Spatial distribution of reference evapotranspiration (ET0) is essential in water resources planning and management, especially in semi-arid areas. In this paper, a digital elevation model is used in an ‘interpolate-then-calculate’ approach to calculating the spatially distributed ET0 using the physically based Penman–Monteith equation in the Taoer river basin in China. The results show the following. (1) Of 11 interpolation methods, the Inverse Distance Weighting (IDW) method was found to be best for interpolating wind speed and a tri-variate secondary trend surface method was found most suitable for interpolating mean air temperature and relative humidity. Spatial modelling of the radiation environment considered the effects of elevation, slope and aspect. (2) Monthly values in January for the three meteorological variables showed larger spatial variations than in July, and just the reverse of net surface radiation. (3) The resulting ET0 calculated at each grid cell with 200 m resolution and its spatial variation showed strong seasonal variation. Lower ET0 was found in high-elevation southern Great Xingan mountains in the northwest basin, while higher values were located in the plains adjacent to the lower reach. (4) The ET0 distribution by the ‘interpolate-then-calculate’ approach better reflected the effects of topography than that of the ‘calculate-then-interpolate’ approach.
Keywords: interpolate-then-calculate, Penman–Monteith equation, reference evapotranspiration, spatial distribution, Taoer river basin, topographic influences