Comparison of spatial interpolation methods for yield response factor of winter wheat and its spatial distribution in Haihe basin of north China

Yield response factor (K _y) is an important basis for implementing efficient irrigation and optimal water allocation. Because K _y varies in different sites, understanding its spatial distribution plays an important role in optimization irrigation in Haihe basin. After determining the K _y and ET₀ of winter wheat, an exponentially increasing function was found between the two parameters. Then, spherical and exponential semivariograms were chosen as proper theoretical models for ET₀ and K _y, respectively, with R ² of more than 0.970. By comparing six interpolation methods as well as two procedures, i.e. ‘calculate first, interpolate later’ (CI) and ‘interpolate first, calculate later’ (IC), IC-RK (residual kriging) was considered as an optimal method in interpolating K _y. Mapping of K _y for winter wheat indicated an increasing trend from the western and northern mountainous region to the eastern plain region in the basin, with the K _y of 0.783–1.668 for the dry growing season, 0.760–1.460 for the average growing season and 0.749–1.293 for the wet growing season. Moreover, the K _y values were more than 1.0 over the most of this basin, indicating that yield loss was more important than evapotranspiration deficit, and there were greater effect of water stress on the yield of winter wheat.