Decision-makers urgently need fit-for-purpose, actionable climate information to address worsening risks to water security due to climate change. To help address this need, we pioneer new work that builds on a recently established standardized benchmarking framework to assess downscaled precipitation simulations and explore how benchmarking historical, fundamental model performance impacts confidence in regional precipitation projections. Inspired by frameworks used by the Coupled Model Intercomparison Project (CMIP) community to navigate the “ensemble of opportunity” that explore the spread, independence, and performance of CMIP models, we assess future confidence across three related categories: model agreement, dependence, and performance. Using data from the Coordinated Regional Downscaling Climate Experiments (CORDEX)-Australasia domain and Southern Australia as a case study, we highlight how benchmarking only “fundamental,” historical model performance improves confidence in regional precipitation projections across these categories of confidence. We find that including fundamentally flawed models in an ensemble can indicate a false level of confidence and model agreement in regional climate projections in annual and wet-season total rainfall and rainfall extremes. We find that our benchmarked subset reduces the spread in projections across season-rainfall index combinations without underestimating observed natural variability or end-of-century uncertainties due to the climate change response. By taking a novel approach to address the coupled interdependencies between global climate models and regional climate models in our dynamically downscaled ensemble, we find that we can increase confidence though historical benchmarking spatially and temporally without over-constraining plausible future projections.

This study applies a benchmarking framework to assess a 39-member ensemble of regional climate models which have dynamically downscaled CMIP6 models over the CORDEX Australasian region. Four modelling centres contributed regional climate models to this ensemble using three regional climate models (CCAM, WRF, and BARPA-R), and a total of five model configurations. Assessment is conducted over the Australian continent using a separation into four major climate zones over a 30 year historical climatological period (1985-2014). Rainfall and near-surface temperatures are compared against six benchmarks measuring mean state patterns, spatial and temporal variance, seasonal cycles, long-term trends and selected extreme indices. Benchmark thresholds are derived either from previous studies or comparison with the driving model ensemble. Major model biases vary between ensemble members and include dry biases in northern and southern Australia, winter wet biases and a persistent low bias in the winter diurnal temperature range across all the modelling centres. Daily variability at large length-scales is comparable in the driving global climate model and downscaled regional climate model length-scales, and long-term trends are largely determined by the driving global climate model. Overall, the ensemble was deemed to be fit for purpose for impact studies. Strengths and weaknesses of the systematic benchmarking framework used here are discussed.