Key points are not available for this paper at this time.
Terrestrial evaporation (E) is a keystone flux linking water, energy and carbon cycles. Consequently, monitoring of E at high temporal and spatial resolution over an extended period is crucial to diagnose climate change and its influence on the acceleration of the global hydrological cycle. As E cannot be directly observed from space, a modelling approach is required to derive E from global, observational remote sensing and meteorological datasets1. The array of available approaches ranges from purely data-driven E retrievals2 to physically-based estimates from traditional land surface models. In this presentation, we introduce the fourth version of the Global Land Evaporation Amsterdam Model (GLEAM), a hybrid evaporation model that harnesses the synergy between process-based modelling and machine learning. The conceptual backbone of the model, a soilvegetation water balance module, is updated from earlier GLEAM versions with new representations of interception loss3, plant access to groundwater4 and potential evaporation. Additionally, earlier empirical evaporative stress functions are replaced by deep neural networks trained on eddy-covariance and sapflow data to better represent the complex physiological response of vegetation to multiple environmental stressors5. Future research directions include the increase in temporal resolution to sub-daily and the training of the stress functions in an end-to-end differentiable modelling framework6. GLEAM4 continuous, daily datasets at 0.1 spatial resolution covering the period 19802023 including evaporation and its components, soil moisture, potential evaporation and evaporative stress estimates will be openly available via www.gleam.eu upon publication. References 1Fisher, J. B., et al., The future of evapotranspiration: Global requirements for ecosystem functioning, carbon and climate feedbacks, agricultural management, and water resources, Water Resour. Res., 53, 26182626, 2017, https://doi.org/0.1002/2016WR020175. 2 Jung, M., Koirala, S., Weber, U. et al. The FLUXCOM ensemble of global land-atmosphere energy fluxes, Sci. Data, 6, 74, 2019, https://doi.org/10.1038/s41597-019-0076-8 3Zhong, F., Jiang, S., van Dijk, A. I. J. M., Ren, L., Schellekens, J., and Miralles, D. G.: Revisiting large-scale interception patterns constrained by a synthesis of global experimental data, Hydrol. Earth Syst. Sci., 26, 56475667, 2022, https://doi.org/10.5194/hess-26-5647-2022 4Hulsman, P., Keune, J., Koppa, A., Schellekens, J., Miralles, D. G., Incorporating plant access to groundwater in existing global, satellite-based evaporation estimates, Water Resour. Res., 59, e2022WR033731, 2023, https://doi.org/10.1029/2022WR033731 5Koppa, A., Rains, D., Hulsman, P. et al., A deep learning-based hybrid model of global terrestrial evaporation, Nat. Commun., 13, 1912, 2022, https://doi.org/10.1038/s41467-022-29543-7 6Shen, C., Appling, A.P., Gentine, P. et al., Differentiable modelling to unify machine learning and physical models for geosciences, Nat. Rev. Earth. Environ., 4, 552567, 2023, https://doi.org/10.1038/s43017-023-00450-9
Bonte et al. (Sat,) studied this question.