Can Satellite Products Recognise Extreme Precipitation Over Southeastern South America?

Abstract

Extreme precipitation events (EPEs) are becoming increasingly frequent and intense in southeastern South America (SESA). The limited rain gauge network in SESA could be overcome using satellite‐based synthetic precipitation data. This study analyses the capability of satellite products IMERG Final Run V06, PERSIANN, PERSIANN CCS‐CDR and PDIR‐NOW in capturing extreme precipitation characteristics over SESA in the 2001–2020 period. EPEs were characterised by annual maximum values, maximum monthly values, and the 95th and 99th percentiles of precipitation time series. Statistical metrics were applied to evaluate the efficiency of satellite products in representing EPEs compared to observational data. Extreme events characterised by the number of very wet days (R95p), extremely wet days (R99p), and the simple daily intensity index (SDII) were also evaluated. Our results suggest that IMERG and PERSIANN CCS‐CDR accurately represent the annual maximum precipitation averages and provide the best estimates of the maximum precipitation and the average number of events across various precipitation thresholds. IMERG exhibits the lowest BIAS and RMSE for the 95th percentile and performs well in representing R95p and R99p indices. IMERG also accurately represents the average number of events across various precipitation thresholds, although it overestimates precipitation at the 0.1–5 mm threshold. In contrast, uncalibrated products like PERSIANN and PDIR‐NOW exhibit less consistent performance, often underestimating lower‐intensity events ( 50 mm). PERSIANN tends to overestimate SDII values and displays higher error rates for the 95th percentile, while PDIR‐NOW overestimates R95p and R99p indices and estimates SDII with poor performance. Although there are challenges in high‐altitude areas and coastal regions, IMERG and PERSIANN CCS‐CDR show promise in detecting extreme events, particularly for precipitation thresholds above 100 mm. Our findings provide a basis for developing Intensity‐Duration‐Frequency (IDF) curves, essential for hydrological planning, in future work using combined satellite datasets.

Bibliographic Information