Hydroclimatic regime shifts and long-term inflow-outflow trends in a monsoon-dependent reservoir revealed through integrated trend and change point analysis

The study examines trend behaviour and change point behaviour of inflow and outflow pattern of the Sri Ram Sagar Project (SRSP) reservoir in Telangana, India. Based on an integrated analytical methodology that combines trend detection techniques with change point identification tests. The data is examined for 35 consecutive water-years from 1988 to 1989 to 2022–2023. Trend behaviour analysis was conducted using three non-parametric methods, which included Sen slope estimator, the Mann-Kendall test and the Innovative Trend Analysis (ITA) method. To determine the regime change and temporal variation of the hydrology of the reservoir, the change point detection analysis adopted four tests of homogeneity, namely, Pettitt’s test, SNHT, Range test and VNRT. The findings indicated that there was a statistically significant decreasing trend in monsoon inflows especially in June with a Sen slope of − 7.060 MCM/year and Probability value of 0.005. Whereas in monsoon season there was a significant reduction in Sen’s slope of − 90.176 MCM/year. The outflows in the reservoirs also showed considerable declining trend in post-monsoon months in November and December with Sen’s slopes of − 12.743 MCM/year and − 9.921 MCM/year respectively. These declines suggest changes in water conservation practices or shifting seasonal demands. The identification of fundamental regime changes was detected in terms of change point. The inflows in June were observed to be inhomogeneous in all the four experiments, with the change points taking place between 1992 and 2002. Annual inflows showed a dramatic drop of nearly 70% after 1990, falling from an average of 20,177 MCM to 6059 MCM. September outflows, on the other hand, experienced a significant growth (ITA slope = 87.005 MCM/year), which reflected that the enhanced flood-management plans are finding a balance between the risk mitigation and conservation objectives. Overall, the combined multi-test approach has indicated the necessity of integrating statistical models, as both trend-based and change-point-based methods prove to be ineffective in capturing hydrological non-stationarity. These results indicate that there is a great deal of hydroclimatic non-stationarity in SRSP. This underscores the need to adopt fair water management systems and climate-resistant reservoir management in monsoon-dependant regions.