A multivariate approach to drought monitoring: Improving robustness and accuracy through a new drought index in regions with high climate variability, applied to the drought-prone region of Ethiopia

Abstract

Study focus: Assessing, monitoring, and quantifying drought characteristics to develop early warning systems is crucial for identifying the spatial extent and severity of droughts at regional and local scales especially in regions of vulnerable societies relying on local agriculture. Observations and reanalysis from 1981 to 2022 are analyzed for spatiotemporal droughts in Ethiopia. While standard drought indices like Standardized Precipitation Index and Standardized Soil Moisture Index are based solely on precipitation or soil moisture, a new drought index based on precipitation, potential evaporation, surface temperature, soil temperature, and soil moisture is developed, making the index more robust to climate and land use changes. This new Multivariate Standardized Drought Index (MvrSDI) is evaluated focusing on the severity and duration of 2015 and 2022 droughts in Ethiopia. Results show that spatiotemporal comparisons of MvrSDI at 3-, 6-, and 12-month time scales detect drought severity and duration in each drought-prone region of Ethiopia. Further,Mann-Kendall statistic test identifiy a drought trend between 1981 and 2022 an increasing drought severity. New hydrological insight for the region: The MvrSDI effectively assesses and monitors drought impacts on agriculture, proving beneficial for stakeholders focused on environmental sustainability and food security. Its multivariate character makes MvrSDI more robust and therefore a valuable tool for drought monitoring and decision-making in regions with high climate variability and land use changes in drought-prone regions like Ethiopia.