Objective This paper aims to elucidate the change of surface evapotranspiration and its main influencing factors in Yinchuan City of northwestern China in 2000−2020, and provide theoretical basis for determining cities and forests based on water.

Method Based on MODIS evpotranspiration product MOD16, temperature, precipitation, normalized vegetation index (NDVI), slope linear trend and correlation analysis methods were used to calculate crop water stress index by MOD16A3 data. The spatio-temporal variation characteristics of evapotranspiration in Yinchuan City from 2000 to 2020, the difference of evapotranspiration in different land use types and its influencing factors were analyzed.

Result (1) From 2000 to 2020, evapotranspiration in Yinchuan City ranged from 201.2 to 344.2 mm, with an annual average of 273.4 mm. In terms of spatial distribution, the central region around the built-up area was higher, while the western Helan Mountain region and the eastern region were lower. (2) From 2000 to 2020, the evapotranspiration in Yinchuan City showed an increasing trend, with a tendency rate of 5.093 mm/year, and the area with increasing evapotranspiration accounted for the highest proportion (88.30%). It can be seen that the changing trend of evapotranspiration in most regions of Yinchuan City was increasing, and the significant increase area was mainly distributed in the central area around the built-up area. The area of evapotranspiration decreased was 2.44%, and the area of annual evapotranspiration basically remained unchanged was 9.26%. (3) Compared with 2000, the relative change rates of cultivated land, woodland, grassland, shrub land, wetland, water body, artificial surface and bare land in 2020 were −6.73%, 0.33%, −8.34%, −14.87%, 21.95%, 9.05%, 230.45% and −2.08%, respectively. The spatial distribution of evapotranspiration in Yinchuan City was significantly affected by land use type. The average evapotranspiration was cultivated land (322.9 mm), water body (268.5 mm), artificial surface (256.9 mm), forest land (240 mm), wetland (220.3 mm), shrub land (216.7 mm), grassland (201.9 mm) and bare land (196.4 mm) in descending order. (4) The correlation analysis of evapotranspiration with temperature, precipitation and NDVI (2000−2019) in Yinchuan City from 2000 to 2020 showed that evapotranspiration was significantly positively correlated with temperature, precipitation and NDVI.

Conclusion It can be seen that from 2000 to 2020, temperature, precipitation and NDVI are all driving factors affecting the increase of annual evapotranspiration in Yinchuan City, and NDVI is the most important factor. This study is expected to provide theoretical support for the rational and efficient utilization of water resources in Yinchuan City and better research on ecological water demand and ecological environment planning in Yinchuan City of northwestern China.