Climate change impacts and uncertainty on spatiotemporal variations of drought indices for an irrigated catchment

Abstract

Droughts are responsible for severe vegetation loss and declining agricultural yields. As future climate change projections imply an increased risk of extreme events, the occurrence of droughts is potentially accelerating in the future. The influence of climate change on drought events in a Danish agricultural catchment under the emission scenario RCP8.5 are evaluated by three different drought indices covering soil moisture, groundwater and streamflow deficits. The indices are based on results from a hydrological model forced by downscaled climate outputs from 16 Euro-CORDEX climate models (GCM-RCMs), while considering uncertainties among climate model projections. The hydrological model demonstrated a satisfactory ability in modelling historical drought characteristics. The results from the future projections showed that the intensity and frequency of droughts increased towards the end of the century. The spatial patterns of changes in drought were found to be highly dependent on the climate model results, index formulations and assumptions, as well as the hydrogeological properties of the catchment. Groundwater based irrigation in the agricultural areas effectively mitigates soil moisture drought, leading to lower future uncertainty of the ensemble mean (higher model agreement) for soil moisture droughts at irrigated locations. This is obtained at the cost of lower groundwater levels in both reference and future periods resulting in larger uncertainties on the simulated groundwater droughts, because of the addition of irrigation uncertainty. While the joint impact of greater groundwater abstraction and changing dynamics of precipitation and groundwater recharge also leads to larger streamflow variability between model combinations (standard deviation) under RCP8.5 for the downstream discharge station.