Analysis of rainfall variability and spell dynamics in Coimbatore using copula function modelling

This study investigates the coupled dynamics of rainfall and evapotranspiration in Coimbatore, a major urban and agricultural hub in southern India, over 31 years (1994–2024). Aside from areas near the Noyyal River basin and episodic overflow from the Western Ghats, the region is predominantly dry, receiving approximately 665.5 mm of annual rainfall over about 45 rainy days and exhibiting a yearly average temperature of 25.4 °C. Advanced statistical techniques, including Gaussian, t, Frank, Roch-Alegre and BB5 copula-based models, are employed to analyse key hydroclimatic indicators: the conditional probability index (CPI), moisture availability index (MAI) and potential evapotranspiration (PET). These models facilitate forecasting of wet and dry spells and provide insights relevant to drought preparedness, sustainable water management and agricultural planning in monsoon-dependent regions. Rainfall patterns and drought risk are quantified using MATLAB, enabling discrimination between low- and high-rainfall trends that are critical for rainfed agriculture. Wet and dry spell dynamics are further evaluated using the Mann-Kendall and Spearman’s Rho non-parametric tests at the 5 % significance level, supporting the identification of atmospheric demand patterns and characteristic monsoon periods. The results reveal substantial variability in hydrological regimes, characterized by abrupt transitions between extreme wet and dry spells and highlight years of pronounced moisture deficiency alongside periods of excessive moisture. This comprehensive framework, which employs the Gaussian copula function, enhances understanding of climate variability and offers actionable strategies for managing hydroclimatic risks in vulnerable regions.