Drought projections and associated uncertainties over the Arabian Peninsula from CMIP6 models

Abstract

Drought frequency and severity have increased over the water-stressed Arid regions. This research employs multiple CMIP6 global climate models (GCMs) for projecting droughts over the Arabian Peninsula (AP) until the end of the 21st century. We utilized the standardized precipitation index (SPI) and standardized precipitation evapotranspiration index (SPEI) to generate projected future statistics of droughts along with uncertainties assessment from inter-model spread, scenarios, timescale, and methods therein. For this purpose, after a meticulous analysis, we first identify the most suitable GCMs for better representation of AP's drought spatiotemporal pattern over the historical period (1985-2014). Our results indicate an increase in potential evapotranspiration (PET), which dominates simulated drought statistics relative to the precipitation. The projected evolution of the SPEI, which is derived from both precipitation and PET, indicates droughts  consistently increasing from low to high emission scenarios, In contrast, the SPI, owing to relatively-weaker amplification of the precipitation shows a moderately increasing wetness, except for a few northern regions where both indices evolve in agreement The fidelity of the simulated precipitation by many models over the historical period is also relatively poor compared to the PET, which may also be potentially adding to the uncertainties. In general, the principal sources of uncertainty in drought projections evolve from the choices of index, followed by scenarios, and inter-model variability, whereas methods and timescale mostly impact the magnitude of the trend in drought statistics.