Modeling of short-term meteorological drought under changing climate in Gujarat, India

Abstract

Climate change exacerbates droughts globally, challenging traditional stationary methods in identifying natural hazards like droughts. This study examines short-term drought characteristics over Gujarat using three indices: Standardised Precipitation Evapotranspiration Index (SPEI), Standardised Precipitation Index (SPI), and Reconnaissance Drought Index (RDI). Non-stationary models were developed using the Generalised Additive Models for Location, Scale, and Shape (GAMLSS) framework, incorporating local and large-scale climatic covariates and time across 11,700 model combinations. Drought patterns were analysed at fortnightly, monthly, and three-month timescales over 60 grid points. Results indicate that non-stationary droughts are more frequent, severe, and localised than stationary cases. The agreement between stationary and non-stationary methods improves with longer timescales. Among local covariates, diurnal temperature range (DTR) has the highest impact on drought characteristics across all indices and timescales. The Kutch region experiences the highest drought frequency with significantly reduced rainfall, with index values often below ? 1. This study highlights the critical role of non-stationary methods in accurately assessing drought patterns in climatically variable regions like Gujarat. The findings emphasise the need for advanced statistical approaches to enhance drought management and mitigation strategies by incorporating local and large-scale climate influences. © The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2025.