planning, development, design, operation, and management. Flooding; droughts; upland erosion; streambank erosion; coastal erosion; sedimentation; nonpoint source pollution; water pollution from industrial, domestic, agricultural, and energy industry sources; migration of microbes; salinity and alkalinity of soils; deterioration of lakes; acid precipitation; disappearance of beaches; desertification of land; degradation of land; decay of rivers; irrigation of agricultural lands; proper management of water resources; conjunctive use of surface and groundwater; reliable design of hydraulic structures; and justifying the need for river training works are some of the critical environmental problems which are solved using watershed models. These models are also employed in military operations. For example, the U.S. Department of Defense (DOD) employs watershed simulation to support military as well as civilian operations, in environmental management of approximately 200,000 km2 _{of land on military installations in the U.S., and}

flood control and river improvement (Downer and Ogden, 2004). Watershed models help understand dynamic interactions between climate and land surface hydrology. For example, vegetation, snow cover, and the permafrost active layer are some of the features which are quite sensitive to the lower boundary of the atmospheric system. The water and heat transfer between the land surface and atmosphere significantly influence hydrologic characteristics and yield, in turn, lower boundary conditions for climate modeling (Kavvas et al., 1998).

4 Watershed Models

Assessment of the impact of climate change on national water resources and agricultural productivity is made possible by the use of watershed models. Water allocation requires integration of watershed models with models of physical habitat, biological populations, and economic response. Estimating the value of instream water use allows recreational, ecological, and biological concerns to compete with traditional consumptive uses, such as agriculture, hydropower, municipality, and industry (Hickey and Diaz, 1999). Watershed models are utilized to quantify the impacts of watershed management strategies linking human activities within the watershed to water quantity and quality of the receiving stream or lake (Mankin et al., 1999; Rudra et al., 1999) for environmental and water resources protection.

2. INVENTORY OF WATERSHED MODELS

In 1991, the Bureau of Reclamation prepared an inventory of 64 watershed hydrology models classified into 4 categories, and the inventory has been updated over the past several years. Burton (1993) compiled the Proceedings of the Federal Interagency Workshop on Hydrologic Modeling Demands for the 1990s, which contains several important watershed hydrology models. Singh (1995b) edited a book that summarized 26 popular models from around the globe. The Subcommittee on Hydrology of the Interagency Advisory Committee on Water Data (1998) published proceedings of the First Federal Interagency Hydrologic Modeling Conference which contains many popular watershed hydrology models developed by federal agencies in the United States. Wurbs (1998) listed a number of generalized water resources simulation models in seven categories and discussed their dissemination. Singh and Frevert (2002a, b) edited two books that contain 38 models. There are still some popular models which have not yet been presented under one cover, and that constitutes the rationale for preparing this book.

3. DEVELOPMENT OF WATERSHED MODELS

The digital revolution started with the advent of computers in the 1960s. The power of computers has since increased exponentially. The digital revolution also triggered two other revolutions, namely, numerical simulation and statistical simulation. As a result, advances in watershed models have occurred at an unprecedented pace since the groundbreaking development of the Stanford Watershed Model (SWM) by Crawford and Linsley in 1966. SWM was the first attempt to model virtually the entire hydrologic cycle. During the decades of the 1970s and the 1980s, a number of mathematical models were developed. Indeed there has been a proliferation of watershed hydrology models since, with growing emphasis on physically based models. Examples of such watershed hydrology models are Storm Water Management Model (SWMM) (Metcalf and Eddy, Inc., 1971), Precipitation-Runoff Modeling System (PRMS) (Leavesley et al., 1983), National Weather Service (NWS) River Forecast System (Burnash et al., 1973), Streamflow Synthesis and Reservoir Regulation

Introduction 5

(SSARR) (Rockwood, 1982), Systeme Hydrologique Europeen (SHE) (Abbott et al., 1986a, b), TOPMODEL (Beven and Kirkby, 1979), Institute of Hydrology Distributed Model (IHDM) (Morris, 1980), and others. All of these models have since been significantly improved. SWM, now called Hydrological Simulation Program-Fortran (HSPF), is far more comprehensive than its original version. SHE has been extended to include sediment transport and is applicable at the scale of a river basin (Bathurst et al., 1995). TOPMODEL has been extended to contain increased catchment information, more physically based processes, and improved parameter estimation. Development of new models or improvement of the previously developed models continues today. Today, many federal agencies in the United States have their own models or some variants of models developed elsewhere. Singh and Frevert (2002c) traced the evolution of watershed models before and during the computer era.

3.1 Currently Used Watershed Models

There are several well-known general watershed models currently in use in the U.S. and elsewhere. These models vary significantly in the model construct of each individual component process partly because these models serve somewhat different purposes. The Hydrologic Engineering Center’s Hydrologic Modeling System (HEC-HMS) is considered the standard model in the private sector in the U.S. for design of drainage systems, quantifying the effect of land use change on flooding, etc. The National Weather Service (NWS) model is the standard model for flood forecasting. HSPF and its extended water quality model are the standard models adopted by the Environmental Protection Agency. The Modular Modeling System (MMS) model of the USGS. is a widely used model for water resources planning and management works, including a number of those under the purview of the U.S. Bureau of Reclamation. The University of British Columbia (UBC) and distributed hydrologic model (WATFLOOD) are popular in Canada for hydrologic simulation. The runoff routing model (RORB) and WBN models are commonly employed for flood forecasting, drainage design, and evaluating the effect of land use change in Australia. TOPMODEL and SHE are the standard models for hydrologic analysis in many European countries. The HBV model is the standard model for flow forecasting in Scandinavian countries. The ARNO, LCS, and TOPIKAPI models are popular in Italy. The Tank models are well accepted in Japan. The Xin’anjiang model is a commonly used model in China.

3.2 Comparison of Watershed Models

The World Meteorological Organization (WMO) sponsored three studies on intercomparison of watershed hydrology models. The first study (WMO, 1975) dealt with conceptual models used in hydrologic forecasting. The second study (WMO, 1986) dealt with an intercomparison of models used for simulation of flow rates, including snowmelt. The third study (WMO, 1992) dealt with models for forecasting streamflow in real time. Except for the WMO reports, no comprehensive effort has been made to compare most major watershed hydrology models. However, efforts have been made to compare models of

6 Watershed Models

some component processes. Also, developers of models have usually compared their models with one or more other models.

3.3 Strengths

There exists a multitude of watershed models, and their diversity is so large that one can easily find more than one watershed model for addressing any practical problem. Thus diversity is one the major strengths of the current generation of models. Many of these models are quite comprehensive in that they can be applied to a range of problems. In many cases models mimic reasonably well the physics of the underlying hydrologic processes in space and time. They are also distributed in space and time. Several of the models attempt to integrate ecosystems and ecology, environmental components, biosystems, geochemistry, atmospheric sciences, and coastal processes with hydrology. This reflects the increasing role of watershed models in tackling environmental and ecosystems problems.

3.4 Deficiencies

Although watershed models have become increasingly more sophisticated, there is a long way to go before they become “household” tools. The most ubiquitous deficiencies of the models are their lack of user-friendliness, large data requirements, lack of quantitative measures of their reliability, clear statement of their limitations, and clear guidance as to the conditions for their applicability. Also, some of the models cannot be embedded with social, political, and environmental systems.

4. DATA SYSTEMS

The data needed for watershed models are hydrometeorologic, geomorphologic, agricultural, pedologic, geologic, hydraulic, and hydrologic. Hydrometeorologic data include rainfall, snowfall, temperature, radiation, humidity, vapor pressure, sunshine hours, wind velocity, and pan evaporation. Agricultural data include vegetative cover, land use, treatment, and fertilizer application. Pedologic data include soil type, texture, and structure; soil condition; soil particle size diameter; porosity; moisture content and capillary pressure; steady-state infiltration, saturated hydraulic conductivity, and antecedent moisture content. Geologic data include data on stratigraphy, lithology, and structural controls, depth, and areal extent of aquifers. For confined aquifers, hydraulic conductivity, transmissivity, storativity, compressibility, and porosity are needed. For unconfined aquifers, data on specific yield, specific storage, hydraulic conductivity, porosity, water table, and recharge are needed. Geomorphologic data include topographic maps showing elevation contours, river networks, drainage areas, slopes and slope lengths, and watershed area. Hydraulic data include roughness, flow stage, river cross-sections, and river morphology. Hydrologic data include flow depth, streamflow discharge, base flow, interflow, stream-aquifer interaction, potential, water table, and drawdown. Each data set is examined with respect to homogeneity,

Introduction 7

completeness, errors, and accuracy. Storage, handling, retrieval, processing, management, analysis, and manipulation of data are other important issues in data processing. Observed data frequently correspond to different space and time scales. Booij (2003) discussed determination and integration of appropriate spatial scales for river basin modeling.

4.1 Remote Sensing and Space Technology

Remote sensing and radar and satellite technology are being increasingly utilized in obtaining synoptic data regarding spatial distribution of meteorological inputs, soil and land use parameters, and initial conditions; inventories of water bodies, such as dams, lakes, swamps, flooded areas, and rivers; mapping of snow and ice conditions; and water quality parameters (Engman and Gurney, 1991). Digital imagery provides mapping of spatially varying landscape attributes. The Landsat Thematic Mapper (TM), Multispectral Scanner (MSS), or Systeme Probatoire d’ la Terre (SPOT) produce satellite imagery which, in conjunction with aerial photos and terrain data, provide data for mapping and classification of land use, and vegetative land cover. The airborne Light Detection and Ranging (LIDAR) technology is providing accurate real-time flood inundation maps.

The Next Generation Weather Radar (NEXRAD), Weather Surveillance Radars-88 Doppler (WSR-88), among others, are being employed to near real-time high-resolution precipitation volume and intensity over space and real-time. The Soil (now Natural Resources) Conservation Service collects real-time data on snowpacks from a network of about 500 snowpack telemetry sites (SNOTEL) located in remote mountainous areas of the western U.S. These point measurements are augmented by satellite remote sensing to provide spatial and temporal distribution of snowpack properties. The National Operational Hydrologic Remote Sensing Center of the National Weather Service provides data on real-time snow water equivalent for river basins in more than 25 states through its airborne gamma radiation measurements, and maps areal extent of snow cover for more than 4000 river basins nationwide through satellite data from the Advanced Very High Resolution Radiometer (AVHRR) and Geostationary Operational Environmental Satellite (GOES).

With the vastly improved capability to observe hydrologic data, remote sensing and space technology are being increasingly coupled with watershed models for real-time flood forecasting, weather forecasting, forecasting of seasonal and/or short-term snowmelt runoff, evolution of watershed management strategies for conservation planning, development of reporting services for drought assessment and forecasting, mapping of groundwater potential to support the conjunctive use of surface water and groundwater, inventorying of coastal and marine processes, environmental impact assessment of large-scale water resource projects, flood-damage assessment, and development of a remote information matrix for irrigation development, to name but a few. Walker et al. (2003) argue that these hydrologic observation tools require concurrent advances in hydrologic assimilation in order for the vast amounts of data to be useful for hydrologic models.

8 Watershed Models

4.2 Digital Terrain and Elevation Models

Digital mapping represents the three-dimensional nature of natural landscapes. Digital terrain (DTM) or digital elevation (DEM) models automatically extract topographic variables, such as basin geometry, stream networks, slope, aspect, flow direction, and so on from raster elevation data. Three schemes for structuring elevation data for DEMs are triangulated irregular networks (TIN), grid networks, and vector or contour-based networks (Moore and Grayson, 1991). The most widely used data structures are grid networks. Although most efficient, Mark (1978) remarked that grid structures for spatially dividing watersheds are not appropriate for many hydrologic and geomorphologic applications. The grid and vector networks are useful for planning purposes. Hydrologic models with a spatial structure are being increasingly based on DEMs or DTMs (Moore et al., 1988).

4.3 Chemical Tracers

Tracers can provide a wealth of information on the flow of water, its origin, source, flow paths, etc. Stable isotopes have been used for defining conceptual models of water flow (Stewart and MacDonnell, 1991). Radiogenic isotopes, both natural and anthropogenic, have been used as tracers (Rose, 1992). Chlorofluorocarbons have been employed to trace flow paths in groundwater systems (Dunkle et al., 1993).

4.4 GIS and DBMS

Geographical information systems (GIS), data base management systems (DBMS), and graphic and visual design tools are employed for processing of large quantities of data (Singh and Fiorentino, 1996). These are being integrated with watershed hydrology models for designing, calibrating, modifying, evaluating, and comparing watershed hydrology models. The use of GIS permits subdividing a watershed into hydrologically homogeneous subareas in both horizontal and vertical domains. With GIS, it is possible to delineate soil loss rates, identify potential areas of nonpoint source agricultural pollution, and map groundwater contamination susceptibility. GIS enhances the ability to incorporate spatial details and with much better resolution of terrain, streams, and drainage areas, and the ability to delineate more appropriate grid layers for a finite-element or finite-difference watershed model is enhanced. Vieux (1991, 2004) discussed several aspects of the use of GIS in watershed modeling.

4.5 Spatial Description of Topography

The various methods of simplifying watershed geometry can be divided into (1) grid methods and (2) conceptual methods (Singh, 1996). Either method subdivides the watershed into subareas that are linked together by routing elements. A grid method attempts to maintain model flow patterns similar to those in the prototype watershed response. This concept was introduced by Bernard in 1937. These days, different types of grid structures, such as the finite-element grid, rectangular grid, boundary-fitted coordinate grid, etc. are used, depending on the numerical scheme of a model.

Introduction 9

Conceptual methods represent watershed geometry using a network of elemental sections, including plane, triangular section, converging section, diverging section, and channel. Each element represents a particular portion of the watershed. These elements may be arranged to provide a detailed representation of the gross topographic features of a watershed, regardless of its geometric complexity. Lane and Woolhiser (1977) suggested a statistical procedure to select an appropriate geometric simplification of a watershed.

5. MODELING TECHNOLOGIES 5.1 Artificial Neural Network (ANN)

ANNs have an ability to capture a relationship from given patterns, and this makes them suitable for employment in the solution of large-scale complex problems, such as pattern recognition, nonlinear modeling, classification, association, and control. Because ANNs have the ability to recursively learn from data and can result in significant savings in time required for model development, they are particularly suited for modeling nonlinear systems where traditional parameter estimation techniques are not convenient. Preliminary concepts and hydrologic applications of ANNs have been detailed by American Society of Civil Engineers (ASCE) (2000a, b). The book edited by Govindaraju and Rao (2000) contains a variety of applications of ANNs to hydrologic modeling.

In hydrologic applications, a three-layer feed-forward type of artificial neural network is commonly considered. In a feed-forward network, the input quantities are fed into input layer neurons, which, in turn, pass them on to the hidden-layer neurons after multiplication by a weight. A hidden-layer neuron adds up the weighted input received from each input neuron, associates it with a bias, and then passes the result on through a nonlinear transfer function. The output neurons do the same operation as does a hidden neuron. The back-propagation algorithm finds the optimal weights by minimizing a predetermined error function (E). The network learns by adjusting the biases and weights that link its neurons. Before training could begin, a network’s weights and biases are set equal to small random values. Also, due to the nature of the sigmoid function used in the back-propagation algorithm, all external input and output values are standardized before passing them into a neural network. Without standardization, large values input into an ANN would require extremely small weighting factors to be applied, and this could cause a number of problems (Dawson and Wilby, 1998).

5.2 Fuzzy Logic (FL)

A general fuzzy system has basically four components—fuzzification, fuzzy rule base, fuzzy output engine, and defuzzification. Fuzzification converts each piece of input data to degrees of membership by a look-up in one or more of several membership functions. Intuition, inference, rank ordering, angular fuzzy sets, neural networks, genetic algorithms, and inductive reasoning are among many ways to assign membership values or functions to fuzzy variables. The

10 Watershed Models

fuzzy rule base contains rules that include all possible fuzzy relations between inputs and outputs. These rules are expressed in the IF-THEN format. All the uncertainties, nonlinear relationships, and model complications are included in the descriptive fuzzy inference procedure in the form of IF-THEN statements. The fuzzy inference engine takes into account all the fuzzy rules in the fuzzy rule base and learns how to transform a set of inputs to corresponding outputs. Defuzzification converts the resulting fuzzy outputs from the fuzzy inference engine to a number. There are many defuzzification methods, such as center of gravity (COG) (centroid), bisector of area (BOA), mean of maxima (MOM), leftmost maximum (LM), and rightmost maximum (RM), etc. (Jantzen 1999; Sen 1999). The details of the FL algorithm are available in the literature (McNeill and Thro 1994; Jantzen 1999; Sen 1998, 1999; and Tayfur et al. 2003).

5.3 Genetic Algorithms

Genetic algorithms are search techniques employing the mechanics of natural selection and genetics. The search algorithm is formed by combining the concept of survival of the fittest among string structures having a structured yet randomized information exchange with some of the innovative flair for human search (Goldberg, 1989). The genetic algorithms differ from traditional search techniques in many ways (Buckles and Petry, 1994) and have attractive advantages. Simpson et al. (1994) compared genetic algorithms to other techniques for pipe optimization. Wang (1991) developed a genetic algorithm for calibrating conceptual rainfall-runoff models. Srivastava et al. (2002) employed a genetic algorithm for watershed optimization of best management practices. Savic et al. (1999) developed a genetic programming approach to structured system identification for rainfall-runoff modeling. Sen and Oztopal (2001) presented algorithms for the classification and prediction of precipitation occurrence. Tang and Mays (1998) employed genetic algorithms for optimal operation of soil aquifer treatment systems. Gentry et al. (2003) discussed the efficacy of genetic algorithms for investigating small-scale aquitard leakage.

6. MODEL CALIBRATION AND VALIDATION

Recent advances in automated watershed model calibration have focused on four main issues: (1) development of specialized techniques for handling errors present in data, (2) search for a reliable parameter estimation algorithm, (3) determination of an appropriate quantity of information-rich data, and (4) efficient representation of the uncertainty of the calibrated model (structure and parameters) and translation of uncertainty into the model response. Critical issues pertaining to calibration data are the amount of data necessary and sufficient for calibration and the quality of data resulting in the best parameter estimates. However, our understanding to address such issues is less than complete. To account for data errors, maximum likelihood functions have been developed for measuring the closeness of the model and the data.

Optimization methods have been developed for parameter estimation. A typical automatic parameter estimation methodology requires four elements: (1)

Introduction 11

objective function, (2) optimization algorithm, (3) termination criteria, and (4) calibration data. The choice of an objective function influences parameter estimates as well as the quality of model results. Rao and Han (1987) analyzed several objective functions in calibrating the urban watershed runoff model. Diskin and Simon (1979) proposed guidelines and made recommendations for selecting an objective function in model calibration.

Sorooshian and Gupta (1995) discussed several optimization methods, including direct search methods, gradient search methods, random search methods, multistart algorithms, and shuffled complex algorithms. The first two are local search methods and the remaining are global search methods. The shuffled complex evolution (SCE-UA) global optimization algorithm has, however, been found to be consistent, effective, and efficient in locating the globally optimum hydrologic model parameters.

Termination criteria are needed in an iterative search algorithm to determine when the slope of the function response surface is zero and the function value is minimum. Criteria include the function convergence, parameter convergence, and maximum iterations and their limitations. Proper choice of calibration data may mitigate difficulties encountered in model calibration.

Gupta et al. (1999) discussed the status of automatic calibration for hydrologic models. They presented a global optimization algorithm and compared it with multilevel expert calibration. Their analysis suggests that simple split sample testing of model performance is not adequate and more robust model evaluation criteria are needed.

Watershed models are verified using a split-sample approach, Monte Carlo simulation, assessment of model uncertainty, and propagation of errors. Wagener et al. (2002) presented a toolkit for development and testing of hydrologic models. Eight objective functions were discussed. The toolkit permits quick implementation and evaluation of model structures to identify the most suitable one for the task under consideration. Recognizing the model uncertainty and critically analyzing the limitations of the existing model testing methods, Kuczera and Franks (2002) developed a probabilistic framework for model testing, including a Bayesian paradigm, articulation of errors, and data augmentation strategies. What is usually not done is to assess the validity of the model for a range of conditions or for a variety of data sets and to delineate limitations for which the model is valid.

Madsen et al. (2002) compared three different methods for calibration of rainfall-runoff models. The methods employed various calibration strategies utilizing multiple objectives and permitting user intervention on different levels and different stages during calibration. Vogel and Sankarasubramanian (2003) validated a model without calibration. They argued that traditional approaches of model validation based on goodness of fit between model predictions and observations might lead to misleading results. They proposed an approach based on the evaluation of the ability of a model to represent the observed covariance structure of the model input and output. Loechle and Ice (2002) reviewed criteria for evaluating watershed models. Various goodness of fit statistics typically used to evaluate the performance of a model are valid only when a

12 Watershed Models

single output variable is tested. They then proposed a Pareto set approach for calibration and evaluation of multicriteria models.

7. EXPERT SYSTEMS

Although the area of artificial intelligence is very appealing, it somehow has not attracted much attention in the hydrologic community. Gashing et al. (1981) probably were the first to develop a knowledge-based expert system for water resource problems. Underlying this system was SWM/HSPF. Simanovic (1990) described an expert system for selection of a suitable method for flow measurement in open channels.

8. MODEL RELIABILITY

When a model is used outside of the conditions for which it was calibrated and verified, the question often arises: How reliable is the model output? Melching (1995) provided a comprehensive review of reliability estimation of watershed models. Central to determining the overall model reliability is the determination of uncertainty inherent in modeling. There are four types of uncertainty: (1) natural randomness, (2) data, (3) model parameters, and (4) model structure. The model output reliability is a function of these uncertainties. Monte Carlo simulation, Latin hypercube simulation, mean value first order second moment method, advanced first order second moment method, Rosenblueth’s point estimation method, and Harr’s point estimation method are some of the popular estimation methods (Melching, 1995; Singh, 2004).

9. EMBEDDING OF WATERSHED MODELS

With growing technologies triggered by the information revolution, remote sensing, satellite technology, geographic information systems, visual graphics, and data base management, hydrologic models are getting increasingly more sophisticated and are being integrated with environmental and ecological management and other process models.

The future of watershed hydrology models will be shaped by increasing societal demand for integrated environmental management; growing need for globalization by incorporation of geological, biological, chemical, and physical aspects of the hydrological cycle; assessment of the impact of climate change; rapid advances in remote sensing and satellite technology, GIS, DBMS, and expert systems; enhanced role of models in planning and decision making; mounting pressure on transformation of models to user-friendly forms; and clearer statement of reliability and risk associated with model results.

Application of watershed hydrology models to environmental management will grow in the future. The models will be required to be practical tools— readily usable in planning and decision making. They will have to be interfaced with economic, social, political, administrative, and judicial models. Thus,

Introduction 13

watershed models will become a component in the larger management strategy. Furthermore, these models will become more global, not only in the sense of spatial scale but also in the sense of hydrologic details. Increasing fusion of biological and chemical courses in undergraduate curricula emphasizing hydrology is a healthy sign in that direction, and will help achieve this goal.

10. FUTURE OF WATERSHED MODELS

Watershed hydrology models will have to embrace rapid advances occurring in remote sensing and satellite technology, geographical information systems, data base management systems, error analysis, risk and reliability analysis, and expert systems. With use of remote sensing, radar, and satellite technology, our ability to observe data over large and inaccessible areas and to map these areas spatially is vastly improved, making it possible to develop truly distributed models for both gauged and ungauged watersheds. Distributed models require large quantities of data which can be stored, retrieved, managed, and manipulated with use of GIS and DBMS. This is possible because of the literally unlimited computing capability available these days and will be even more so in the future. If watershed hydrology models are to become practical tools, then they will have to be relatively easy to use with a clear statement as to what they can and cannot do. They will need to assess the errors and determine how they propagate, define the reliability with which they accomplish their intended functions, and require the user to possess only a minimal amount of hydrologic training. Furthermore, the models will have to “learn” from the user as well as from empirical experience. Many of these functions can be performed by the use of expert systems in watershed hydrology modeling. Usually, the user is interested in what a model yields and its accuracy, and how easy it is to use, not the biology, chemistry, physics, geology, and hydrology it is based on.

The models will have to be described in simple terms such that the interpretation of their results would not tax the ability of the user. They must be designed to serve a practical end, and their constituency is one of users. After all, hydrologic models are to be used, not to be confined to academic shelves. Thus, model building will have to gravitate around the central theme of their eventual practical use in integrated environmental management. Although much progress has been made in mathematical watershed hydrology models, there is still a long way to go before the models will be able to fully integrate rapidly evolving advances in information, computer, and space technology, and become “household” tools. Hydrologists are being challenged but we have no doubt that they will meet the challenge.

Although much progress has been achieved in watershed hydrology models, there is a greater road ahead. A basic question is: What modeling technology is better? Because of the confusion, the technology developed decades ago is still in use in many parts of the world. This state of affairs is partly due to the lack of consensus as to the superiority of one type of technology over another. Also, we have not been able to develop physically based models in a true sense and define

14 Watershed Models

their limitations. Thus, it is not always clear when and where to use which type of a model.

11. OBJECTIVE OF THIS BOOK

This book contains those models that have not been reported in the earlier books. Clearly, there were several worthy models that were not included in these books. The motivation for putting together this book stemmed from the desire to provide, under one cover, a comprehensive account of some of those popular mathematical models of watershed hydrology that were not included earlier. The objective of this volume is to include a variety of models that span a range of characteristics, such as representativeness, comprehensiveness, broad-based applicability, and use of modern tools. Based on these considerations, 24 models were selected for inclusion in this volume. It is hoped that these models fulfill the intended objective. Because there is a large number of models available these days, any number of combinations of the models could satisfy the intended objective equally well. Therefore, in any model selection, personal bias is unavoidable, and the models included here may reflect our personal bias. This model selection in no way implies an endorsement of the models included or a rejection of those not included. We personally have used and like very much some of the models not included here.

REFERENCES

Abbott, M. B., Bathurst, J. C., Cunge, J. A., O’Connell, P. E., and Rasmussen, J., 1986a. An introduction to the European Hydrologic System-Systeme Hydrologique Europeen, SHE, 1: History and philosophy of a physically-based, distributed modeling system. J. of

Hydrol., 87:45-59.

Abbott, M. B., Bathurst, J. C., Cunge, J. A., O’Connell, P. E., and Rasmussen, J., 1986b. An introduction to the European Hydrologic System-Systeme Hydrologique Europeen, SHE, 2: Structure of a physically-based, distributed modeling system. J. of Hydrol., 87: 61-77.

ASCE, 2000a. Artificial neural networks in hydrology. 1: Preliminary concepts. J.

Hydrol. Engrg., ASCE, 5(2):115-123.

ASCE, 2000b. Artificial neural networks in hydrology. 2: Hydrologic applications. J.

Hydrol. Engrg., ASCE, 5(2):124-137.

Bathurst, J. C., Wicks, J. M., and O’Connell, P. E., 1995. The SHE/SHESED basin scale water flow and sediment transport modeling system. Chapter 16, in Computer Models of

Watershed Hydrology, edited by V. P. Singh, Water Resources Publications, Littleton,

CO, pp. 563-594.

Bernard, M. 1937. Giving areal significance to hydrologic research on small areas. in: Headwaters control and use. Paper presented at the Upstream Engineering Conference in 1936 in Washington, D.C., U.S. Dept. of Agriculture, Soil Conservation Service, Washington, D.C.