Accounting for Water Useand Productivity

x

SWIM Papers

In an environment of growing scarcity and competition for water, increasing the produc-tivity of water lies at the heart of the CGIAR goals of increasing agricultural productivity,protecting the environment, and alleviating poverty.

TAC designated IIMI, the lead CGIAR institute for research on irrigation and watermanagement, as the convening center for the System-Wide Initiative on Water Manage-ment (SWIM). Improving water management requires dealing with a range of policy, in-stitutional, and technical issues. For many of these issues to be addressed, no single cen-ter has the range of expertise required. IIMI focuses on the management of water at thesystem or basin level while the commodity centers are concerned with water at the farmand field plot levels. IFPRI focuses on policy issues related to water. As the NARS are be-coming increasingly involved in water management issues related to crop production,there is a strong complementarity between their work and that of many of the CGIAR cen-ters that encourages strong collaborative research ties among CGIAR centers, NARS, andNGOs.

The initial publications in this series cover state-of-the-art and methodology papersthat assisted the identification of the research and methodology gaps in the priority projectareas of SWIM. The later papers will report on results of SWIM studies, including inter-sectoral water allocation in river basins, productivity of water, improved water utilizationand on-farm water use efficiency, and multiple uses of water for agriculture. The papersare published and distributed both in hard copy and electronically. They may be copiedfreely and cited with due acknowledgment.

Randolph BarkerSWIM Coordinator

ii

SWIM Paper 1

Accounting for Water Use and Productivity

David Molden

International Irrigation Management InstituteP O Box 2075, Colombo, Sri Lanka

ii

The author: David Molden is the Research Leader of the Performance Assessment Programof the International Irrigation Management Institute, Colombo, Sri Lanka.

Several ideas and critical comments were received when preparing this document. Theauthor would like to acknowledge the creative inputs and comments from HaraldFrederiksen, Andrew Keller, David Seckler, Chris Perry, Wim Bastiaanssen, R.Sakthivadivel, Daniel Renault, and Yvonne Parks.

Molden, D. 1997. Accounting for water use and productivity. SWIM Paper 1. Colombo, SriLanka: International Irrigation Management Institute.

/ water management / irrigation management / water supply / terminology / performance indexes/ water use / water allocation / productivity /

ISBN 92-9090-349 X© IIMI, 1997. All rights reserved.

Responsibility for the contents of this publication rests with the author.

iiiiii

Contents

Acronyms iv

Glossary v

Foreword vii

Abstract ix

Background 1

Objectives 2

Water Balance Approach 3

Water Accounting Definitions 4

Accounting Components at Use, Service, and Basin Levels 7

Examples of Water Accounting 10

The Water Accounting Research Agenda 13

Literature Cited 15

iv

Acronyms

CGIAR Consultative Group on International Agricultural ResearchCIAT Centro Internacional de Agricultura TropicalCIMMYT Centro Internacional de Mejoramiento de Maize y TrigoET EvapotranspirationICRAF International Council for Research in AgroforestryIIMI International Irrigation Management InstituteIRRI International Rice Research InstituteM&I Municipal and Industrial usesNARS National Agricultural Research System(s)NGOs Nongovernmental OrganizationsSGVP Standardized Gross Value of ProductionSWIM System-Wide Initiative on Water ManagementTAC Technical Advisory Committee of the CGIAR

CGIAR Centers

CIAT Centro Internacional de Agricultura TropicalCIFOR Center for International Forestry ResearchCIMMYT Centro Internacional de Mejoramiento de Maize y TrigoCIP Centro Internacional de la PapaICARDA International Center for Agricultural Research in the Dry AreasICLARM International Center for Living Aquatic Resources ManagementICRAF International Centre for Research in AgroforestryICRISAT International Crops Research Institute for the Semi-Arid TropicsIFPRI International Food Policy Research InstituteIIMI International Irrigation Management InstituteIITA International Institute of Tropical AgricultureILRI International Livestock Research InstituteIPGRI International Plant Genetic Resources InstituteIRRI International Rice Research InstituteISNAR International Service for National Agricultural ResearchWARDA West Africa Rice Development Association

v

Glossary

Available water: the amount of water available to a ser-vice or use, which is equal to the inflow less thecommitted water.

Basin or sub-basin accounting: the macro scale of wateraccounting for all or part of water basins, includ-ing several uses of water.

Closed basin: a basin where utilizable outflows arefully committed.

Committed water: the part of outflow that is reservedfor other uses.

Depleted fraction: the fraction of inflow or availablewater that is depleted by process and non-processuses. Depleted fraction can be related to gross in-flow (Depleted Fraction of Gross Inflow), net in-flow (Depleted Fraction of Net Inflow), or avail-able water (Depleted Fraction of Available Water).

Domain: the area of interest where accounting is to bedone, bounded in time and space.

Equivalent yield: a yield value for a base crop derivedfrom a mixture of crops by using local prices toconvert yields between crops.

Fully committed basin: a water basin that has been de-veloped to the extent that all water has been al-located or, in other words, all outflows are com-mitted.

Gross inflow: the total amount of inflow crossing theboundaries of the domain.

Net inflow: the gross inflow less the change in storageover the time period of interest within the do-main. Net inflow is larger than gross inflow whenwater is removed from storage.

Non-depletive uses of water: uses where benefits are de-rived from an intended use of water without de-pleting water.

Non-process depletion: depletion of water by uses otherthan the process that the diversion was intendedfor.

Open basin: a basin where uncommitted utilizable out-flows exist.

Process depletion: that amount of water diverted anddepleted to produce an intended good.

Process fraction: the ratio of process depletion total todepletion (Process Fraction of Depleted Water) oravailable water (Process Fraction of Available Wa-ter).

Productivity of water: the physical mass of productionor the economic value of production measuredagainst gross inflow, net inflow, depleted water,process depleted water, or available water.

Standardized gross value of production: a standard meansof expressing productivity in monetary terms byconverting equivalent yield of a base crop intomonetary units using world prices.

Uncommitted outflow: outflow from the domain that isin excess of requirements for downstream uses.

Use level accounting: the micro scale of water account-ing such as an irrigated field, a household, or aspecific industrial process.

Utilizable water: outflow from a domain that could beused downstream.

Water depletion: a use or removal of water from a wa-ter basin that renders it unavailable for furtheruse.

Water services level accounting: the mezzo scale of wa-ter accounting for water services such as irriga-tion services or municipal services.

vii

Foreword

Water accounting is a procedure for analyzing theuses, depletion, and productivity of water in a waterbasin context. It is a supporting methodology usefulin assessing impacts of field level agriculturalinterventions in the context of water basins, theperformance of irrigated agriculture, and allocation ofwater among users in a water basin. It is beingdeveloped as one of the activities of the System-WideInitiative on Water Management (SWIM) of theCGIAR. The purpose of the first phase of this SWIMWater Accounting Project is to develop standardizedwater accounting procedures. The specific objectivesof the project are to:

1. Develop and formalize accounting standards fortracking water depletion within water basins.

2. Develop, jointly with the major commodity cen-ters, an accounting procedure for determining thestatus of, and measuring changes in, the sustain-able output per unit of water effectively depletedby various crops.

3. Apply and test the procedures for water use anddepletion by irrigated agriculture as a componentof selected SWIM and NARS research projects.

4. Disseminate the accounting procedures to NARSoperating in both water resources planning andcrop research.

This paper is aimed at fulfilling the first twoobjectives. It is to be used as a tool to disseminateinformation about standardized water accountingprocedures both for CGIAR centers and for NARS,including those involved in managing irrigation andwater resources. IIMI will coordinate activities withIRRI, ICRAF, CIAT, and CIMMYT. A second phase ofwater accounting and other SWIM projects will usethe procedures developed here in more detailedinvestigations. Ultimately, it is intended that wateraccounting will evolve into a set of generally acceptedand standardized practices.

David Molden

ix

Abstract

This paper presents a conceptual framework for wateraccounting and provides generic terminologies andprocedures to describe the status of water resourceuse and consequences of water resources relatedactions. The framework applies to water resource useat three levels of analysis: a use level such as anirrigated field or household, a service level such as anirrigation or water supply system, and a water basinlevel that may include several uses. Water accountingterminology and performance indicators aredeveloped and presented with examples at all thethree levels. Concepts and terminologies presented

are developed to be supportive in a number ofactivities including: identification of opportunities forwater savings and increasing water productivity;developing a better understanding of present patternsof water use and impacts of interventions; improvingcommunication among professionals and communica-tion to non-water professionals; and improving therationale for allocation of water among uses. It isexpected that with further application, these wateraccounting concepts will evolve into a robust,supporting methodology for water basin analysis.

1

With growing population and limited waterresources, there is an increasing needworldwide to manage water resources bet-ter. This is especially true when all or nearlyall water resources in a basin are allocatedto various uses. Effective strategies for ob-taining more productivity while maintain-ing or improving the environment must beformulated. Wastes and nonproductive usesmust be carefully scrutinized to identify po-tential savings. Effective allocation proce-dures that minimize and help resolve con-flicts must be developed and implemented.To assist in accomplishing these tasks,improved procedures to account for waterresource use and productivity are required.

Due to vastly different types and scalesof use, communicating about water betweenprofessionals and non-water professionals isquite difficult. Policy decisions are oftentaken without a clear understanding of con-sequences on all water users. As competi-tion for a limited supply of water increases,it becomes increasingly important to clearlycommunicate about how water is beingused, and how water resource develop-

ments will affect present use patterns.As irrigation is a large consumer of

water, developments in irrigation have pro-found impacts on basin-wide water use andavailability. Yet, planning and execution ofirrigation interventions often take placewithout consideration of other uses. One ofthe main reasons for this restricted view ofirrigation workers is inadequate means todescribe how irrigation water is being used.Irrigation efficiency is the most commonlyused term to describe how well water isbeing used. But increases in irrigation effi-ciency do not always coincide with in-creases in overall basin productivity ofwater.

Irrigation within a basin context hasbeen dealt with by several researchers.Bagley (1965) noted that failure to recognizethe boundary characteristics whendescribing irrigation efficiency can lead toerroneous conclusions, and noted that waterlost due to low efficiencies is not lost to alarger system. In the field of water rights,there is often a clear distinction betweenconsumption and diversion from a

Accounting for Water Use and Productivity

David Molden

All science depends on its concepts. These are ideas which receive names.They determine the questions one asks, and the answers one gets.They are more fundamental than the theories which are stated in terms of them.

Sir G. Thompson

Background

2

hydrologic cycle (Wright 1964). Bos (1979)identified several flow paths of waterentering and leaving an irrigation project,clearly identifying water that returns to awater basin and is available fordownstream use. Willardson (1985) notedthat efficiency of a single irrigation field isof little importance to the hydrology of abasin, except when water quality isconsidered, and concluded that “basin-wideeffects of increasing irrigation efficiencymay be negative as well as positive.” Bosand Wolters (1989) pointed out that theportion of water diverted to an irrigationproject that is not consumed, is notnecessarily lost from a river basin, becausemuch of it is being reused downstream. Itwas shown that high reuse actuallyincreases overall efficiency (Wolters and Bos1990). Van Vuren (1993) listed severalconstraints on the use of irrigationefficiency and pointed out situations whenlower efficiencies are tolerable. Palacios-Velez (1994) argues that “water that is lost isnot always necessarily wasted.”

While these works recognizedweaknesses in using efficiency terms, and

scale effects in moving analysis from farmto irrigated area and to river basin, analternative means of describing waterresource use was not presented. Working atabout the same time, alternative terms wereproposed to describe use of water withinbasins by Keller and Keller (1995) witheffective efficiency, and by Willardson et al.(1994) with the use of fractions. Recently,works by Seckler (1992, 1993, 1996), Keller(1992), Keller et al. (1996), and Perry (1996b)describe many of the considerations to bedealt with in describing irrigation in waterbasins.

At this time, a common framework isrequired to describe water use in basins. Aframework and common language todescribe the use and productivity of waterresources are presented in the paper. Thiswork is developed from an irrigationperspective so that we can betterunderstand the impacts of irrigationinterventions at a water basin scale. It isdeveloped in a general manner to describeany water resource use in order to enhancecommunications between practitioners indifferent water resource fields.

Objectives

This paper presents concepts and defini-tions necessary to account for water use,depletion, and productivity. The accountingprocedures and standards given here aredesigned to be universally applicable forevaluating water management within andamong all sectors. A goal of this approach isto develop a generic, common language foraccounting for uses of water. This concep-tual framework provides:

1. the terminology and a procedure thatcan be applied to describe the present

status and consequences of water re-sources related actions carried out inagriculture and other water sectors;

2. a common means for reporting resultsof water-related agronomic trials and ir-rigation interventions so that impactscan be better understood in a water ba-sin context; and

3. examples of water accounting at threelevels to test and demonstrate the util-ity of the methodology.

3

Levels of Analysis

Researchers in agriculture, irrigation, andwater resources work with spatial scales ofgreatly different magnitudes. Agriculturalresearchers often focus on a field level or aplot level dealing with crop varieties andfarm management practices. Irrigation spe-cialists focus on a set of fields tied togetherby a common resource—water. Water re-source specialists are concerned with otheruses of water beyond agriculture, includingmunicipal, industrial, and environmentaluses.

An understanding of the interactionsamong these levels of analysis helps us tounderstand the impacts of our actions. Aperceived improvement in water use at thefarm level may improve overall productiv-ity of water in a basin, or it may reduceproductivity of downstream users. Onlywhen the intervention is placed in the con-text of a larger scale of analysis can the an-swer be known. Similarly, basin-wide stud-ies may reveal general concepts about howwater can be saved or productivity of waterincreased, but field level information on

how to achieve savings or increase waterproductivity is required. Therefore, threedifferent levels of water use are defined forwhich water accounting procedures are de-veloped:

• Macro level: basin or subbasin level cov-ering all or part of a water basin, in-cluding several uses of water

• Mezzo level: water services level, suchas irrigation or municipal water ser-vices

• Micro level: use level, such as an agricul-tural field, a household, or an environ-mental use

The water accounting methodology isdeveloped in a manner such that it can begenerically used for irrigation, municipal,industrial, environmental, or other uses ofwater. But the focus of this paper will be onirrigation services and use of water, andemphasis will be on quantities of water atthe field and irrigation service levels. In thefuture phases, concepts and examples willbe presented from multiple uses of waterand water quality.

Water Balance Approach

The water accounting methodology is basedon a water balance approach. Water bal-ances consider inflows and outflows frombasins, subbasins, and service and use lev-els such as irrigation systems or fields. Aninitial step in performing a water balance isto identify a domain of interest by specify-ing spatial and temporal boundaries of thedomain. For example, a domain could be anirrigation system bounded by its headworksand command area, and bounded in timefor a particular growing season. Conserva-

tion of mass requires that for the domainover the time period of interest, inflows areequal to outflows plus any change of stor-age within the domain.

In a purely physical sense, flows of wa-ter are depicted by a water balance. To de-velop and use water resources for their ownneeds, humans change the water balance.Water accounting considers components ofthe water balance and classifies them ac-cording to uses and productivity of theseuses.

4

Conceptually, the water balance ap-proach is straightforward. Often though,many of the components of the water bal-ance are difficult to estimate or are notavailable. For example, groundwater in-flows and outflows to and from an area ofinterest are difficult to measure. Estimatesof actual crop consumptive use at a regionalscale are questionable. And drainage out-flows are often not measured, as more em-phasis has been placed on knowledge of in-flows to irrigation systems or municipalwater supply systems. In spite of the limita-tions, experience has shown that even agross estimate of water balances for use inwater accounting can be quite useful tomanagers, farmers, and researchers. Water

balance approaches have been successfullyused to study water use and productivity atthe basin level (for example, Owen-Joyceand Raymond 1996; and Hassan and Bhutta1996), at the irrigation service level (for ex-ample, Perry 1996b; Kijne 1996; and Helal etal. 1984), and at the field level (for example,Mishra et al. 1995; Rathore et al. 1996;Bhuyian et al. 1995; and Tuong et al. 1996).Binder et al. (1997) use a regional balancetechnique quantifying municipal, industrial,and irrigation process uses to provide anearly recognition of changes in quantity andquality of water. Often, first order estimatesprovide the basis for a more in-depth analy-sis that provides important clues on increas-ing water productivity.

Water Accounting Definitions

The art of water accounting is to classifywater balance components into water usecategories that reflect the consequences ofhuman interventions in the hydrologiccycle. Water accounting integrates waterbalance information with uses of water asvisualized in figure 1. Inflows into the do-main are classified into various use catego-ries as defined below.

Gross inflow is the total amount of waterflowing into the domain from precipitationand surface and subsurface sources.

Net inflow is the gross inflow plus anychanges in storage. If water is removedfrom storage over the time period of inter-est, net inflow is greater than gross inflow;if water is added to storage, net inflow isless than gross inflow. Net inflow water iseither depleted, or flows out of the domainof interest.

Water depletion is a use or removal of waterfrom a water basin that renders it unavail-able for further use. Water depletion is akey concept for water accounting, as it isoften the productivity and the derived ben-efits per unit of water depleted we are in-terested in. It is extremely important to dis-tinguish water depletion from water di-verted to a service or use, because not allwater diverted to a use is depleted. Water isdepleted by four generic processes, the firstthree described by Seckler (1996) and Kellerand Keller (1995). A fourth type of deple-tion occurs when water is incorporated intoa product.

The four generic processes are:

• Evaporation: water is vaporized fromsurfaces or transpired by plants

• Flows to sinks: water flows into a sea,saline groundwater, or other location

5

where it is not readily or economicallyrecovered for reuse

• Pollution: water quality gets degradedto an extent that it is unfit for certainuses

• Incorporation into a product: by a pro-cess such as incorporation of irrigationwater into plant tissues

Process depletion is that amount of water di-verted and depleted to produce an intendedgood. In industry, this includes the amountof water vaporized by cooling, or convertedinto a product. For agriculture, it is watertranspired by crops plus that amount incor-porated into plant tissues.

Non-process depletion occurs when divertedwater is depleted, but not by the process it

was intended for. For example, water di-verted for irrigation is depleted by transpi-ration (process), and by evaporation fromsoil and free water surfaces (non-process).Outflow from coastal irrigation systems andcoastal cities to the sea is considered non-process depletion. Deep percolation flows toa saline aquifer may constitute a non-pro-cess depletion if the groundwater is notreadily or economically utilizable. Non-pro-cess depletion can be further classified asbeneficial or non-beneficial. For example, avillage community may place beneficialvalue on trees that consume irrigation wa-ter. In this case, the water depletion may beconsidered beneficial, but depletion bythese trees is not the main reason why wa-ter was diverted.

Committed water is that part of outflow thatis committed to other uses. For example,

FIGURE 1.Water accounting.

6

downstream water rights or needs may re-quire that a certain amount of outflow berealized from an irrigated area. Or, watermay be committed to environmental usessuch as minimum stream flows, or outflowsto sea to maintain fisheries.

Uncommitted outflow is water that is not de-pleted, nor committed, and is thus availablefor a use within a basin or for export toother basins, but flows out due to lack ofstorage or operational measures. For ex-ample, waters flowing to a sea that is in ex-cess of requirements for fisheries, environ-mental, or other beneficial uses are uncom-mitted outflows. With additional storage,this uncommitted outflow can be trans-ferred to a process use such as irrigation orurban uses.

A closed basin1 (Seckler 1992) is one wherethere are no utilizable outflows in the dryseason. An open basin is one where uncom-mitted utilizable outflows exist.

In a fully committed basin, there are nouncommitted outflows. All inflowing wateris committed to various uses. In this case,major options for future development arereallocation among uses, or importing waterinto the basin.

Available water is the net inflow less theamount of water set aside for committed usesand represents the amount of water availablefor use at the basin, service, or use levels.Available water includes process and non-process depletion, plus uncommitted water.

Non-depletive uses of water are uses wherebenefits are derived from an intended usewithout depleting water. In certain circum-stances, hydropower can be considered anon-depletive user of water if water di-verted for another use such as irrigationpasses through a hydropower plant. Often,a major part of instream environmental ob-

jectives can be non-depletive when outflowsfrom these uses do not enter the sea.

Performance Indicators

Performance indicators for water accountingfollow depleted fraction and effective effi-ciency concepts presented by Willardson, etal. (1994) and Keller and Keller (1995). Wa-ter accounting performance indicators arepresented in the form of fractions, and interms of productivity of water.

Depleted Fraction (DF) is that part of the in-flow that is depleted by both process and non-process uses. Depleted fraction can be definedin terms of net, gross, and available water.

1. DFnet =

2. DFgross =

3. DFavailable =

Process Fraction (PF) relates process deple-tion to either total depletion or the amountof available water.

4. PFdepleted =

5. PFavailable =

The process fraction of depleted water(PFdepleted) is analogous to the effective effi-ciency concept forwarded by Keller andKeller (1995) and is particularly useful inidentifying water savings opportunitieswhen a basin is fully or near fully commit-ted. When there is no uncommitted water,process fraction of depleted water is equalto the process fraction of available water.

Productivity of Water (PW) can either be re-lated to the physical mass of production orthe economic value of produce per unit vol-

DepletionNet Inflow

DepletionGross Inflow

DepletionAvailable Water

Process DepletionTotal Depletion

Process DepletionAvailable Water

1This water accountingdefinition differs from astrict hydrologic defini-tion of a closed basinwhere outflows go onlyto internal seas, lakes, orother sinks.

7

ProductivityNet Inflow

ProductivityDepletion

ProductivityProcess Depletion

PW net inflow

DF net

PW depleted

PF net

ume of water. Productivity of water can bemeasured against gross or net inflow, de-pleted water, process depleted water, oravailable water. Productivity of water has abroader basis than water use efficiency(Viets 1962), which relates production ofmass to process depletion (transpiration orevapotranspiration for irrigated agriculture).Here it is defined in terms of net inflow, de-pleted water, and process depletion.

6. PWinflow =

7. PWdepleted =

8. PWprocess =

The following relationships exist be-tween productivity and water indicators.

9. PWdepleted =

10. PWprocess =

For an irrigated service area, these areexternal indicators of system performancerelating the output of irrigated agricultureto its main input, water. IIMI’s external in-dicators (Perry 1996a and Molden et al.,forthcoming) draw from this water account-ing list for a minimum set of indictors andinclude the productivity of water related toprocess depletion.

Accounting Components at Use, Service, and Basin Levels

Field Level

The use level of analysis for irrigation istaken at the field level with inflows andoutflows shown in table 1. This is the levelwhere crop production takes place—theprocess of irrigation. Agricultural researchat this level is often aimed at increasingproductivity per unit of land and water andconserving water. The key question is:Which water? Again, it is important to un-derstand the category of water againstwhich production is being measured, or thecategory of water that is being conserved.

At the field level, the magnitudes of thecomponents of the water balance are a func-tion of crop and cultural practices. Differentcrops, and even different varieties of crops,will transpire water at different rates. Irriga-tion techniques influence evaporative losses,and volumes of deep percolation and sur-face runoff. For example, drip irrigationminimizes these components, while surface

application induces depletion by evapora-tion. Also, the amount of water deliveredinfluences runoff and deep percolation.Other cultural practices such as bunding,mulching, and crop spacing affect theamount of water stored in the soil, and theamount of runoff and deep percolation.Water accounting procedures attempt tocapture the effects of different crop and cul-tural practices on how water is used anddepleted at the field level.

At the field level, it is sometimes im-possible, and oftentimes unnecessary toknow the fate of outflows. Only when mov-ing up to the service and basin levels canwe determine whether to classify outflowsas committed or uncommitted. By account-ing for water use at the field level, thenplacing it in the context of irrigation serviceand basin levels, it is possible to match fieldlevel interventions with requirements at theirrigation service level, or water basin level,or both.

8

TAB

LE 1

.W

ater

acc

ount

ing

com

pone

nts

at fi

eld,

ser

vice

, and

bas

in le

vels

.

Fie

ldIr

rigat

ion

serv

ice

Bas

in/s

ubba

sin

Infl

ow

*irr

igat

ion

appl

icat

ion

*su

rfac

e di

vers

ions

*pr

ecip

itatio

n*

prec

ipita

tion

*pr

ecip

itatio

n*

tran

s-ba

sin

dive

rsio

ns*

subs

urfa

ce c

ontr

ibut

ions

*su

bsur

face

sou

rces

*gr

ound

wat

er in

flow

*su

rfac

e se

epag

e flo

ws

*su

rfac

e dr

aina

ge s

ourc

es*

river

inflo

w in

to b

asin

Sto

rag

e ch

ang

e*

soil

moi

stur

e ch

ange

in a

ctiv

e ro

ot z

one

*so

il m

oist

ure

chan

ge*

soil

moi

stur

e ch

ange

*re

serv

oir

stor

age

chan

ge*

rese

rvoi

r st

orag

e ch

ange

*gr

ound

wat

er s

tora

ge c

hang

e*

grou

ndw

ater

sto

rage

cha

nge

Pro

cess

dep

leti

on

*cr

op tr

ansp

iratio

na*

crop

tran

spira

tion

*cr

op tr

ansp

iratio

n*

mun

icip

al a

nd in

dust

rial u

ses

*fis

herie

s, fo

rest

ry, a

nd o

ther

non

-cro

p de

plet

ion

*de

dica

ted

envi

ronm

enta

l wet

land

s

No

n-p

roce

ss d

eple

tio

n*

evap

orat

ion

from

soi

l sur

face

,*

evap

orat

ion

from

free

wat

er a

nd s

oil s

urfa

ces,

wee

ds,

*ev

apor

atio

n fr

om fr

ee w

ater

and

soi

l sur

face

, wee

ds,

incl

udin

g fa

llow

land

sph

reat

ophy

tes,

and

oth

er n

on-c

rop

plan

tsph

reat

ophy

tes,

and

oth

er n

on-c

rop

plan

ts*

wee

d ev

apot

rans

pira

tion

*la

tera

l or

vert

ical

flow

to s

alt s

inks

*flo

w to

sin

ks (

salin

e gr

ound

wat

er, s

eas,

oce

ans)

*flo

w to

sin

ks (

salin

e gr

ound

wat

er, s

eas,

oce

ans)

*flo

w to

sin

ks (

salin

e gr

ound

wat

er, s

eas,

oce

ans)

*ev

apor

atio

n fr

om p

onds

/pla

yas

*ev

apor

atio

n fr

om p

onds

/pla

yas

*w

ater

ren

dere

d un

usab

le d

ue to

*w

ater

ren

dere

d un

usab

le d

ue to

deg

rada

tion

of q

ualit

y*

wat

er r

ende

red

unus

able

due

to d

egra

datio

n of

qua

lity

degr

adat

ion

of q

ualit

y*

ET

from

nat

ural

veg

etat

ion

Ou

tflo

w*

deep

per

cola

tion

*in

stre

am c

omm

itmen

ts s

uch

as e

nviro

nmen

t and

fish

erie

s*

inst

ream

com

mitm

ents

suc

h as

env

ironm

ent a

nd fi

sher

ies

*se

epag

e*

dow

nstr

eam

com

mitm

ents

*do

wns

trea

m c

omm

itmen

ts*

surf

ace

runo

ff*

for

M&

I use

with

in ir

rigat

ion

serv

ice

*ou

tflow

com

mitm

ents

to m

aint

ain

envi

ronm

ent

*un

com

mitt

ed o

utflo

ws

*un

com

mitt

ed o

utflo

ws

a Cro

p e

vap

otra

nsp

iratio

n m

ay b

e co

nsid

ered

pro

cess

dep

letio

n w

hen

it is

imp

ract

ical

to s

epar

ate

evap

orat

ion

and

tran

spira

tion

com

pon

ents

, or

whe

n se

par

atio

n of

term

s d

oes

not a

dd

toth

e an

alys

is.

Not

es: M

&I =

Mun

icip

al a

nd in

dus

tria

l use

s.E

T =

Eva

pot

rans

pira

tion.

9

Irrigation Service Level

At the service level, the focus is on irriga-tion service analysis (table 1). Similar wateraccounts could be developed for municipaland industrial uses.

The boundaries for an irrigation systemtypically include groundwater underlyingthe irrigated area, whereas for the fieldlevel the boundary would be taken as thebottom of the crop root zone. Changes instorage take place in the soil, the ground-water, and surface storage. As compared tothe field level, there are more opportunitiesfor non-process depletion, such as evapora-tion from free water surfaces and phreato-phytes.

Water diverted primarily for irrigationoften provides the source of water for otheruses such as for fisheries, drinking, bathing,and industrial use. Some of this water maybe committed to these uses and not avail-able for crop transpiration. Municipal usesof irrigation service water are typically notlarge, but they may represent a significantproportion of depletion during low flowperiods and have an important impact onoperating rules. Another commitment is toensure that water is delivered to meetdownstream rights or requirements. It isvery common to have downstream irriga-tion diversions dependent on irrigation re-turn flows and water rights can be violatedwhen these outflows are not available.These outflows, whether remaining in ca-nals or flowing through drains, can be con-sidered committed uses of water. The water

available at the irrigation service level is thediversion to irrigation less the committeduses.

Basin and Subbasin Levels

At the basin level, several process uses ofwater are considered, including agricultural,municipal and industrial uses (table 1). Themajor inflow into a basin is precipitation.Other inflows could be river inflows into asubbasin, trans-basin diversions, or ground-water originating from outside the basin. Atthe outflow of a basin, it is important toconsider commitments such as water re-quired to remove salts and pollutants fromthe basin, and water required to maintainfisheries.

Through water accounting, changes inwater use patterns can be analyzed. For ex-ample, changes in watershed vegetation canhave a profound impact on basin-wide wa-ter accounts. Reducing forest cover may re-duce evaporation but induce non-utilizableor even damaging flood flows unless sur-plus storage is available. Converting a forestto agricultural use with water conservationpractices may make water available in wa-ter-deficit seasons, or drought years. Con-verting from agricultural land to native veg-etation may have the impact of reducingdownstream flows. Using water accountingto note these factors allows decision makersto start to understand the consequences oftheir actions and to indicate where more in-depth studies would be most profitable.

10

To illustrate water accounting, three ex-amples are chosen from the use, service,and basin levels. The use level and servicelevel examples are taken from the Bhakrasystem in India, and the basin level ex-ample is drawn from the Nile River inEgypt.

Field-Level Accounting Example

As a field-level example, results of model-ing trials based on field experiments(Bastiaanssen et al. 1997) carried out in theHisar and Sirsa Circles of the Bhakra com-

mand area in India are reported in a wateraccounting format (table 2). In this area, thewater duty falls short of potential crop re-quirements as water is scarce relative toland. In response, farmers typically have astrategy of deficit irrigation, or giving lesswater than the potential crop requirement,thus giving them the opportunity to irrigatemore land.

For both crops, nearly all of the irriga-tion and rainwater applied is depleted lead-ing to a depleted fraction of nearly 1.0. Theprocess fraction of depleted water is quitehigh at 0.73, due to small amounts ofevaporation. On an annual basis, there is

Examples of Water Accounting

TABLE 2.Field-level water accounts of Sirsa Circle of the Bhakra system in India: 1991 wheat-cotton rotation.a

Wheat Cotton Annual(mm) (mm) (mm)

InflowIrrigation 324 393 717Precipitation 42 206 252Subsurface 0 0 0Lateral seepage flows 0 0 0

GROSS INFLOW 366 599 969

Storage Change +14 +22 +24NET INFLOW 352 577 945

DepletionTranspiration (process) 291 401 692Evaporation (non-process) 60 175 251

TOTAL DEPLETION 351 576 943

OutflowSurface runoffDeep percolation 1 1 2

TOTAL OUTFLOW 1 1 2

PerformanceDepleted Fraction (gross) 351/366=0.96 576/599=0.96 943/969=0.97Process Fraction (depleted) 291/351=0.83 401/576=0.70 692/943=0.73Production (kg/ha) 4,000 2,380Production per unit net inflow (kg/m3) 4,000/3,520=1.14 2,380/5,770=0.41Production per unit total depletion (kg/m3)b 4,000/3,510=1.14 2,380/5,760=0.41Production per unit process depletion (kg/m3) 4,000/2,910=1.37 2,380/4,010=0.59

aTreatment: Computer simulation of typical on-farm water management practices.bDepletion includes process depletion of transpiration and non-process depletion of evaporation.

11

some rainfall between seasons, and someevaporation from fallow land.

Yields were reported as 4.0 tons perhectare for wheat and 2.38 tons per hectarefor cotton while relative transpiration (theratio of actual to potential transpiration) forwheat and cotton was, 0.78 and 0.60,respectively, lower than 1.0 as a result of thepractice of deficit irrigation used. In thiscase, transpiration and evaporation arereported separately so that productivity pertotal depletion and per process depletioncan be computed.

It is meaningful to compare values ofmass of production per unit of water di-verted or depleted, when comparing likecrops. But when different crops are com-pared, mass of output is not as meaningful.There is a clear difference between 1 kg ofstrawberries and 1 kg of rice produced percubic meter of water depleted. One approachis to convert yields into value of productionusing local prices. A second approach is touse Standardized Gross Value of Production(SGVP). SGVP is used to measure economicproductivity to allow comparisons acrossdifferent agricultural settings by using worldprices of various crops (Perry 1996a, Moldenet al., forthcoming). To calculate SGVP, yieldof a crop is converted into an equivalentyield of a predominant, traded field cropusing local prices. Then this mass of produc-tion is converted into a monetary unit usingworld prices.

Service-Level Accounting Example

Inflow sources for irrigation include sourcesthat originate from outside the irrigationsystem. Accounts for the Sirsa Circleserving 430,000 ha in the Bhakra commandarea (Boels et al. 1996) are shown in table 3.They are based on computer simulationscalibrated to local situations. Crop

TABLE 3.Service level accounts of sirsa circle of the Bhakra system in India:1977–1990.

Component value Total(mm/year) (mm/year)

Inflow

Gross Inflow 652

Surface diversions 402

Precipitation 191

Subsurface sources from

outside domain 59

Surface drainage sources from

outside domain

Storage change 98

Surface n.a.

Subsurface 98

Net Inflow 554

Depletive use

Process depletion, (ET) 533

Non-process depletion n.a.

Flows to sinks n.a.

Other evaporation n.a.

ET from non-crop vegetation n.a.

Total Depletion 533

Outflow

Total utilizable outflow 21

Surface outflow 21

Subsurface outflow n.a.

Committed water 21

Domestic use n.a.

Industrial use n.a.

Environmental use n.a.

Downstream uses 21

Uncommitted water 0

Available water 533

Indicators

Depleted fraction (gross) 533/652 = 0.82

Depleted fraction (net) 533/554 = 0.96

Depleted fraction (available) 533/533 = 1.00

Process fraction (depleted) 533/533 = 1.00

Process fraction (available) 533/533 = 1.00

Notes: ET = Evapotranspiration. n.a. = Data not available.

12

evapotranspiration was considered as theprocess depletion. No information wasavailable on non-process depletion andproductivity. It was assumed that all theoutflow was required by downstream users,and thus is shown as committed water.

Striking in this example is the relativelyhigh depleted fraction. This is in part dueto the on-farm practice of deficit irrigationillustrated in the example above. In thisexample, there is no uncommitted water, soavailable water is equal to depleted water.Because non-process depletion was notseparated from process depletion(evapotranspiration) and no other non-process depletion was identified, theprocess fraction of depleted water is 1.0.

The change of storage term stands outas 15 percent of the gross inflow and wasadded to groundwater storage over theperiod 1977 to 1990. In fact, it is estimatedthat over this area, groundwater is rising atan average rate of 0.6 m per year. For long-term sustainability, this situation of a risingwater table has to be dealt with. In general,over time periods of several years, a largepositive or negative change of storagerepresents issues of sustainabilitycorresponding to a rise in water table ormining from groundwater.

Basin-Level Accounting Example

Water accounts are shown for the Nile Riverdownstream of the High Aswan Dam inEgypt (table 4). Figures used in the accountsare based on data presented by Abu-Zeid(1992) and estimates presented by Keller(1992) for the water year 1989–1990. Manyof the figures and estimates require furtherscrutiny, and the example is presented toillustrate the use of water accounting.

The inflow is derived almost entirelyfrom releases from the High Aswan Dam

TABLE 4.Basin-level accounts of the Nile River downstream of the High AswanDam:1989–1990 agricultural year.

Component value Total(km3) (km3)

InflowGross Inflow 53.7

Surface diversions 53.2Precipitation 0Subsurface sources from

outside subbasin 0.5Surface drainage sources from

outside subbasin

Storage change 0Surface 0Subsurface 0

Net Inflow 53.7

Depletive useProcess depletion 36.4

Evapotranspiration 34.8M&I 1.6

Non-process depletionFlows to sinks n.a. 3.2Other evaporation 3.2(phreatophytes, free water surface)

Total Depletion 39.6

OutflowTotal Outflow 14.1

Surface outflow from rivers 1.8Surface outflow from drains 12.3Subsurface outflow 0

Committed Water 9.8Navigation 1.8Environment maintenancea (assumed) 8.0

Uncommitted Outflow 14.1–9.8 4.3Available Water 53.7 – 9.8 43.9Available for irrigation 43.9–1.6 (M&I) 42.3

IndicatorsDepleted fraction (gross and net) 39.6/53.7 0.74Process fraction (depleted) 36.4/39.6 0.92Process fraction (available) 36.4/43.9Gross value of production

(1992 US$)b 6,450 millionProductivity per unit of water 6.45/36.3 0.19

depleted by irrigationProductivity per unit of water 6.45/41.9 0.15

available to irrigationProductivity per unit inflow 6.45/55.3 0.12

aThis estimate is made by the author and is made for illustrative purposes.bSource: Ministry of Agriculture and Land Reclamation, 1990.

Notes: ET = Evapotranspiration. M&I = Municipal and industrial uses.

13

and was recorded at 53.2 km3 (cubic kilome-ters). Inflow from groundwater was esti-mated at 0.5 km3, while rainfall and othersources were negligible during this year.The storage change of the groundwater wasassumed to be zero for the annual cycle.Evapotranspiration was estimated at 34.8km3. Depletive use by municipal and indus-trial (M&I) uses was estimated at 1.6 km3

by assuming that 20 percent and 30 percentof diversions are depleted in the Nile Deltaand Valley, respectively. There is consider-able return flow from M&I back to the Nilewater system. Other non-process depletionconsidered was evaporation from free watersurfaces, fallow land, and phreatophytes,estimated at 3.2 km3. The total depletion ofthe net inflow of 53.7 km3 is 39.6 km3.

Measured outflows are 1.8 km3 fromthe Nile River to the sea, and 12.3 km3 from

the drains to the sea. But 1.8 km3 of thisoutflow during that particular year wascommitted to maintain water levels fornavigation. Some drainage outflow isrequired to maintain the environment atpresent levels, and is roughly estimatedhere at 8 km3 based on the need to removesalts and pollutants from the Nile, and tomaintain freshwater fisheries. With thisestimate for environmental commitments,the remaining uncommitted outflow is 4.3km3.

The depleted fraction of net and grossinflows for the basin is 0.74 (in this casegross inflow equals net inflow). Theprocess fraction of depleted water is 0.92and the process fraction of available wateris 0.83. A very high percentage of bothdepleted and available water is depleted bythe intended processes.

The Water Accounting Research Agenda

The terminology and framework presentedhere are developed for general use under avariety of conditions. The concept is meantto be supportive for a number of activities,including identification of opportunities forwater savings and increasing water produc-tivity; support of the decision process forwater allocation; general water audits andperformance studies; and conceptualizingand testing interventions. A few exampleswere provided here for illustration, but forthe procedures and terminology to developinto standards these need to be tested andrefined under a variety of conditions.

Water accounting requires water bal-ances. Some important shifts in irrigationwater measurement programs are required.Measurement of irrigation water is often fo-cused on diversions. Yet to complete the

water balance, better knowledge is requiredof the drainage outflow. Estimates of non-process depletion of water are rarely madeat the irrigation system and basin levels.

Critical in the water accounting processis the estimation of evaporation and transpi-ration. Our uncertainty about results in-creases when shifting from use level to ser-vice level to basin level. New technologiesand techniques of remote sensing showpromise in improving estimates of evapora-tion and transpiration and for separatingprocess and non-process evaporation(Bastiaanssen, forthcoming).

The interaction between water accountsat the field level and at the irrigation serviceand basin levels needs to be better illus-trated. Claims of water savings based onfield trials should be made by placing the

14

consequences of field-level practices interms of service and basin-level analyses.Reporting of field-scale results in a wateraccounting format will assist researchersand designers to point out appropriatefield-level strategies that will achieve watersavings and increases in water productivity.

At the service and basin levels, there isa need for documentation on the depletionof water by other nonagricultural users ofwater, and the returns from that depletion.It is widely known that within irrigated ar-eas and within basins, there are other im-portant uses of water. Oftentimes, we knowthe diversion to these uses, but we have noidea how much is depleted. Some detailedcase studies should shed light on the deple-tion of water by these uses.

Quality of water plays an extremelyimportant role in water depletion and waterproductivity. A means of accounting fordepletion due to pollution is required. Thisis particularly true in basins where water isfully or near fully committed. Often, inthese cases, recycling intensifies and dilu-tion is not a viable option. While recyclingof water is a viable water saving strategy

the effects of pollution loading on produc-tivity, the environment, and health requireclear accounting beyond the water quantityestimates presented here.

In the field of financial accounting,there exist accepted, standardized proce-dures for performing audits, and auditingof accounts is a widely accepted practice.This is not true in the water field. Therewill be an increasing pressure for account-ability for water use with increasing scarcityof water. One direction for further researchstarting from water accounting is the devel-opment of standardized, widely acceptedprocedures for performing water audits.

Water accounting will assist in a betterunderstanding of present patterns of wateruse, improving communication among pro-fessionals and communication to non-waterprofessionals, improving the rationale forallocation of water among uses, and foridentification of means to achieve watersavings and increases in water productivity.It is expected that over a short period oftime this framework will develop into com-monly applied practices and presentationsto help us better utilize our water resources.

15

Literature Cited

Abu-Zeid. 1992. Water resources assessment for Egypt. In Roundtable on Egyptian Water Policy, ConferenceProceedings, ed. Abu-Zeid and D. Seckler. Water Research Center, Ministry of Public Works and Water Re-sources, Cairo, Egypt and Winrock International, Arlington, Virginia, USA.

Bagley, J. M. 1965. Effects of competition on efficiency of water use. Journal of Irrigation and Drainage Division ofthe American Society of Civil Engineers 91 (IR 1): 69–77.

Bastiaanssen, W. G. M. Forthcoming. Does information from satellite remote sensing have potential to support themanagement of irrigated water basins? Research Report. Colombo, Sri Lanka: International Irrigation Man-agement Institute.

Bastiaanssen, W. G. M., J. K. Schakel, R. Singh, S. Kumar, and R. K. Jhorar. 1997. Analysis and recommendationsfor integrated on-farm water management in Haryana, India: A model approach. Report 118. Wageningen, TheNetherlands: Dienst Landbouwkundig Onderzoek (DLO)-Winand Staring Centre.

Bhuyian, S. I., M. A. Sattar, and M. A. K. Khan. 1995. Improving water use efficiency in rice irrigation throughwet seeding. Irrigation Science (16): 1–8.

Binder, C., R. Schertenleib, J. Diaz, H. P. Bader, and P. Baccini. 1997. Regional water balance as a tool for watermanagement in developing countries. Water Resources Development 13 (1): 5–20.

Boels, D., A.A.M.F.R. Smit, R. K. Jharar, and J. Singh. 1996. Analysis of water management in Sirsa District inHaryana: Model testing and application. Report 115. Wageningen, The Netherlands: Dienst LandbouwkundigOnderzoek (DLO)-Winand Staring Centre.

Bos, M. G. 1979. Der Einfluss der Grosse der Bewasserungs einheiten auf die verschienden Bewasserungswirkungsgrade. Zeitschrift für Bewasserungs Wirtschaft, Bonn 14(1): 139–155.

Bos, M. G., and W. Wolters. 1989. Project or overall irrigation efficiency. In Irrigation theory and practice, 499–506, ed. J. R. Rydzewski and C. F. Ward, 499–506. Proceedings of the International Conference held at theUniversity of Southampton, 12–15 September, 1989. London, UK: Pentech Press.

Hassan, G. Z., and M. N. Bhutta. 1996. A water balance model to estimate groundwater recharge in RechnaDoab, Pakistan. Irrigation and Drainage Systems (10): 297–317.

Helal, M. A. Nasr, M. Ibrahim, T. K. Gates, W. O. Ree, and M. Semaika. 1984. Water budgets for irrigated regionsin Egypt, Egypt Water Use and Management Project. Technical Report 47. Cairo, Egypt: Egypt Water and Man-agement Project.

Keller, J., 1992. Implications of improving agricultural water use efficiency on Egypt’s water and salinitybalances. In Roundtable on Egyptian Water Policy, Conference Proceedings, ed. Abu-Zeid and D. Seckler. WaterResearch Center, Ministry of Public Works and Water Resources, Cairo, Egypt and Winrock International,Arlington, Virginia, USA.

Keller, A., and J. Keller. 1995. Effective efficiency: A water use concept for allocating freshwater resources. WaterResources and Irrigation Division Discussion Paper 22. Arlington, VA, USA: Winrock International.

Keller, A., J. Keller, and D. Seckler. 1996. Integrated water resource systems: Theory and policy implications. Re-search Report 3. Colombo, Sri Lanka: International Irrigation Management Institute.

Kijne, C. J. 1996. Water and salt balances for irrigated agriculture in Pakistan. Research Report 6. Colombo, SriLanka: International Irrigation Management Institute.

Ministry of Agriculture and Land Reclamation. 1990. The annual bulleting of agricultural economics. Cairo, Egypt:Ministry of Agriculture and Land Reclamation.

Mishra, H. S., T. R. Rathore, and V. S. Tomar. 1995. Water use efficiency of irrigated wheat in the tarai region ofIndia. Irrigation Science 16 (2): 75–80.

Molden, D. J., R. Sakthivadivel, C. J. Perry, and C. de Fraiture. Forthcoming. Indicators for comparing the perfor-mance of irrigated agriculture. Research Report. Colombo, Sri Lanka: International Irrigation ManagementInstitute.

Owen-Joyce, S. J., and L. H. Raymond. 1996. An accounting system for water and consumptive use along theColorado River, Hoover Dam to Mexico. U.S. Geological Survey Water-Supply Paper 2407. Washington, D.C.,USA: Government Printing Office.

Palacios-Velez, E. 1994. Water use efficiency in irrigation districts. In Efficient water use, ed. H. Garduno and F.Arreguin-Cortes. Montevideo: UNESCO/ROSTLAC.

16

Perry, C. J. 1996a. Quantification and measurement of a minimum set of indicators of the performance of irrigationsystems. January, 1996. Colombo, Sri Lanka: International Irrigation Management Institute. Duplicated.

Perry, C. J. 1996b. The IIMI water balance framework: A model for project level analysis. Research Report 5.Colombo, Sri Lanka: International Irrigation Management Institute.

Rathore, A. L., A. R. Pal, R. K. Sahu, and J. L. Chadhary. 1996. On-farm rainwater and crop management forimproving productivity of rainfed areas. Agricultural Water Management 31: 253–267.

Seckler, D. 1996. The new era of water resources management: From “dry” to “wet” water savings. Research Report1. Colombo, Sri Lanka: International Irrigation Management Institute.

Seckler, D. 1993. Designing water resources strategies for the twenty first century. Water Resources and IrrigationDivision Discussion Paper 16. Arlington, Virginia, USA: Winrock International.

Seckler, D. 1992. Irrigation policy, management and monitoring in developing countries. In Roundtable onEgyptian Water Policy, Conference Proceedings, ed. Abu-Zeid and D. Seckler. Water Research Center, Ministryof Public Works and Water Resources, Cairo, Egypt and Winrock International, Arlington, Virginia, USA.

Tuong, T. P., R. J. Cabangon, and M. C. S. Wopereis. 1996. Quantifying flow processes during land soaking ofcracked rice soils. Soil Science Society of America Journal 60 (3), (May–June 1996).

Van Vuren, Gerrit. 1993. Constraints in the use of irrigation efficiency coefficients. ICID Fifteenth Congress. TheHague. The Netherlands: International Commission on Irrigation and Drainage (ICID).

Viets, F. G. 1962. Fertiliser and efficient use of water. Advances in Agronomy 14: 223–264.Willardson, L. S. 1985. Basin-wide impacts of irrigation efficiency. Journal of Irrigation and Drainage Engineer-

ing 111 (3), (September 1985).Willardson, L. S., R. G. Allen, and H. D. Frederiksen. 1994. Universal fractions and the elimination of

irrigation efficiencies. Paper presented at the 13th Technical Conference, USCID, Denver, Colorado, Octo-ber 19-22, 1994. Duplicated.

Wolters, W., and M. G. Bos. 1990. Interrelationships between irrigation efficiency and the reuse of drainage water.Symposium on Land Drainage for Salinity Control in Arid and Semi-Arid Regions, Cairo, Egypt, 25 Febru-ary-2 March 1990. 3: 237–245. Presented papers. Session III-Session IV-Session V. Cairo, Egypt: Ministry ofPublic Works and Water Resources.

Wright, K. R. 1964. Water rights engineering under appropriation system. Journal of the Irrigation and DrainageDivision, ASCE 90 (IR 1): 9–15.

Accounting for Water Useand Productivity