Multivariate Drought Characteristics and Joint Probability Analysis Under Climate Change: A Comparative Index-Based Study in Korea

This study analyses the impact of climate change on drought occurrence in the Doam Watershed using three distinct drought indices: precipitation, temperature and soil moisture. Historical data reveal shifting drought patterns influenced by changing hydrological conditions and precipitation variability. The Standardised Precipitation Index (SPI) and Standardised Precipitation Evapotranspiration Index (SPEI) indicate a trend towards more severe droughts in spring and winter, with less severity in summer. However, the Palmer Drought Severity Index (PDSI) shows increasing severity and longer duration for droughts from winter to spring. The periodicity of drought recurrence ranges from three to 6 years, with SPI3 indicating a short interval of approximately 1 year, SPI6 and SPI12 cycles of 4–6 years and PDSI a cycle of about 6 years. Future projections suggest that droughts will become more severe and persistent, particularly, under the SSP585 scenario. Joint return period analysis highlights an increased likelihood of future droughts, with shorter return periods for specified severity or duration compared to historical data. Drought severity data fit best with a log-normal distribution, while drought duration data follow an exponential distribution. Copula-based joint distribution functions were developed, with Frank and normal copula functions selected based on evaluation criteria. These models facilitated a deeper analysis of the interdependence between drought severity and duration. The findings emphasise the need for targeted measures to address annual winter and spring droughts, as well as large-scale, extreme droughts recurring approximately every 6 years.