Abstract: Water resources are scarce for inland river basins in arid regions of northwestern China,where water consumption is mainly for agriculture. Especially in arid regions,evapotranspiration（ET） is the main link to the water cycle and has an important impact on environmental changes. It is therefore important to accurately estimate ET and irrigation efficiency for predicting climate change and for rationally allocating water resources in inland river basins. The Shule River Basin is one of three inland river basins in the Hexi Corridor,and its ecological environment is fragile given its large annual average ET. Water use accounts for a large proportion ,and the utilization rate of agricultural irrigation water is high in this region. In this study,we estimated the daily evapotranspiration （ET） of the Shule River Basin （SRB） based on the Surface Energy Balance Algorithms for Land （SEBAL） model with satellite data （from the moderate-resolution imaging spectroradiometer known as MODIS） and meteorological data from March in 2017 through February in 2018. Then we applied a time-scale extension to obtain monthly and seasonal ET by using daily ET and meteorological data. We also analyzed the spatiotemporal distribution characteristics of ET to derive daily and seasonal ET. Then,combining the ET results from the SEBAL simulation and data collected on irrigation and precipitation for the Changma irrigation district （CM） during 2017–2018,we calculated the coefficient of irrigation-water effective utilization （“η”） for the CM on the basis of the water-balance equation,which was used for irrigation-efficiency assessment for the CM.
The main conclusions are as follows：（1） The variation range of average daily ET values was 0.55–5.03 mm across the SRB during 2017–2018,which showed a single peak change and obvious difference in spatial distribution. （2） There were significant differences in seasonal ET in the SRB;summer ET was the highest （201.83 mm）,followed by spring and autumn,and the winter ET was the lowest （53.92 mm）. The spatial pattern of seasonal ET shows that ET in the upper reaches of SRB was significantly higher than that in the middle and lower reaches,and ET decreased from the southeast to the northwest in the SRB. （3） The quantity of irrigation water in different irrigation periods resulted in the ET differences in the CM. The high-ET area was mainly in the middle and southeastern parts of the irrigation district,and the low-ET area was mainly in the northwestern part and along the margins of the district. （4） The “η” of CM showed a downward trend during the irrigation period,in which the values of “η” for the spring,summer,autumn,and winter irrigation periods were 0.76,0.71,0.69,and 0.55,respectively. The CM had a higher irrigation efficiency,with the annual “η” of 0.67.
Through simulation of monthly ET in the SRB for the period 2017–2018,the annual ET spatial distributions and fluctuations were more detailed and clearer. We found that ET of irrigation land could be accurately estimated by remote-sensing and meteorological data. The “η” of the irrigation district could easily be obtained from ET and measured irrigation data in each irrigation period,which can reflect CM irrigation efficiency during the year in the SRB.

Key words: evapotranspiration, SEBAL model, Shule River Basin, coefficient of irrigation water effective utilization, Changma irrigation district