Design of climate station network in mountain catchments

Abstract

In the Jizera Mountains (Czech Republic) the density of climate station network was tested in relation to spatial data interpolation, and watershed management targets. Point weather data (precipitation, air temperature, humidity and wind velocity) were interpolated by the nearest neighbourhood (NN), inverse distance weighting (IDW), spline (SPL), hypsometric (HYP) and kriging (KRI) methods. The results were assessed by the root mean square error (RMSE). The interpolation effectiveness showed the following order: HYP, IDW, KRI, NN and SPL. The advantage of the hypsometric method was recognised, particularly, by providing reasonable outputs in marginal catchments of the region and outside of the main instrumented area. However, in case of a higher density of observation points (11 hectares per station), all interpolation methods manifested comparable and realistic outputs in the focused mountain watersheds.