An applicability evaluation of version 05 IMERG precipitation products over a coastal basin located in the tropics with hilly and karst combined Landform, China

ABSTRACT

Version 05 of the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) includes IMERG-E, IMERG-L, and IMERG-F with spatial resolution of 0.1° × 0.1°. Evaluation over a hilly and Karst landform combined basin in the tropics is necessary to be done to verify the acceptability of the use of IMERG V05. In this study, IMERG Data (ID) are compared with Ground Precipitation Data (GPD) from 15 stations over the Nanliujiang River Basin in China. The comparison is quantified using some statistical indices to analysis detection capability, correlation, and error evaluation. The results show that IMERG-F performs better than IMERG-E and IMERG-L. All three products can detect precipitation under different rainfall intensity effectively. A positive relationship between Pearson’s correlation coefficient (r) and temporal scales has been found within both single station and whole basin scales. At a single station scale, r of 14 stations are greater than 0.6 (daily scale) and 0.8 (monthly scale), Relative Bias (RB) ranges from 0.458 to 0.013, and the value of Root-Mean Square Error (RMSE) on daily scale and monthly scale is ranging from 7.38 to 12.51 mm and from 40.17 to 69.01 mm, respectively. At the whole basin scale, products got r values in daily, monthly, and seasonal scale within the range of 0.698 to 0.741, 0.867 to 0.918 and 0.940 to 0.950, respectively. RB in summer is smaller than 0.35, and a more underestimation of precipitation occurs during the winter season of the Northern Hemisphere. The values of monthly RMSE are distributed from 2.841 to 18.949 mm, while the seasonal RMSE has a higher value in summer. Conclusively, IMERG V05 can be considered as an auxiliary tool for precipitation observation in the regions with similar characteristics.

Acknowledgements

This study was jointly supported by the National Natural Science Foundation of China (Grant Number 41807197), the Natural Science Foundation of Guangxi Province (Grant Numbers 2018GXNSFAA138042 and GuikeAB7195073). Thanks for the Editors of the IJRS and reviewers helping us revise the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [41807197]; Natural Science Foundation of Guangxi Province [2018GXNSFAA138042] [GuikeAB17195073].