Uncertainty Estimation of Global Precipitation Measurement through Objective Validation Strategy

Abstract:

Since Tropical Rainfall Measuring Mission (TRMM) has been launched in 1997 as the first satellite mission dedicated to measuring precipitation, the spatiotemporal gaps of precipitation observation have been filled significantly. On February 27^th, 2014, Dual-frequency Precipitation Radar (DPR) satellite has been launched as a core observatory of Global Precipitation Measurement (GPM), an international multi-satellite mission aiming to provide the global three hourly map of rainfall and snowfall. In addition to Ku-band, Ka-band radar is newly equipped, and their combination is expected to introduce higher precision than the precipitation measurement of TRMM/PR. In this study, the GPM level-2 orbit products are evaluated comparing to various precipitation observations which include TRMM/PR, in-situ data, and ground radar. In the preliminary validation over intercross orbits of DPR and TRMM, Ku-band measurements in both satellites shows very close spatial pattern and intensity, and the DPR is capable to capture broader range of precipitation intensity than of the TRMM. Furthermore, we suggest a validation strategy based on ‘objective classification’ of background atmospheric mechanisms. The Japanese 55-year Reanalysis (JRA-55) and auxiliary datasets (e.g., tropical cyclone best track) is used to objectively determine the types of precipitation. Uncertainty of abovementioned precipitation products is quantified as their relative differences and characterized for different precipitation mechanism. Also, it is discussed how the uncertainty affects the synthesis of TRMM and GPM for a long-term satellite precipitation observation records which is internally consistent.