Unsaturated zone model complexity for the assimilation of evapotranspiration rates in groundwater modelling

Abstract. The biophysical processes occurring in the unsaturated zone have a direct impact on water table dynamics. Representing these through application of models different complexity has an estimates volumes flowing between and aquifer. These fluxes, known as net recharge, are often used shared variable that couples to groundwater models. However, recharge always affected by degree uncertainty, model–data fusion methods, such data assimilation, can be inform coupled reduce uncertainty. This study assesses effect (conceptual versus physically based) update model outputs, assimilation actual evapotranspiration rates, for water-limited site South Australia. Actual rates assimilated because they been shown related dynamics thus form link remote sensing deeper parts soil profile. Results quantified using standard metrics, root mean square error Pearson correlation coefficient, reinforced calculating continuous ranked probability score, which is specifically designed determine more representative stochastic It found that, once properly calibrated reproduce evapotranspiration–water dynamics, simple conceptual may sufficient this purpose; one configuration over other should motivated specific purpose simulation information available.