Evaluation of Terrain-Considered Spatial Interpolation Methods on Temperature and Precipitation in Complex Underlying Surface Region
To explore the applicability of terrain-considered spatial interpolation method in complex underlying surface region, the interpolation tests using regional automatic weather station observation data of monthly accumulated precipitation and average temperature in Chongqing in 2017 were designed by Cokriging, PRISM and IDW interpolation methods. The results showed that when a large number of samples are used, for average temperature, Cokriging and IDW have similar interpolation errors, which are lager than PRISM’s. For accumulated precipitation, two terrain-considered methods have similar errors and both are slightly lower than IDW. The accuracy of all three methods would decrease when using fewer observation samples, but PRISM has the lowest decrement rate in average temperature interpolation and highest decrement rate in precipitation interpolation. PRISM and Cokriging have the lowest errors for temperature and precipitation respectively when using small amounts of samples. Further analysis shows that PRISM can reduce the temperature interpolation error of complex terrain region significantly, but since the errors of accumulated precipitation are mainly based on precipitation intensity, the terrain-considered interpolation method may not have obvious advantage.