The Role of Anthropogenic External Forcing in Australian Rainfall Trends

The seasonal mean of rainfall can be considered to consist of an externally-forced "signal", and a "noise" component related to internal variability. In the global climate system, the externally-forced component is related to changes in radiative forcing, notably anthropogenically induced changes in greenhouse gases, aerosols, stratospheric ozone and land-use. The external forcing variability can therefore be estimated from a multi-model ensemble of Coupled atmosphere-ocean General Circulation Models (CGCMs). Here, we examine the co-variability between the externally-forced components of the atmospheric circulation and Australian rainfall in regions of significant observed trends. In this way, the role of external forcing in Australian rainfall trends, and the mechanisms for this, are investigated.

A multi-model ensemble comprising 11 selected CGCMs from the Coupled Model Intercomparison Project Phase 5 (CMIP5) dataset is used to estimate the externally-forced co-variability between the Southern Hemisphere 500 hPa geopotential height and rainfall for north-west Australia in summer, and winter rainfall for south-west Australia. Both regions have significant trends in observed rainfall in the latter part of the 20th century. In the CMIP5 historical experiment ensemble, the dominant mode of externally-forced co-variability has a rainfall response pattern in both seasons that closely matches the structure of the observed trends. Coupled with this, the 500 hPa geopotential height pattern reflects the expansion of the tropical troposphere in the models, related to increasing historical greenhouse gas concentrations. In Australia north-west summer, this enhances the regional monsoon circulation, resulting in the observed positive rainfall trend. For Australia south-west winter, meridional tropical expansion drives a decrease in the strength of the mid-latitude jet, with anti-cyclonic pressure anomalies upstream of south-west Western Australia. This results in the observed decrease in wintertime rainfall.

In the CMIP5 RCP8.5 experiment the atmospheric circulation and rainfall patterns in both seasons are projected to continue, with increased magnitude, in response to the increased anthropogenic greenhouse gas concentrations in the second half of the 21st century. In contrast, with greenhouse gas stabilization the CMIP5 RCP4.5 experiment, there is no coherent externally-forced co-variability between Australian rainfall and the atmospheric circulation.