Temporal assessment of rainfall in the Itatiaia National Park, Brazil: an assimilation of orbital and surface data

Abstract Damage caused by extreme rainfall variability, including both excess and scarcity, poses significant environmental challenges. This study evaluated the performance of the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) using observations from an automatic weather station (ITWH1080) located in the lower sector of Itatiaia National Park, Brazil’s oldest Conservation Unit. Hourly data from the station (May 2021-December 2022) were aggregated to daily values and compared with CHIRPS data (2001–2022). Analyses were conducted for the rainy (October-March) and dry (April-September) seasons. CHIRPS showed strong agreement with ground observations (correlation > 0.96), confirming its reliability for seasonal rainfall assessment, although it underestimated precipitation during the rainy season and overestimated it during the dry season. The largest underestimation occurred on 16 March 2022 (-182 mm), while the greatest overestimation was observed on 5 April 2022 (+ 69 mm). A significant decreasing trend in precipitation was detected in July (Zmk = -2.4), indicating a progressive reduction in dry-season rainfall. Discrepancies between observed and estimated data are likely related to complex topography, dense forest cover, and persistent cloudiness. Overall, CHIRPS effectively captured rainfall seasonality and long-term variability, demonstrating its applicability for hydroclimatic analyses in protected areas of the Atlantic Forest biome.