23STUDENT Enhancing Agricultural Drought Assessment in the Krishna Basin through Non-Stationary Standardized Soil Moisture Index (NSSMI)

The study introduces a novel approach to improve agricultural drought assessment in the Krishna River basin, India, through the development of a Non-Stationary Standardized Soil Moisture Index (NSSMI) incorporating time as a covariate. Overcoming limitations of conventional drought indices under the influence of global climate change, the NSSMI was calculated at a 3-month time scale using 40 years of monthly soil moisture data (1980-2019) from seven locations in the Krishna basin. A time-sliding window approach was applied to detect non-stationarity in the soil moisture data. Various probability distributions including gamma, Weibull, log normal, and gev were evaluated for fitting the data, with the Kolmogorov-Smirnov (K-S) test indicating the gev distribution as the most suitable. Trend analysis through the Mann-Kendall test revealed distinct patterns of negative, positive, or no trend in the shape (γ), scale (σ), and location (µ) parameters across the grids. Comparative analysis with the traditional Standardized Soil Moisture Index (SSMI) demonstrated the NSSMI's superior ability to capture drought characteristics in the Krishna River basin. This underscores its efficacy in robustly monitoring agricultural drought amidst a changing environment. Given the evolving climate scenario, these findings offer crucial insights for effective agricultural drought assessment. Future research endeavors will explore the potential inclusion of climatic and temperature variables as covariates in NSSMI computation, providing a comparative perspective on its performance against the present study's outcomes. This innovative approach holds promise for enhancing agricultural drought resilience in the face of ongoing climate challenges.