Climate models are the best tools we have for estimating how the climate system will change in the future. However, computing resource limitations restrict the resolution of these models - so we can only make predictions about large-scale changes (e.g. continental scales) with any confidence. It would be far more useful to be able to make predictions at smaller scales (e.g. at the scale of a city).

To solve this issue, we can use a process called downscaling, where a high-resolution regional model is embedded in a low-resolution global model. Using this technique, we can obtain high-resolution data (albeit over a limited domain) using relatively modest computing resources. Most previous downscaling has been done using models of the atmosphere only. In this project, we have developed the first coupled ocean and atmosphere regional model for Australia. Using this, we can test the importance of air-sea interactions for producing regional projections.