A detailed framework for the characterization of rainfall climatology in semiarid watersheds

The identification of spatial and temporal rainfall climatology patterns is crucial for hydrometeorological studies over semiarid watersheds, which frequently face water distribution conflicts and socioeconomic issues due to water scarcity. Thus, the objective of this study was to propose a comprehensive approach for the characterization of rainfall climatology over semiarid watersheds. Monthly rainfall time series (1962–2015) with up to 30% of gaps measured in 56 rain gauges in the Piranhas-Açu Watershed—Brazilian semiarid region—were used. Data gaps were filled through a combination of simple spatial interpolation techniques. Principal component analysis and cluster analysis identified two homogeneous rainfall subregions in the basin: C1, in the upper portion, and C2, in the middle and lower portions. Rainfall volumes in C2 were up to 23.5% smaller than those in C1, due to orographic structures which contribute to aridity in this region. Rainfall anomalies were calculated in each cluster through the modified Rainfall Anomaly Index (mRAI) and were associated with the phases of the El Niño–Southern Oscillation (ENSO) and the Atlantic Meridional Mode (AMM). In years when the ENSO (AMM) was in its positive (negative) phase, there was a higher probability of occurrence of months with above-average rainfall, while the opposite was also true. Results showed that the effects of the patterns are mutually influenced, which has been previously found at larger scales. Finally, mRAI trends were identified through the Mann-Kendall test, which indicated significant negative trends in C1 and C2, especially during the wet season