Spatial heterogeneity of precipitation trends in the interior of northern Northeast Brazil: role of Atlantic and Pacific SST variability (1979–2020)

Abstract

Precipitation variability in the MATOPIBA region, located in northern Northeast Brazil, exhibits pronounced spatial heterogeneity and is strongly influenced by large-scale ocean–atmosphere interactions. This study investigates trends, non-stationarity, and ocean–atmosphere teleconnections of daily precipitation over the period 1979–2020 using CPC/NOAA rainfall data and sea surface temperature (SST) fields from the Tropical Atlantic and Equatorial Pacific. A cluster-based regionalization was applied to identify areas with homogeneous precipitation behavior, followed by trend detection, non-stationarity testing, correlation analyses with seasonal SST indices, and a lead–time assessment based on the Atlantic Multidecadal Oscillation (AMO). The results indicate a widespread predominance of negative precipitation trends across MATOPIBA, with marked spatial contrasts among sub-regions. Formal tests reveal significant non-stationary behavior, with a structural regime shift identified in 1995 that coincides with the transition of the AMO to its positive phase. Correlation analyses, evaluated using a Student’s t-test with effective degrees of freedom to account for serial correlation, reveal a robust and spatially coherent association between precipitation variability and North Tropical Atlantic SST, particularly during the austral autumn (MAM), with statistically significant negative correlations covering 19.1% of the region. In contrast, correlations with South Atlantic and Equatorial Pacific SSTs are weaker, spatially fragmented, and frequently not statistically significant. The lead–time analysis further indicates a statistically significant negative relationship between AMO anomalies during the pre–rainy season (October–November) and rainfall during the subsequent core rainy season (January–March), indicating a robust lead–lag statistical association that provides early diagnostic insights into seasonal rainfall variability. These findings highlight the dominant role of North Atlantic variability in modulating rainfall over the interior sector of northern Northeast Brazil and underscore the importance of accounting for spatial heterogeneity and regime shifts when interpreting regional climate signals, with implications for climate diagnostics and seasonal climate assessment in this agricultural frontier.