Impact of In-Situ Observation Sites Configuration on Spatial Interpolation: A Case Study on Air Temperature

Meteorological variables, such as air temperature, are fundamental for improving our understanding of weather and the energy process. Spatial continuous information about these variables, generally obtained from interpolation using limited ground observations, is often required. However, interpolation method as well as in-situ observational site density and the sites configuration will affect the interpolated results. Therefore, we evaluate the impact of different in-situ observational configurations under five scenarios on air temperature interpolation in a heterogeneous oasis in Zhangye City, Northwest China. The results show that at validation sites in different scenarios, the mean absolute errors between the interpolated and observed temperature values varied from 0.26 to 0.43 °C whereas pixel-by-pixel comparisons indicated an increase in difference, with R² values ranging from 0.04 to 0.71. The findings suggest that under a same observation density, the sites location and the sampling configuration significantly affects the interpolated air temperature.