Comprehensive evaluation of mainstream gridded precipitation datasets in the cold season across the Tibetan Plateau

Abstract

The Tibetan Plateau (TP). This study evaluated the accuracy of six mainstream gridded precipitation products, including the Asian precipitation dataset by calibrating the GPM-era IMERG (AIMERG), Climate Hazards group InfraRed Precipitation with Stations (CHIRPS), China Meteorological Forcing Dataset (CMFD), ECMWF Re-Analysis version 5 (ERA5-Land), Integrated Multi-satellite Retrievals for Global Precipitation Measurement (IMERG) and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Cloud Classification System-Climate Data Record (PERSIANN-CCS-CDR), in the cold season across the TP. The observations and these six precipitation products all revealed high precipitation occurs in the eastern and southeastern TP, and low precipitation occurs in the Qaidam Basin. But in the western regions with sparse gauge networks, the spatial patterns revealed by these six products were different. The precipitation occurrences in the cold season detected by CMFD had the highest accuracy, followed by AIMERG and IMERG, while CHIRPS had the worst accuracy. CMFD, IMERG and AIMERG outperformed CHIRPS, ERA5-Land and PERSIANN-CCS-CDR against observations at the annual and monthly scale, ERA5-Land and PERSIANN-CCS-CDR overestimated while AIMERG, CHIRPS, CMFD and IMERG underestimated cold season precipitation. CMFD had the highest accuracy when using the total adjusted observations as a reference at monthly scale, while IMERG had the highest accuracy when only using the adjusted precipitation by Geonor gauge as a reference.