Representation of precipitation in models limited by small scale processes

We wanted to find out if climate models could simulate the complicated relationships between Australian precipitation and the larger atmospheric flow patterns that influence it.  Widely recognised patterns include El Nino and other natural phenomena, but a pattern of new concern is the gradual shift of climate zones away from the equator that has been reported in recent studies, and is likely caused by humans.  Due to our subtropical location this could further dry much of the continent.

Figure: Climate model skill scores for the atmospheric flow patterns (variability; shown in blue) and precipitation (shown in green). Observations are based on assimilated data products, specifically ERA-Interim and Australia Water Availability Project, shown as the darkest colours.


The good news is that models are doing a reasonably good job of simulating the relationships between both types of flow pattern, and the influence each has on precipitation.  The bad news is that recent model versions have not improved in either respect compared to those from 5-10 years earlier.

Interestingly, in cases where patterns did improve, this did not seem to help precipitation.  This suggests that it is poor representation of small-scale processes in the models, rather than their simulation of large-scale flow patterns, that is holding back progress in accurately simulating Australian precipitation.

UNSW logo ANU logo Monash logo UMelb logo UTAS logo