Spatial-temporal accuracy assessment of TRMM, CHIRPS, PERSIANN and GPM satellites precipitation estimation products over the complex Lebanese territory

The accuracy of satellite precipitation estimates (SPEs), which are prospective substitutes for gauge observations taken on the ground, is essential for hydrological modelling and water resources management. However, the efficient use of SPEs is difficult, particularly in regions with limited ground-based networks, high climatic unpredictability and harsh topography. The purpose of this study is to evaluate the performance accuracy of four popular operational SPEs: the Tropical Rainfall Measuring Mission (TRMM), the Climate Hazards Group Infra-Red Precipitation with Station (CHIRPS), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN) and Global Precipitation Measurement (GPM) over Lebanon, a country of the Eastern Mediterranean. The performance accuracy was spatially and temporally assessed against ground-based observations from 12-gauge stations at daily and monthly scales during 2000–2020 using several statistical parameters. Overall, the results reveal a higher accuracy with the monthly timescale datasets of the four SPEs as compared to daily time scale, indicating lower reliability. CHIRPS outperforms the other products in providing accurate monthly SPEs across diverse landscapes, especially in the rainy season. However, when it comes to coastal areas with abundant humidity, GPM stands out as the top performer. Moreover, Lebanon's complex topography is indicated to significantly affect SPEs accuracy by humidity, altitude and proximity to mountain ranges.