Multiscale Comparative Evaluation of the GPM and TRMM Precipitation Products Against Ground Precipitation Observations Over Chinese Tibetan Plateau
Qiong Li,Jiahua Wei,Jianguo Yin,Zhen Qiao,Wang Peng,Haiyue Peng
DOI: https://doi.org/10.1109/jstars.2020.3047897
IF: 4.715
2021-01-01
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:This study aims to evaluate the performances of two latest released global precipitation measurement (GPM)-era satellite precipitation final run products (integrated merged multisatellite retrievals IMERG-V06B and Global Satellite Mapping of Precipitation GSMaP-V07) and one tropical rainfall measuring mission (TRMM)-era product (TMPA-3B42-V07) at hourly, daily, and monthly scale over the Tibetan plateau (TP), with special focus on the performances at different rain intensities, subbasins, and elevations at daily scales. Besides, hourly scale evaluations were carried out for one grid box value of IMERG-V06B and GSMaP-V07 versus 15 rain gauges on the rainy season, 2019. Results indicated that: First, GSMaP-V07 outperformed TMPA-3B42-V07 and IMERG-V06B with higher correlation coefficient and lower relative bias, mean absolute error, and root-mean-squared error values at daily and monthly scale. Second, satellite products showed overestimation during light rain and underestimation throughout moderate rain, heavy rain, and rainstorm; Third, satellite products have performed relatively better in lower elevation (<3000 m) regions, but they greatly overestimated the precipitation (more than 50%) at high-elevation (>4000 m). Fourth, satellite precipitation products performed better in the source region of the Yellow River, Yangtze River, Lancang River, and Nujiang River basins, whereas the products greatly overestimated the precipitation in the source region of the Yarlung zangbo River basin. In arid climate regions (Qiangtang Basin and Qaidam Basin), GSMaP-V07 had better performances than other products; and finally, although GSMaP-V07 showed higher overestimation than IMERG-V06B product at hourly scale, it has better consistency with rain-gauge observations. For satellite precipitation data application within the TP, we recommend the GSMaP-V07 over the IMERG-V06B. Furthermore, improvement on the mult-satellite rainfall retrieval algorithm is required.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geography, physical