California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in Californ

Available online at www.sciencedirect.comJournal of Integrative Agriculture2013, 12(8): 1371-1388心ScienceDirectAugust 2013RESEARCH ARTICLECalifornia Simulation of Evapotranspiration of Applied Water and AgriculturalEnergy Use in CaliforniaMorteza N Orang', Richard L Snyder2, Shu Geng23, Quinn J Hart, Sara Sarreshteh2, Matthias Falk2, DylanBeaudette, Scott HayesI and Simon Eching'2 Department of Land, Air and Water Resources, University of California, Davis, CA 95616, USA3 School of Environment and Energy, Shenzhen Graduate School, Peking University, Shenzhen 518055, P.R.ChinaAbstractThe Califormia Simulation of Evapotranspiration of Applied Water (CaI-SIMETAW) model is a new tool developed by theCalifornia Department of Water Resources and the University of California, Davis to perform daily soil water balance anddetermine crop evapotranspiration (ET), evapotranspiration of applied water (ETgw), and applied water (AW) for use inCalifornia water resources planning. ETw is a seasonal estimate of the water needed to irrigate a crop assuming 100%irrigation efficiency. The model accounts for soils, crop cofficients, rooting depths, seepage, etc. that influence crop waterbalance. It provides spatial soil and climate information and it uses historical crop and land-use category information toprovide seasonal water balance estimates by combinations of detailed analysis unit and county (DAU/County) over California.The result is a large data base of ET。and ET that will be used to update information in the new California Water Plan (CWP).The application uses the daily climate data, i.e., maximum (T) and minimum (T) temperature and precipitation (Pp), whichwere derived from monthly USDA-NRCS PRISM data (PRISM Group 2011) and daily US National Climate Data Center(NCDC) climate station data to cover California on a 4 kmx4 km change grid spacing. The application uses daily weather datato determine reference evapotranspiration (ET.), using the Hargreaves-Samani (HS) equation (Hargreaves and Samani 1982,1985). Because the HS equation is based on temperature only, ET。from the HS equation were compared with CIMIS ET。atthe same locations using available CIMIS data to determine correction factors to estimate CIMIS ET。from the HS ET。toaccount for spatial climate differences. Cal SIMETAW also employs near real-time reference evapotranspiration (ET.)information from Spatial CIMIS, which is a model that combines weather station data and remote sensing to provide a grid ofET. information. A second database containing the available soil water holding capacity and soil depth information for all ofCalifornia was also developed from the USDA-NRCS SSURGO database. The Cal-SIMETAW program also has the ability togenerate daily weather data from monthly mean values for use in studying climate change scenarios and their possibleimpacts on water demand in the state. The key objective of this project is to improve the accuracy of water use estimates forthe California Water Plan (CWP), which provides a comprehensive report on water supply, demand, and management inCalifornia. In this paper, we will discuss the model and how it determines ETw for use in water resources planning.Key words: soil water balance, crop water requirements, weather generator, water resource planning, crop coffcient, energy usetion of Evapotranspiration of Applied Water or Cal-INTRODUCTIONSIMETAW was specifically designed to provide the bestpossible information on agricultural water demand forThe daily soil water balance model California Simula-use in the California Water Plan, updated every five中国煤化工Received 17 October, 2012 Acepted 10 January, 2013YHCNMHGCorrespondence Morteza N Orang. Tel: +1-916-6537707, E-mail: morang @ water.ca gov⑥2013,CAAS. All nghts rseved. PulisedbylEseviertd.1372Morteza N Orang et al. .years to present the status and trends of California'swater requirement. It differs from irrigation efficiency,water- dependent natural resources; water supplies; andwhich includes the crop water requirements, water usedagricultural, urban, and environmental water demandsfor frost protection, and leaching requirements, i.e.,for a range of plausible future scenarios. Californiabeneficial uses, divided by AW over a cropping season.agriculture is a multibillion dollar industry, number oneA major goal of this project was to improve infor-producer in the nation, and largest consumer of water.mation on current and future water demand. Cal-The agricultural water demand is high and increasing SIMETAW was developed for water demand planningbecause water supplies are limited and competition forand it can help to plan for the effects of climate changethose supplies is growing. The main factors that areas well as for current climate conditions. Improve-causing increases in agricultural water demand are thements to the input information and data processing inpopulation growth and demand for food and fiber. At Cal-SIMETAW greatly enhances our ability to rapidlythe same time, the demand for urban and environmen-and accurately determine ETaw for 20 crop categoriestal water uses is increasing. The California Depart- and 4 land-use categories by each DAU/County withinment of Water Resources (DWR) and the University ofCalifornia. All of the ET w calculations are done on aCaliformia, Davis (UC Davis) are keenly aware of thedaily basis, so the estimation of effective seepage ofneed for good planning, and Cal-SIMETAW model wasgroundwater, effective rainfall and, hence, ETw isdeveloped to address the planning needs. The Cal-greatly improved over earlier methods. In addition, theSIMETAW computer application program was writtenuse of the widely adopted Penman-Monteith equationusing Microsoft C# for calculations and Oracle Spatialfor reference evapotranspiration (ET) and improved11g for data storage, as a tool to help DWR obtain ac-methodology to apply crop coefficients for estimatingcurate estimates of crop evapotranspiration (ET),crop evapotranspiration (ET) is used to improve ETWevapotranspiration of applied water (ET) for agricul-accuracy for climate change and long-range water re-tural crops, and urban landscapes, which account for source planning.most evapotranspiration losses and water contributionsCal-SIMETAW uses batch processing to read (1) thefrom ground water seepage, precipitation, and irrigation.climate data, (2) the surface/crop coefficient values,Crop evapotranspiration is computed as the product(3) growth dates to estimate annual curves, (4) soilof reference evapotranspiration (ET) and a crop coef- information, (5) crop and irrigation information, andficient (K) value, i.e., ET =ET ,xK, and ET , which is(6) surface area of each crop and land-use category onequal to the seasonal evapotranspiration minus watereach of the 482 DAU/Counties. Then, the programsupplied by stored soil moisture, effective rainfall, andcomputes daily ET, K. factors, ET。, daily water balance,seepage from canals. Cal-SIMETAW accounts for effective rainfall, ETw, etc. for every surface withincontributions from rainfall and for ground water seep-each of the 482 DAU/Counties over the period of record.age from the rivers and canals when spatial informa-The water balance model is similar to that used in thetion on the depth to water table is available on theSimulation of ET of Applied Water (SIMETAW) applica-same 4 kmx4 km grid spacing used to characterizetion program, which was also developed as a coopera-soils within California.tive effort between the UC Davis and the DWR (SnyderCal-SIMETAW has the capability to estimate appliedet al. 2012). The main difference between the originalwater (AW) by crop and land-use category for eachSIMETAW model and Cal-SIMETAW is that SIMETAWdetailed analysis unit and county (DAU/County) com-uses historical or generated climate data to determine abination in the state. Applied water is estimated as thedaily water balance for individual cropped fields within aET divided by the mean seasonal irrigation systemwatershed region having one set of ET。estimates,application efficiency. Thus, the AW supplies estimateswhereas Cal-SIMETAW uses historical or generated cli-of the diversions needed by DWR to plan its futuremate data and batch fles of soil and climate data to com-water demand for irrigated agriculture. Seasonal sys-pute daily water balance for 20 crop categories, 4 land-tem application efficiency is an estimate of the fractionuse categories中国煤化IDAUCountyof AW irigation water is used to contribute to the crops regions that exhYHCNMHGdemandand⑥2013, CAAS. Alights reseved. Published by EsevierLtd.California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in California1373rainfall. Cal-SIMETAW was designed to reduce the timethe USDA-NRCS SSURGO database (SSURGO 2011).needed for data input and to improve the water use/de-The developed data base covers all of California on themand estimates needed for the California Water Plan.same 4 kmx4 km grid as was used in the SSURGOThe simulation component of Cal-SIMETAW is use-database.ful for studying the effect of climate (e.g., temperature,Using mean soil characteristics and climate and ET。humidity, CO2 concentration, and rainfall) change on crop information from the 4 kmx4 km grid, Cal- SIMETAWevapotranspiration (ET) and evapotranspiration of ap- estimates the mean soil characteristics and ET。infor-plied water (ET、). One of main features of Cal-mation by DAU/County. The PRISM climate data baseSIMETAW is that it can simulate daily weather data from(PRISM Group 201 1), the Hargreaves Samani equation,monthly climate data, and the simulated data are used toand a calibration factor to convert ETys to ET。are usedestimate reference ET。Because of this feature, Cal-to estimate reference evapotranspiration (ET). CropSIMETAW allows the examination of the impact of mul-evapotranspiration is estimated using the single CrOPtiple management scenarios on agricultural water demandcoefficient approach (Doorenbos and Pruitt 1977; Allenusing GCM scenarios and regional downsizing models.et al. 1998). Up to 20 crop and 4 land-use categoriesUsing different climate change scenarios ftom GCMare used to determine weighted crop coefficients tomodels and a downsizing model to determine means ofestimate crop evapotranspiration (ET) using the singlemonthly climate data for 2030 and 2050, Cal-SIMETAWcrop coefficient approach (Doorenbos and Pruitt 1977;can simulate daily weather data from the monthly meanAllen et al. 1998). A daily water balance is computedof solar radiation, maximum and minimum temperature,using input soil and crop information and ET。Thewind speed, and dew point temperature data to deter-model can use daily observed climate data or it canmine ET。ET, and ETW for 20 crop categories and 4generate simulated daily climate data from monthly dataland-use categories in each of the 10 hydrologic regionsto estimate daily ET。Information from Spatial CIMIS,in Califormia. The ability to change the CO, concentra-which is a model that combines weather station datation was included in Cal-SIMETAW to more accuratelyand remote sensing to provide a grid of ET。informa-estimate the effect of climate change on ET。in additiontion is also used by Cal SIMETAW to estimate nearto changes in temperature and humidity.real-time ET。Cal-SIMETAW is used by DWR to estimate cropevapotranspiration (ET) and evapotranspiration of ap-MODEL DESCRIPTIONplied water (ET ), which is the sum of net irrigationapplications needed to produce a crop. Thus, ET pro-vides an estimate of the water needed to achieve fulling Microsoft C# for numerical calculations, graphics,evapotranspiration in addition to that water supplied byetc. and Oracle software for data storage. In the Cal-preseason soil moisture and in-season effective rainfallSIMETAW project, soil and climate database were devel-assuming 100% application efficiency. Dividing the EToped to spatially characterize ET and ETw Oracle soft-by the mean seasonal application efficiency (AE) pro-ware was used to store the historical daily climate data,vides an estimate of the seasonal water diversions neededi.e, maximum (T) and minimum (T) temperatures andto produce a fully irigated crop. The application effi-precipitation (P ), which were derived from monthlyciency is the ratio of irrigation water applied that con-PRISM data that cover Califonia on a4 kmx4 km grid tributes to evaporanspiration to the total aplied water.spacing. Because the PRISM data are monthly andA first guess for the ET_w would be SET, which is ,daily data are needed to determine ET , daily NCDCthe seasonal total ET。, minus the change in stored soilclimate station data (from October 1921 to Septemberwater during the season and minus any in-season ef-2010), were used with the PRISM data to estimate dailyfective rainfall. Therefore, ET = SET. -SR. -ASW,T, T, and P。The daily climate data development iswhere SR。is the seasonal effective rainfall and ASW=described later in this paper.SW; -SW, is ch' 中国煤化ri'e initial soilA second database containing the soil water holdingwater content (SFontent (SW).capacity and soil depth information was developed from If the seasonalTYHCNMHGCorrectly, the⑥2013, CAAS. Alights reseved. Published by EsevierLtd.1374Morteza N Orang et al.2NA= =SET -SR.-ASW. The Cal-SIMETAW model useshumidity, and wind speed data were lacking from mostcrop, soil, and climate or weather data to determine theclimate data sets prior to development of CIMIS, i.e.,ETw using the sum of a daily soil water balance. Thethe California Irrigation Management Information Sys-generated ETw information provides an estimate of ag-tem (Snyder and Pruitt 1992). Since only tempera-ricultural water demand and thus is important for the ture data were available prior to 1986, it was decidedCalifornia Water Plan.to use daily maximum and minimum temperatures andIn addition to using historical data, the weather gen-the Hargreaves and Samani (1982, 1985) equation toerator in Cal-SIMETAW can simulate regional dailycalculate reference evapotranspiration (ETg) as anweather data from monthly climate data that areapproximation for ET。Using recent climate data fromdownscaled from a GCM“General Circulation Model"CIMIS, comparisons were made between ETHs andto estimate ET。, ET。and ETw . Using crop coefficientET。and calibration factors were developed to esti-data for 20 crop and 4 land-use categories, Cal.mate ET。from ETHs as a function of wind speed andSIMETAW estimates daily ET , SET, and ETg fromsolar radiation. In general, ETHs was lower than ET。2030 and 2050 climate projections for each of the 10under windy conditions and it was higher than ET。hydrologic regions within Califormia for use in the Cali- under calm conditions. Using approximately 130formnia Water Plan.CIMIS weather stations distributed across the state, a4 kmx4 km grid of correction factors for the ETHsPurpose of detailed analysis/County units (DAU/equation was developed. There are many daily tem-Countyperature and precipitation weather stations in Califormia,but the PRISM data set (PRISM Group 2011) pro-DWR has subdivided California into 482 DAU/Counties,vided a long-term GIS data base of historical dailywhich are geographic areas having relatively uniformmaximum and minimum temperature and precipita-ET。throughout the region. The regions are used fortion on the same 4 kmx4 km grid as the correctionestimating water demand by agricultural crops and otherfactor GIS map. Thus, using the PRISM historicalsurfaces for water resources planning. DAUs are basedtemperature data to compute ETs and the calibrationon watershed and other factors related to water trans-factors, Cal-SIMETAW is able to produce ET。esti-fer and use within the region, which are often split bymates on a 4 kmx4 km grid over the state from Octo-counties. DAU/Counties are the smallest study areasber 1921 to September 2010.used by DWR. The largest study areas comprise theten hydrologic regions. Land use surveys are periodi-ET。correction factorscally completed within each DAU/County by DWRstaff, and the percentages of each crop within a mul- NationalClimate Data Center (NCDC) stations were pairedtiple crop/land-use category are recorded for most DAU/with neighboring CIMIS stations from 1986 throughCounty regions. Using the percentages of each crop2010. Corresponding data for the paired stations werewithin a DAU/County, the individual crop coefficientsselected from the University of California Integrated Pestand growth rates are analyzed to determine a weightedManagement (UC IPM) site (t:/ipm.ucdavis.edu). Themean K。curve for each category. Thus, each DAU/daily Penman and Monteith equation was used to calcu-county can have as many as 20 crop and 4 land-uselate reference evapotranspiration (ET ) using daily CIMIScategories with weighted mean K。curves (Fig. 1).data and the HS equation was used to calculate ET。(ET)using daily Ty and T。data. The correction factor (Cp)Reference evapotranspiration(ET。)was calculated as: C= =ET /ETys. Spatial interpolationwas completed using ARC GIS and a 4 km gridded ras-Weather and climate data are commonly used to calcu-ter map for CF was produced (Fig. 2). The CF valueslate standardized reference evapotranspiration (ET) forfell within 15% of 1.0. The CF values were archived forshort canopies (Monteith 1965; Monteith and Unswortheach 4 kmx4 km中国煤化工s were stored1990; Allen et al. 1998, 2005), but solar radiation,in files designatYHCNMHGmber.⑥2013, CAAS. Alights reseved. Published by EsevierLtd.California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in California1375Spatial CIMIS ET。programto satisfy the Penman-Monteith ET。equation (Hartet al. 2000).The Califormia Irigation Management Information Sys-tem (CIMIS) is a program developed by UC DavisReal-time Cal-SIME TAWand operated by DWR to help farmers, turf and land-scape managers and other resource managers to de-Cal-SIMETAW provides a method to analyze histori-velop water budgets that improve irrigation schedul-cal data to determine trends in agricultural watering and monitor water stress. CIMIS weather sta-demand, but it is also useful for near real-time detions are located at key agricultural and municipal sitesmand estimates. Although there are about 130 CIMISthroughout Califormia to collect comprehensive, timely,weather stations in California, many locations haveweather data on an hourly basis and to disseminatelimited weather data for ET。estimation, so there arethe weather and ET , data to help farmers and land-gaps in the spatial data. To resolve this problem, DWRscape professionals to improve the efficient use ofand UC Davis used satellite data and developed spatialirrigation water. For the Spatial CIMIS program,CIMIS to estimate ET。on a 4 kmx4 km grid over theweather data collection system is combinedstate. Since the Spatial CIMIS uses the same grid aswith NOAA Geostationary Operational EnvironmentalCal-SIMETAW and it provides near real time ET。Satellite (GOES) visible satellite data to to extend the(i.e., up through the previous day), the output fromreference evapotranspiration (ET) estimates to areasSpatial CIMIS was incorporated into Cal-SIMETAWnot well covered by CIMIS and to provide daily spa-and to develop near real-time daily maps of crop ETc.tial ETo maps. The maps are calculated on a (4 kmx4Spatial CIMIS is available and explained on the CIMISkm) square grid, which is a high spatial resolutionwebsite (CIMIS 2011).when compared to the density of CIMIS stations. Thehourly GOES satellite images are used to estimatecloud cover which are used in turn to modify clearVerification of ET。datasky radiation estimates. These are combined with in-terpolated CIMIS weather station meteorological data Results from Cal-SIMETAW were validated againstHR and DAU/County boundariesSan Francisco BaySouthCoastSacramento RiverSan Joaquin RiverNorth LahontanSouth Lahontan! Colorado RiverDetail Analysis UnitCounty lineCrrection factor: HSto PM ET。1 High: 1.14111Low: 0.821548中国煤化工:Fig. 2 Correction f:ing Hargreaves-Fig. 1 California study area map showing hydrologic regions, detailed Samani ET。(ETH:YHC N M H Gfor California.analysis units (DAU), and counties.ET=HTHSxCF.⑥2013, CAAS. Alights reseved. Published by EsevierLtd.1376Morteza N Orang et al.spatial CIMIS ET。estimates from 2004 to present (Figs.10一Spatial CIMIS3-6).Cal-SIMETAWCIMIS network station measurements are amongthe most reliable direct datasets of daily weather vari-ables including solar radiation (R ), maximum air tem-perature (Tm), minimum air temperature (T), windspeed (U), dew point temperature (T), and etc. Ref-erence evapotranspiration (ET), computed by the daily200420052006(24-h) Penman-Monteith equation, has been recom-Time (d)mended by both America Society of Civil Engineers(ASCE) and United Nation FAO. As a final verifica-Fig. 3 Comparison of daily ET。estimates versus time from Cal-tion of our calibrated Hargreaves Samani equation for SIMETAW and Spatial CIMIS for PRISM grid number 50-60, Januaryestimating ET , a comparison of the calibrated ETHS2004-July 2007.from Cal-SIMETAW and CIMIS-based estimates ofET。with data from Davis, California are shown inFigs. 7-9. The results show that estimates of ET。fory=0.98x1990-2007 closely approximate ET。 values fromR2-0.92冒8CIMIS. The mean ET。estimates from Davis for the看7period of 1990-2007 were 3.90 and 3.94 mm with层6standard deviations of 2.25 and 2.52 mm for the cali-brated Hargreaves-Samani model and CIMIS,respectively. The difference between the two ap-proaches was small (roughly 1%).Crop and land-use categories0↑23↓56789 7oSpatial CIMIS ET。estimates (mmd')Daily soil water balance is the key component of theETgw model. The calculations require input of weather Fig. 4 Comparison between daily ET。estimates from Cal-or climate data, soil depth and water-holding capacity,SIMETAW versus Spatial CIMIS for PRISM grid number 50-60from January 2004 through July 2007.crop root depth, and seasonal crop coefficient curves.Because there are thousands of soil and cropping pat-tern combinations (including differences in cropping soil depth, and rooting depth information for all of Cali-seasons), it is impossible to account for all combina-fornia was developed from the USDA-NRCS SSURGOtion in the state. The biggest limitation is the lack ofdatabase (SSURGO 2011). The developed database cov-both historical and current cropping patterners all of California on the same 4 kmx4 km grid for allinformation. In recent years, however, the croppinglocations that are included in the PRISM database, whichinformation has dramatically improved and refine-covers most of California. There are about 26 300ments are likely in the future. To overcome the prob-PRISM grids in the model's database for California.lem of too many crop and soil combinations, the cropswere separated into 20 crop and 4 land-use catego- Crop cofficientsries that consist of surfaces with similar character-istics (Table 1).Crop evapotranspiration is estimated as the product ofreference evapotranspiration (ET) and a crop coeffi-Soils characteristics and rooting depthscient (K ) value. Cron cnefficients are commonly de-veloped by mea中国煤化工r., and deter-A database containing the soil water holding capacity,mining the rYHCNMHGoftheCal-⑥2013, CAAS. Alights reseved. Published by EsevierLtd.California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in California1377240-Spatial CIMIS220- ---- CaI-SIMETAW200-180-160-140-120-100-8060-4020Time (mon-yr)Fig. 5 Comparison of monthly total ET。estimates versus time for Cal-SIMETAW and Spatial CIMIS for PRISM grid number 50-60,January 2004-December 2006.240Calculated from Cal-SIMETAW220 --1.006612------ Obtained from CIMIS200- R2-0.9908190 ]? 10-140 -120 .“80-601990199199219931994Time (d)0十204068010012014016180200220240Fig. 7 Comparison of daily ET。estimates for Cal-SIMETAW andSpatial CIMIS ET。cstimates (mm mon')CIMIS at Davis, California within the PRISM grid 99-62 from1990 to 1994.Fig. 6 Comparison between monthly ET。estimates from Cal-SIMETAW versus Spatial CIMIS for PRSIM grid number 50-60valuesare used to estimate daily K。values during afrom January 2004 to December 2006.season.One of main objectives of this project was to refineSIMETAW crop coefficient values were developed inand improve crop coefficient values for 20 crop cat-California, but some were adopted from Doorenbos andegories on each of the 482 DAU/Counties within thePruitt (1977) and Allen et al. (1998). While crop coef-state using the County Ag Commissioner reportsficients are continuously developed and evaluated, Cal-(CDFA) and DAU boundaries. Crop categories thatSIMETAW was designed for easy updates of both Krepresent individual crops have seasonal crop coeffi-and crop growth information. Also, K。values needcient (K) curves, but categories containing multipleadjustment for microclimates, which are plentiful andcrops do not have a single seasonal K。curve. Usingextreme in California. A microclimate K。correctionthe percentages of each crop within a DAU/County,based on the ET。rate is included in the Cal-SIMETAWthe crop coeffil中国煤化工analyzed tomodel. The K。values and corresponding growth datesdetermine a wiFor each cropfMHCNM HGare included by crop in the model. These dates and Kccategory.⑥2013, CAAS. Alights reseved. Published by EsevierLtd.1378Morteza N Orang et al.Estimated from CaI-SIMETAWObtained from CIMIS1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Time (mon)Fig. 8 Comparison of monthly mean ET。estimates versus time for Cal- SIMETAW and CIMIS at Davis, California within the PRISM grid99-62 from 1990 to 2007 time period.-1.0087x8about 10 to 75% ground cover, the Kc value increasesR2=0.9725linearly from K B to K C. The K。values are typically a7constant value during midseason, so K C=K, D. Dur-ing late-season, the Kc values decrease linearly fromKD to K E at the end of the season (Fig. 10).Doorenbos and Pruitt (1977) provide estimatednumber of days for each of the four growth periodsto help identify the end dates of growth periods. Be-cause there are climate and varietal differences,however, and because it is difficult for growers toknow when the inflection points occur, irrigators of-Cal-SIMETAW ET。(mm)ten find this confusing. To simplify this problem, per-centages of the season from planting to each inflec-Fig. 9 Comparison of monthly mean ET。for CIMIS versus Cal-tion point rather than days in growth periods are usedSIMETAW at Davis, California within the PRISM grid 99-62 from1990 to 2007.(Fig. 10). Irrigation planners need only enter the plant-ing and end dates and the intermediate dates are deter-mined from the percentages, which are easily storedField and row cropsin a computer program.During initial growth of field and row crops, a de-Field and row crop K。values are calculated using afaultK; =K B=K A unless it is overridden by entering anmethod similar to that described by Doorenbos and initial growth K. based on rainfall or imrigation frequency.Pruitt (1977) and Allen et al. (1998). A generalizedThe values for K C=K D depend on the difference incurve is shown in Fig. 10. In their method, the season (1) light interception, (2) crop morphology effects onis separated into initial (date A-B), rapid (date B-C),turbulence, and (3) physiological differences betweenmidseason (date C-D), and late season (date D-E) the crop and reference crop. Some field crops aregrowth periods. K。values are denoted K A, K B, K C,harvested before senescence, and there is no late sea-K.DandKEattheendsoftheA,B,C,D,andEgrowthson drop in K。(for example, silage corn and fresh mar-dates, respectively. During initial growth, the K。val-ket tomatoes). Relatively constant annual K。values areues are at a constant value, so K A=K B. During thepossible for some crops (for example, turfgrass andrapid growth period, when the canopy increases frompasture) with 1中国煤化工"YHCNMH G .⑥2013, CAAS. AlI ights reserved. Published by EsevierLd.California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in California1379Table 1 Crop and land-use category numbers, symbols and descriptionsLand-use Crop symbolSurface category descriptionGGrain (wheat, wheat_ winter, wheat spring, barley, oats, misc._ grain & hay)RIRice (rice, rice. _wild, rice. flooded, rice-upland)CCottonSISugar beet (sugar-beet, sugar_ beet_ late, sugar. beet early)CorrDDry bean:S/SfflowerFIOther field crops (flax, hops, grain_ sorghum, sudan,castor-beans, misc._ field, sunflower, sorghun/sudan_ hybrid, millet, sugarcaneAAlfalfa (alalfa, alfalfa _mixtures, alfalfa cut, alfalfa_ annual)1(P/Pasture (pasture, clover, pasture_ mixed, pasture_ native, misc._ grasses, turf_ farm, pasture_ bermuda, pasture_ rye, klein_ _grass, pasture_ fescue)1TI1:Tomato fresh (tomato fresh. tomato frestCucurbits (cucurbits, melons, squash, cucumbers, cucumbers_ fre:esh_ market, cucumbers_ machine-harvest, watermelon)Onion & garlic (onion & garlic, onions, onions_ dry, onions_ green, garlic)Potatoes (potatoes, potatoes_ sweet)16THTruck_ Crops_ misc (artichokes, truck. _crops, asparagus, beans_ green, carrots, celery, lettuce, peas, spinach, bus h_ berries, strawberries,peppers, broccoli, cabbage, cauliflower)17AlAlmond & pistacios18o1Orchard (deciduous) apples, apricots, walnuts, cerries, peaches, nectarines, pears, plums. prunes, figs, kiwis)19Citrus & subtropical (grapefruit, lemons, oranges. dates, avocados, olives. jojoba)2(Vineyards (grape_ _table, grape. _raizin, grape. wine)21UUrban landscape (cool-season turf, warm- season turf, golf course, open water)22R\Riparian (marsh, tules, sedges, high water table meadow, teese, shrubs, duck marsh)23Native vegetation (grassland, light brush, medium brush, heavy brush, forest, oak. woodland)24Water surface (river, stream, channel delivery, freshwater Jlake, brackish _saline, wastewater)Planting, 10%Cgr 75%Cp/T-01.47100%3-.2-75%.1-50%、0十20%.9-.8-2 0.7-.6-.4-.3-.2十0.0+Mar-04Apr-04May-04Jun-04Jul-04Aug-04Sep-04Oct-04Nov-04Growth date (mon-yr).... Initial stage - . - Rapid growth-Mid-season - .●Late seasonFig. 10 Hypothetical crop cofficient curve for field and row crops using percentage of the season to delineate growth dates. The seasonends when transpiration (T) from the crop ceases (T).Some field crops and landscape plants (type-2 crops)bare soil K。should be used. The bare soil K。valuehave fixed K。values all year. However, if the signifi-serves as a baseline for the crop coefficient, and thecant rainfall frequency is sufficient to have a higher K。higher of the fix中国煤化工K. is used tofor bare soil than for the selected crop, then the higherestimate ETMHCNMHG⑥2013, CAAS. Alights reseved. Published by EsevierLtd.1380Morteza N Orang et al.Tree and vine crop K valuesthan mature crops. The following equation is used toadjust the mature K。values (K。) as a function of per-Deciduous tree and vine crops, without a cover crop,centage ground cover (C).have K. curves that are similar to field and row cropsbut without the initial growth period (Fig. 11). De- .If sinCg. π> 1.0then Kc=Kcm elseI702_faultK. B, K C=K D=K 2 and KcE K 3 values are in-cluded in Cal-SIMETAW. The season begins withK.=Kcmsir「ex. π(1)rapid growth at leaf out when the Kc increases from70 2J_KB to K C. The midseason period begins at approxi-mately 70% ground cover. Then, unless the crop isSubtropical cropsimmature, the K。is fixed between dates C and D, .which corresponds to the onset of senescence. Forimmature crops, the canopy cover may be less thanFor mature subtropical orchards (for example, citrus),70% during the midseason period. If so, the K。willusing a fixed K。during the season provides acceptableET。estimates. If higher on any given date, however,increase from K.C up to the K D as the canopy coverincreases, so the Cal-SIMETAW model accounts fothe bare soil K。replaces the orchard K。For an imma-K. changes of immature tree and vine crops. Duringture orchard, the mature K。values (K ) are adjustedlate season, the K decreases from K D to K E, whichfor their percentage ground cover (C ) using the fol-lowing criteria.occurs when the transpiration is near zero.Initially, the K. value for deciduous trees and vines(K.B) is selected from a table of default values.|sinCg_ π≥1.0then K.=Kcm or else[702However, the ET is mainly soil evaporation at leaf out,so Cal-SIMETAW contains the methodology to deter-(2)|Cg. πmine a corrected K B based on the bare soil evaporation.K.=Kan/sin|70 2Immature deciduous tree and vine crops use less water1.4710%Cq100% :70%1.1-35%0.9-0.8-义0.7-0.6-0.5-0.4-0.3-Leaf out70%CgLeaf drop0.1pI0.0Jan-04 Feb-04 Mar-04Apr-04 May-04Jun-04Ju1-04 Aug-04 Sep-04 Oct-04 Nov-04Growh date (mon-yr)- - Rapid growth- Mid-season - - ●Late seasonFig. 11 Hypothetical crop cofficient curve for deciduous tre and vine crops using per中国煤化工。; growth dates.There is no initial growth period, so the season starts at leaf out on date B.MHCNM HG⑥2013, CAAS. Alights reseved. Published by EsevierLtd.California Simulation of Evapotranspiration of Applied Water and Agricultural Energy Use in California1381Cover crop corrections1.00With a cover crop, the K. values for orchards and vinesy-2.54r