Main content area

Evaluation and comparison of CHIRPS and MSWEP daily-precipitation products in the Qinghai-Tibet Plateau during the period of 1981–2015

Liu, Jun, Shangguan, Donghui, Liu, Shiyin, Ding, Yongjian, Wang, Shaoping, Wang, Xiuna
Atmospheric research 2019 pp. 104634
altitude, climate, ecology, hydrometeorology, landscapes, peroxidase, probability, rain, rain gauges, satellites, summer, time series analysis, winter, China
High-resolution, long-term and accurate daily-precipitation is always difficult and rarely measured in the Qinghai-Tibet Plateau (QTP) because of the high altitude and complex terrain. The accuracy of satellite-based gridded precipitation products have been continuously improved recently which is crucial to the study of cryosphere ecology and environment. The goal of this study is to evaluate the accuracy of CHIRPS v2 (Climate Hazards Group Infrared Precipitation with Stations data, version 2) and MSWEP v2 (Multi-source weighted-Ensemble Precipitation, version 2) daily-precipitation products overthe QTP during the period 1981–2015. Validation was done using a time series of daily-precipitation data obtained from 104 hydrometeorological stations distributed over the QTP. Error metrics (The correlation coefficient CC, the relative bias BIAS, and root mean square error RMSE) were used for accuracy evaluation and detectability indicators (probability of detection POD, false-alarm ratio RFA, and critical success index CSI) were used for the analysis of detection capabilities of rainfall occurrence events. The results indicate that when compared to rain gauge observations, CHIRPS and MSWEP daily-precipitation products represent well the spatial and temporal distribution of the mean daily precipitation over the QTP, while both of them overestimate the daily-precipitation (0.18 mm/d for CHIRPS, 0.56 mm/d for MSWEP). MSWEP performed better than CHIRPS according to CC (MSWEP is 0.44, CHIRPS is 0.23) and RMSE (MSWEP is 4,21 mm, CHIRPS is 5.03 mm) and MSWEP showed better detection capabilities with higher POD (0.65), lower RFA (0.50) and CSI (0.39) in the QTP. Both products are less accurate in dry conditions (the north QTP, winter) than in moist conditions (the south QTP, summer). Light precipitation events (0–2 mm/d) are underestimated but heavy precipitation events (2–25 mm/d) are overestimated. CHIRPS and MSWEP have shown great potential to be able to be applied to the precipitation-related study of the QTP. Although the accuracy of MSWEP is higher than that of CHIRPS, the latter has higher spatial resolution (o.o5°) and is more suitable for small-scale studies.