Answers are grounded in this paper and cite their sources.
We present a method that integrates physically based evaporation modelling approaches with high-resolution meteorological data interpolated from land-based observations to estimate daily open water evaporation at local and regional scales across the data-sparse Pilbara region of north-western Australia. The evaporation modelling approach considers water body heat storage and enables water bodies of different sizes, depth and salinity to be modelled. Daily meteorological data were interpolated at 0.01° resolution for the 440,000 km 2 Pilbara region based on the best performing model from up to 14 thin plate spline models per meteorological variable, which used varying combinations of latitude and longitude, elevation and a coastal distance index as spatial predictors. Using the interpolated meteorological data, the evaporation model was run across the study region (at a daily time step with 0.01° spatial resolution) for all combinations of water body size and salinity. Comparison of model results with measurements for three different sized water bodies yielded R 2 > 0.64 and an RMSE of < 1.6 mm d −1 giving confidence in results. Reference water body evaporation was high across the focus area and ranged between 2272 and 2965 mm y −1 . We showed that the spatio-temporal open water evaporation model can be used to estimate evaporation for both existing and proposed water bodies. To simplify model application to different waterbody characteristics we developed an evaporation scaling factor which was used to scale model results from a reference water body (surface area of 1 km 2 , salinity of 10 mS cm −1 and depth of 10 m) to any water body with surface area between 0.01 and 5 km 2 and salinity between 0 and 300 mS cm −1 . The method developed provides the first daily-scale open water evaporation estimates across the Pilbara region and can be applied to other locations where open water evaporation estimates are required for water resource management by using spatially interpolated meteorological grids and/or reanalysis products.
No takes yet. Share an insight, caveat, or question.
McJannet et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: