Multi-source Precipitation Data Fusion Method Based on Geostatistical FLTERSIM Algorithm

Authors

  • Xiaoru Cong, Aifeng Lv, Chunlin Xia

Abstract

Based on rain gauge-measured precipitation data of 17 national reference stations and basic stations in Qaidam Basin and surrounding area within one kilometer distance, as well as the 3B43 remote measurement data of Tropical Rainfall Measuring Mission (TRMM), the multi-point geostatistical FILTERSIM algorithm is used for fusion, while statistics such as Mean Absolute Error (MAD), Root Mean Square Error (RMSE) and Correlation Coefficient (COR) are used to test and analyze the fusion results. Fusion accuracy is evaluated by comparison with classical Bayesian fusion method and Co-Kriging method. The results show that: on a monthly scale, FILTERSIM algorithm has significantly superior precipitation fusion accuracy in Qaidam area than traditional Co-Kriging and classic Bayesian, which is more suitable for precipitation data fusion of special terrain.

Published

2020-04-30

Issue

Section

Articles