clustering, using IMD gridded daily data. Monthly rainfall and extreme rainfall indices were analysed across multiple periods, and the reliability of Innovative Trend Analysis (ITA) was evaluated against the modified Mann-Kendall test. Benchmarking revealed a 93.1% false discovery rate for ITA, demonstrating its limited reliability for rainfall trend detection and underscoring the need for statistically robust methods. Structural break analysis identified no abrupt change point in mean rainfall, but extreme indices exhibited significant distributional differences across the 1960 boundary. Monthly trends showed a clear temporal reversal, with widespread increases during 1961-1990 followed by predominantly decreasing trends in 1991-2020. Extreme indices displayed a higher significance rate than mean rainfall, with declining heavy-rainfall frequency contrasting with consistently increasing intensity. Frequency trends remained monsoon-driven, whereas peak intensity extended into the pre-monsoon season. Teleconnection signals were weak and spatially limited, indicating that regional monsoon dynamics, rather than large-scale ocean-atmosphere forcing, primarily govern rainfall variability in Odisha.
Malayalam et al. (Mon,) studied this question.