Evaluation of CMIP6 models for simulations of diurnal temperature range over Africa

The variability in the diurnal temperature range (DTR), an indicator of climate change, remains limited, especially over Africa, due to the scarcity of observed maximum and minimum temperature data. This work investigates the ability of the Coupled Model Intercomparison Project (CMIP6) to simulate DTR over Africa for the period 1980-2014. Datasets from the Climatic Research Unit (CRU TS4.05) and National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Centre (CPC) gridded temperature datasets are utilized as observed data. Similar to the high variability in topography and climate across the continent, the DTR exhibits high heterogeneity over Africa. The Sahara and its environs record the highest DTR, while Central Africa and coastal areas experience the least, given the thermal inertia of water bodies. CMIP6 models overestimate and underestimate DTR over different parts of the continent. Moreover, the multi-model ensemble mean of CMIP6 models shows significant decreasing trends both in seasonal and annual trends. Overall, five CMIP6 models such as EC-Earth3, ACCESS-CM2, BCC-CSM2-MR, EC-Earth3-veg, and IPSL-CM6A-LR show robust skill scores (0.48-0.54). The findings form the basis for investigating the role of temperature extremes on DTR. Further, the variability in DTR across parts of the continent prompts the need for future assessments to investigate future changes in DTR.