Harun Rashid: Drivers of projected Australian rainfall decline in ACCESS-ESM1.5 large ensembles
Rainfall is an important climate variable, which is challenging to accurately simulate in climate models. Many economic and societal sectors in Australia are highly sensitive to rainfall variability and long-term change. Therefore, it is crucial that climate models reliably represent both present-day rainfall and its future projections. However, rainfall projections in state-of-the-science climate models, such as those from the CMIP6 archive, shows substantial uncertainty.
ACCESS-ESM1.5 is among the best performing CMIP6 models for many important climate variables. Despite this, it projects one of the largest declines in Australian rainfall under the high-emission scenario SSP5-8.5 by the end of the 21st century. In this study, we investigate the drivers of this pronounced drying signal using ACCESS-ESM1.5 historical and SSP5-8.5 simulations, each with 40 ensemble members. We found that the projected drying is related to an eastward shift of tropical convection, from the Maritime Continent toward the eastern equatorial Pacific under future warming. This shift is associated with a strong El Niño-like pattern of mean state warming (relative to the tropical mean) simulated by ACCESS-ESM1.5, a feature also present in many CMIP6 models. We will discuss the physical mechanisms underlying this remotely driven drying over Australia, highlighting the role of large‑scale tropical circulation changes in shaping regional rainfall projections.