Assessment of Tropical Cyclones Over the Australian Region From Downscaled CORDEX‐CMIP6 Simulations

Abstract

Tropical cyclones (TCs) cause major economic losses in Australia due to severe winds and floods; however global climate models (GCMs) are too coarse to properly resolve them. Here, we dynamically downscale ERA5 reanalysis and 16 CMIP6 simulations from 11 GCMs using the Conformal Cubic Atmospheric Model (CCAM) to a ∼10 km resolution over the Australian region. We evaluate whether these simulations can represent TC activity over the Australian region by analyzing TC frequency, genesis, intensity, lifespan, and the role of ocean–atmosphere coupling. The downscaled simulations reproduce observed spatial and seasonal patterns of TC activity, with the highest skill in the central and eastern subregions. CCAM‐ERA5 captures the observed intensity distribution for Categories 2–4 and the relationship between lifespan and intensity. The CCAM‐CMIP6 ensemble reproduces mean TC number well in the central and eastern subregions but underestimates the most intense (Categories 4–5) and medium‐lived TCs (6–8 days) while overestimating Category 1 systems. Ocean coupling has model‐dependent effects rather than producing a systematic improvement. Coupled configurations generally simulate fewer TCs than atmosphere‐only configurations, mainly through reduced weak‐TC numbers, while differences in stronger‐TC intensity are generally small and metric dependent. Overall, the CCAM downscaled simulations provide a realistic representation of Australian TC climatology, particularly in the central and eastern subregions, but remaining biases in the western basin and in the upper tail of TC intensity should be considered in future projection and impact applications.