Long-term statistical analysis of rainfall variability for climate adaptation planning in a semi-arid region

BACKGROUND AND OBJECTIVES: Climate variability and changing weather patterns pose increasing challenges to environmental sustainability and resource management. Understanding long-term rainfall trends at the semi-arid regional scale is crucial for developing effective adaptation strategies. This study evaluates precipitation records from ۱۹۸۰ to ۲۰۱۸ to address the limited analysis of rainfall variability in the districts of Madurai and Theni. It identifies temporal trends at seasonal, monthly, and annual scales to improve forecasting and support adaptive climate planning. Data from ۲۱ strategically placed rain gauge stations were analyzed using statistical methods, including the Mann-Kendall test and Spearman’s Rank Correlation, to detect shifts in rainfall behaviour and produce actionable insights for local decision-making and climate resilience.METHODS: Daily rainfall data from Madurai and Theni districts, located between ۹°۳۲’ and ۱۰°۱۸’ north latitude and ۷۷°۰۰’ and ۷۸°۳۰’ east longitude for Madurai, and between ۹°۵۳’ and ۱۰°۲۲’ north latitude and ۷۷°۱۷’ and ۷۷°۴۵’ east longitude for Theni, were analyzed over thirty-nine years from ۱۹۸۰ to ۲۰۱۸.  Following extensive quality checks, data from twenty-one sites were analyzed for homogeneity and serial correlation.  The Cumulative Sum Control Chart and the Modified Mann-Kendall Test, which were supported by the AnClim and trendchange software packages, were used to identify rainfall trends and shifts.FINDINGS: Rainfall analysis from ۱۹۸۰ to ۲۰۱۸ in Madurai and Theni found ۱۷ of ۳۵۷ series with serial correlation at ۹۰ percent confidence, notably at Tirumangalam, Usilampatti, Gudalore, and Periyakulam.  Annual rainfall increased by ۸.۵۶ millimeters per year at Veerapandi, ۷.۲۱ millimeters per year at Tiruppuvanam, ۵.۴۳ millimeters per year at Edayapatti, ۴.۸۷ millimeters per year at Andipatti, ۴.۵۲ millimeters per year at Bodi, and ۴.۱۱ millimeters per year at Uthamapalayam, but declined by ۶.۷۴ millimeters per year at Usilampatti. The corrected Mann-Kendall test showed the strongest trend at Veerapandi (۹.۰۲), with weaker trends at Tirumangalam and Usilampatti. Although there were some localized changes, no noteworthy turning points were discovered.CONCLUSION: This investigation revealed significant seasonal rainfall variability and trend shifts across Madurai and Theni districts. Serial correlation and season-specific trends were identified, with distinct change-points highlighting climatic sensitivity. The study addresses gaps in previous research by providing localized, season-specific insights through integrated trend and change-point analysis. Policy recommendations include promoting drought-resilient crops, enhancing rainwater harvesting, modernizing irrigation. Additionally, incorporating localized rainfall trends into district-level agricultural and water management planning can support adaptive responses to evolving rainfall patterns in southern Tamil Nadu.

Keywords:

Auto-correlation , Climate adaptation , Cumulative sum control chart , Rainfall Variability , Semi-arid regions