An assessment of CMIP5 multi-model decadal hindcasts over Australia from a hydrological viewpoint

- Explores the potential skill of decadal hindcasts over Australia.
- Precipitation predictions show limited skill in comparison to atmospheric variables.
- Major part of the predictability comes from the long term trend.
- The results suggest possible benefits of using a suitably structured downscaling approach.

SummaryThis study explores the potential skill of Coupled Model Intercomparison Project (CMIP5) decadal hindcasts from 9 GCMs and their ensembles over the period 1960-2010 from a hydrological perspective. This assessment is first conducted using GCM derived rainfall and subsequently extended to the other climate variables such as air temperature and geopotential height, with the intention of assessing the possible benefit of using these variables for formulation of appropriate precipitation modelling alternatives. A total of 78 ensemble members for rainfall, and 22 ensemble members for air temperature and geopotential height, are used in performing this decadal predictability assessment.The results suggest that while precipitation predictions show very limited skill when assessed at annual and multi-annual time-scales, air temperature and geopotential height exhibit greater predictive skill. The major part of the predictability of these variables comes from the long term trend. However, the predictability is not enough to drive impact models at decadal timescales and to influence the policy and decision making. The results indicate possible benefits of using a suitably structured modelling approach to derive precipitation using atmospheric predictors exhibiting better skills than the raw simulated rainfall.