Abstract The main objective of this study is to identify whether there are significant trends or changes in precipitation on seasonal, decadal, annual, and Climatological Normal (CN) scales in a tropical environment with climatic contrasts (seasonally dry and seasonally wet). To this end, we used monthly precipitation series (1961 to 2022) for the state of Alagoas, Northeast Brazil. Statistical analyses were applied (cluster analysis, Mann-Kendall tests, Sen’s slope, Pettitt, Wilcoxon, and Kruskal-Wallis tests) to identify pluviometrically homogeneous sub-regions and detect significant trends/changes in precipitation patterns (p-value < 0.05). It was found that the highest average precipitation occurs on the coast, where it rains three times more than in the Sertão region (semi-arid). Four sub-regions were identified and named, from east to west, as Sub-region 1 (S1), 2 (S2), 3 (S3), and 4 (S4). The wettest season (April to July) and the driest season (October to December) coincide temporally, but diverge in rainfall amounts. In 38.2% of the municipalities, significant differences were observed between the NC, meaning that the most recent normal (1991–2020) was drier. One-third of the municipalities showed significant differences in at least two decades, and the 1990s and 2010s were the driest decades in the entire state. In annual terms, significant trends (p-value < 0.05) were identified in 26 municipalities, with 5 increasing and 21 decreasing, notably S4 (least rainy), in which eight municipalities showed decreasing precipitation trends. It can be concluded that the driest tropical region of the state of Alagoas is experiencing decreasing rainfall, which may favor the emergence of future arid areas. Graphical Abstract This graph presents a concise overview of the main findings and methodologies of this work. Located in the Northeast region of Brazil (Tropical region), the state of Alagoas has 102 municipalities with climatic contrasts (humid in the east and semi-arid in the west). We analyzed, on a monthly scale (1961 to 2022), for 102 municipalities, the impacts of climate change on rainfall variability at various temporal scales (monthly, seasonal, annual, decadal, and climatological normals). Initially, we regionalized areas with homogeneous rainfall characteristics (Cluster Analysis) and identified 4 distinct sub-regions, which were supported by the application of the Silhouette Index. The sub-regions showed the most marked differences in quantity and distribution, with a notable decrease in rainfall from east (humid) to west (semi-arid); however, they share similarities regarding the timing of the rainy season. In the analysis of annual rainfall series, decreasing trends predominated. The application of the triad of non-parametric statistical trend tests (Mann-Kendall, Sen, and Pettitt) supports the results with 95% statistical confidence. With the same level of confidence, we used tests to compare precipitation by decades (Kruskal-Wallis) and by climatological normals (Wilcoxon). This allowed us to affirm that, on average, not all areas/municipalities show a significant trend, nor was the sign of the trend uniform. However, there was a predominance of decreased rainfall in the state, especially in a region that is already semi-arid. This reinforces the need for strategies for decision-making planning, implementation of water infrastructure, and socioeconomic adaptation to local and regional climate change, particularly in a vulnerable region like Northeast Brazil.
Júnior et al. (Tue,) studied this question.