Rainfall Variability Analysis Using Rolling Statistics in Chiang Rai Province

This study analyzed long-term rainfall variability in Chiang Rai Province (1981–2024) using rolling statistics and the non-parametric Mann–Kendall test to detect temporal changes in both the mean and variability of annual rainfall. Annual rainfall data from five meteorological locations were examined using 3-, 5-, 7-, and 12-year moving windows to characterize short-, medium-, and decadal-scale fluctuations. Although previous studies in northern Thailand have examined trends in total or extreme rainfall, multi-scale variability has not been systematically assessed, leaving uncertainty about how rainfall behavior is changing across different temporal windows. Results indicate that mean annual rainfall remains statistically stable (p > 0.05) at most locations, except for Wiang Pa Pao, which shows a significant upward trend (Z = 3.79–5.35). In contrast, the rolling standard deviation increased consistently across all locations, suggesting intensifying interannual variability. These findings indicate that rainfall in northern Thailand has become more unpredictable, with larger departures from the mean despite relatively stable long-term averages. The results point to practical needs for updating design-rainfall criteria and integrating variability-based assessments into regional water-resource and climate-adaptation planning.