Quantifying confidence in projected evapotranspiration and irrigation demand under climate change: A high-resolution ensemble approach

Projected changes in evapotranspiration have important implications for irrigation planning in water-limited regions. This study quantifies mid-century changes in reference evapotranspiration ( ET 0 ), representative crop evapotranspiration ( ET c ), and irrigation water demand across the Murray-Darling Basin (MDB), with careful consideration of model uncertainty. An ensemble of ten downscaled climate projections, comprising five CMIP6 global climate models and two regional climate models, was used to assess two emissions scenarios and two mid 21st century periods. Crop-specific irrigation demand was estimated for almonds and winegrapes by combining ET 0 with phenology-adjusted crop coefficients and effective rainfall ( P e ). Results indicate consistent increases in ET 0 across all regions and seasons, with the largest percentage increases occurring in spring and the largest absolute increases in summer. Rising temperatures were the dominant drivers of ET 0 change, while projected changes in wind speed and net radiation were small and exhibited low inter-model agreement. Signal-to-noise analysis indicated moderate confidence in ET 0 projections, with marginally higher confidence under the lower emissions scenario and in the later projection period. When translated to crop water demand, projected increases in ET c – P e were substantial across the MDB, often exceeding ET 0 changes in relative terms and implying significant increases in absolute irrigation demand, particularly in regions with high baseline water use, although differences between emissions scenarios were frequently smaller than the spread across the model ensemble. Differences between crops were minor relative to the influence of increasing atmospheric demand. Climate driven increases in evaporative demand are projected to raise MDB irrigation requirements, but the considerable model uncertainty identified here indicates a need for adaptive, risk-informed water management strategies.