August 6, 2020 | Published by |

Supervisors: Dr Raktima Dey (raktima.dey@anu.edu.au), Prof Julie Arblaster (julie.arblaster@monash.edu)

Definition of extreme rainfall timing: The month in which extreme rainfall occurs. There are many ways of defining extreme rainfall. The maximum five-day consecutive rainfall or other such indices can be used to study extremes in this project.

Australia experiences some of the world’s most variable rainfall. Previous studies have mostly focused on understanding rainfall variability in terms of frequency and intensity. However, understanding the timing of when extreme rainfall occurs is crucial for seasonal prediction, although it largely remains unexplored. In a previous study led by Dey et al. (2020; submitted) shows that north and south Australia has very distinct seasonality in the timing of extreme rainfall, whereas, in the north extreme rainfall occurs in the summer, and in the south, extreme rainfall occurs in winter/autumn. This study also found that there is a significant relationship between the month when extreme rainfall occurs and the phases of large-scale Pacific drivers (ENSO and IPO). The variation in the relationship between extreme rainfall timing and these large-scale drivers can shift extreme rainfall in southeast Australia by six months, making it extremely challenging to predict extreme rainfall in this region.

The ACCESS-S1 model is used for seasonal prediction of climate in Australia. This project will examine whether ACCESS-S1 model can replicate the observed seasonal pattern of extreme rainfall timing in Australia, and the relationship with large-scale drivers. This study is crucial for improving seasonal prediction of extreme rainfall.

Requirements: Some experience in programming with MATLAB, R, Python or any other software/willing to learn.