ACCURACY EVALUATION OF MULTI-SOURCE SATELLITE PRECIPITATION DATA ALONG THE SICHUAN-TIBET RAILWAY IN CHINA

High spatiotemporal resolution (satellite precipitation data: SPD) are vital for railway planning and risk management, as well as advancing research in geoinformation science and meteorology.Four SPD (Climate Prediction Center Morphing technique: CMORPH, Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record: PERSIANN-CDR, Global Precipitation Measurement IMERG's third-level fusion product: GPM-EARLY, GPM-LATE) were compared with ground-based observations along the (Sichuan-Tibet Railway: STR) in China.Through quantitative and categorical evaluations across different temporal scales, and validation using extreme precipitation indices, the study finds that GPM generally outperforms other datasets, though PERSIANN-CDR exhibits the largest errors.The SPD show the highest correlation on the monthly scale, with significant variability at the annual scale.Errors in PERSIANN-CDR and CMORPH increase with rising monthly precipitation, peaking in July, while GPM demonstrates greater stability.Although PERSIANN-CDR captures precipitation events well, it has a high false alarm rate, whereas GPM-LATE achieves better skill scores.SPD retrieval capabilities decline with increasing precipitation, particularly above 25 mm/day.Spatial analysis shows better data quality in the eastern and southern regions, with GPM performing better.CMORPH accurately captures extreme precipitation indices, followed by PERSIANN-CDR, while GPM tends to overestimate extreme precipitation.These findings are crucial for improving precipitation estimation accuracy and enhancing water resource management and disaster warning systems along the STR.