Precipitation Variability for Protected Areas of Primary Forest and Pastureland in Southwestern Amazônia

Daily and monthly rainfall data provided by surface rain gauges in the Amazon Basin are sparse defective, making it difficult to monitor patterns for certain portions of its territory, this sense, estimations precipitation from remote sensing calibrated with gauge key overcome problem. This paper presents a spatiotemporal analysis distribution Rondônia State, southwestern Amazonia. Data Climate Hazards Group InfraRed Precipitation Station (CHIRPS) were analyzed, using pooled time forty-year period (1981–2020). obtained validated distributed over study region. Pixel-by-pixel trend analyzes developed applying Mann-Kendall test Sen’s slope magnitude trend. The revealed that CHIRPS tendency underestimate values most cases. Among metrics, mean between very good (<±15%) (±15–±35%) observed PBIAS; RMSE range 57.8 mm 107.9 mm; an average agreement level 0.9 SES 0.5; fit linear regression model (average R2 > 0.70) about 64.7% stations. Sen’ spatialization results show reduction approximately −15 year−1, decrease mainly Northern Region Rondônia, which has extensive areas where native forest been replaced pasture.