A51K-0214
First Evaluation of Day-1 IMERG products using the best-available hourly rain gauge network over the Tibetan Plateau

Friday, 18 December 2015
Poster Hall (Moscone South)
Yingzhao Ma1, Guoqiang Tang1, Di Long1, Lingzhi Zhong2, Yan Shen3, Wei Wan1, Yang Hong4 and HydroSky, (1)Tsinghua University, Beijing, China, (2)Chinese Academy of Meteorological Sciences, Beijing, China, (3)National Meteorological Information Center of China, Beijing, China, (4)University of Oklahoma Norman Campus, Norman, OK, United States
Abstract:
Evaluation of the newly launched satellite precipitation programme - Global Precipitation Measurement (GPM) mission in high altitude and complex terrain regions is vital for assessing and then improving the truly global precipitation retrieval algorithm. In this study, the performance of Integrated Multi-satellitE Retrievals for GPM (IMERG) rainfall products was evaluated against hourly rain gauge network in warm season over the Tibetan Plateau (TP). The study revealed that IMERG successfully captured the spatial pattern of gauged-based rainfall values in warm period at both plateau and river basin scales. Over the whole plateau, the south and center part primarily suffered from positive bias, while negative bias was mainly detected in the east and north. Although the correlation between IMERG and gauge data was a litter scattered spatially, their correlation coefficients increased gradually from 0.69 to 0.73 among hourly to daily periods. River-basin uncertainties were characterized by an underestimate in the Yellow river basin (-6.7%), and an overestimate both in the Brahumaputra (+18.4%) and Yangtze river (+9.6%) basins. In addition, there was generally an overestimate within the occurrence of heavy rainfall events, particularly at higher elevation sites. The probability of detection (POD) exhibited an appreciable elevational dependence, trending toward higher values as elevation increased. Also, the IMERG rainfall product reasonably well detected the onset, mature, and end of the Indian monsoon over the TP, where some moisture surge events were not detected by sparse rain gauge observations. Overall, this study provides meaningful guidance about the error characteristics associated with IMERG rainfall estimates and potential values for IMERG algorithm improvements over the complex and high TP area.