Australia's water policy and infrastructure investment decisions require information on water security under a range of plausible future climate conditions. This information is currently provided to the country through numerous, methodologically inconsistent one-off studies for limited geographical regions (for example single catchments, urban regions or states). There would be benefits in adopting agreed and consistent approaches nationally, including ensuring that policy and investment decisions are based upon an accessible, authoritative set of national climate projections for water, and that climate change risks are properly factored into infrastructure, investment and policy related decisions.

To address this need, we are undertaking project to produce a set of consistent, national projections of the impacts of climate change on water and water related variables. This is made possible because of the new, national scale, Australian Water Resource Assessment – Landscape (AWRA-L) model. The project aims to bring together a number of state-of-the-art downscaling techniques together with the CMIP5 ensemble to sample uncertainty along the impact modelling chain, using AWRA-L and other continental-scale hydrological models. Uncertainties due to downscaling of global circulation model (GCM) outputs are considerable (e.g. Jacob et al. 2016, Maraun et al. 2017) and as part of this project, currently available downscaled climate projections for Australia, both statistical and dynamical, will be evaluated and bias-corrected for use as an ensemble of downscaled climate data to force hydrological models. Bias-correction methods will be selected for fitness-of-purpose for hydrological impact modelling, using results from international intercomparison projects (Nikulin et al. 2017), as well as local evaluations (e.g. Frost et al. 2011). The final service aims to support customers with both nationally modelled climate change impacts on water as well as hydrological model ready ensembles of downscaled climate inputs. In this presentation, we present some of the results so far.