Detecting monotonic and non-monotonic rainfall trends across Pernambuco using innovative trend analysis and Mann-Kendall test

Abstract Annual precipitation can change differently across low, medium, and high annual totals, yet this non-monotonic behavior has not been mapped across Pernambuco’s three climatic regions. We analyzed annual precipitation from 820 BR-DWGD grid cells for 1961–2022 using Innovative Trend Analysis (ITA), the Mann-Kendall (MK) test, and Sen’s slope. For regional mean series, ITA slopes were − 2.34, − 0.07, and − 6.28 mm yr⁻¹ for Sertão, Agreste, and Zona da Mata, respectively; ITA detected significant negative trends in Sertão and Zona da Mata. The corresponding Sen’s slopes were − 1.48, − 0.09, and − 3.82 mm yr⁻¹, with a significant MK trend only in Zona da Mata. At the grid-cell scale, ITA identified significant negative and positive trends in 76% and 17% of cells, compared with 26% and 4% for MK. ITA further showed that low annual totals declined across the state, whereas high totals increased in the Agreste and northern Zona da Mata; 30% of cells had non-monotonic trends. These results show that monotonic tests alone can miss distribution-specific changes relevant to drought assessment, water management, and rainfed agriculture.