Xiaoli Su, C. K. Shum, Zhicai Luo

1 Institute of Geophysics, School of Physics, Huazhong University of Science and Technology, Wuhan 430074 China

2 School of Earth Sciences, The Ohio State University, Columbus, OH 43210 USA

Abstract: Predicting and monitoring the spatiotemporal characteristics of heavy rain events are important to hazard pre- paredness, mitigation efforts, and local water resource manage- ment. Using three data sets, namely, the daily rain product from the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) version 04 Final Run, the daily output from European Centre for Medium-Range Weather Forecasts reanalysis data Interim version (ERA-Interim), and the high- quality gauge-satellite merged precipitation product, the spa- tiotemporal patterns of three heavy rain events are investigated for the first time over China in 2016, with the objective of assessing the capability of IMERG product for monitoring heavy rain events. It is found that the daily IMERG Final Run can better capture the spatial and temporal characteristics of heavy rain compared with that from ERA-Interim, but it significantly overestimates the amounts of the heaviest rainfalls by 11%–85% over the example regions. The comparison of regional averaged precipitation demonstrates that time series of precipitation retrieved by the IMERG algorithm agree well with that from gauge-satellite merged data set, with differences less than 10 mm on most days over each region. The statistic metrics demonstrate that the IMERG Final Run has a strong potential for detecting heavy rain events but with a relatively large error. This letter may provide useful feedback and insights for further improving the precipitation retrieving algorithm and the application of such data sets.

doi: https://doi.org/10.1109/LGRS.2018.2793897