Accurate regional precipitation observations are crucial for hydrological and meteorological assessments, particularly amid climate change. However, ground stations struggle to capture precipitation variability, especially in regions with sparse networks. This study evaluates the performance of two satellite precipitation products (SPPs), GPM-IMERG V06 and PERSIANN-CDR, in the semi-arid and urbanized District 1 of Tehran, Iran, in the period from 2013 to 2023. Quantile mapping (QM) was applied for statistical downscaling to improve precipitation estimates at different time scales. The corrected SPP data were compared to synoptic station observations using Taylor diagrams, quantile-quantile plots, and intensity–duration–frequency (IDF) curves. Results indicate that GPM-IMERG outperforms PERSIANN-CDR for most metrics and temporal scales, particularly for short-term precipitation events, making it more suitable for flood warnings. However, both SPPs struggle with extreme precipitation estimation, even after bias correction. QM significantly improves alignment with ground-based observations, reducing systematic errors and enhancing correlation coefficient (CC), root-mean-square error (RMSE), and relative bias (RB) at daily, monthly, and seasonal scales. IDF analysis confirms that bias correction enhances the reliability of SPPs for estimating extreme precipitation, especially for shorter return periods (2–10 years). However, deviations persist for longer return periods (50–100 years), highlighting the limitations of satellite-derived products in capturing extreme events. This study underscores the potential of bias-corrected satellite precipitation products for hydrological applications while emphasizing the need for further refinement to enhance accuracy in estimating extreme precipitation in semi-arid urban regions.
Heir et al. (2025) studied this question.