Performance validation and bias correction of IMERG precipitation data under foggy meteorological conditions in Southeast Asia

Although the satellite precipitation products (SPPs) have been extensively utilized, their performance under fog-affected conditions has not been evaluated. To address this issue, we evaluated the SPPs and improved bias correction for foggy meteorological conditions in Southeast Asia. Three running versions of the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) were evaluated from 2014 to 2022 at the daily timescale. The assessment of the performance for Early-Run (ER), Late-Run (LR), and Final-Run (FR) was conducted based on statistical and categorical indicators. Then, the impact of foggy meteorological conditions on IMERG was analyzed. Finally, artificial neural networks (ANN) were employed to correct fog-affected estimates of precipitation. The main conclusions are: (1) The best performance is demonstrated by the FR, with a correlation coefficient (CC) of 0.443 and a probability of detection (POD) of 0.575. All running versions perform significantly better in the Indochinese Peninsula compared to the Southern Asia Archipelago. (2) Fog adversely affects the accuracy of IMERG precipitation estimates. In the FR, the presence of fog results in a decrease of 6.95 % in CC, and this effect intensifies during the rainy season, with fog reducing CC by 15.46 %. (3) Significant improvement is demonstrated by the ANN correction, enhancing CC by 27.46 % relative to the FR, with particular effectiveness for extreme precipitation overestimation. This study analyzed the performance of IMERG from the perspective of foggy meteorological conditions, providing a reference for researches under special weather conditions.