Prediction of the outflow temperature of large-scale hydropower using theory-guided machine learning surrogate models of a high-fidelity hydrodynamics model

•A reservoir outflow temperature prediction model based on the theory-guided machine learning (TGML) paradigm was constructed.•In the absence of observation data, the training dataset generated by the numerical simulation model can effectively support the training of the TGML model.•The TGML model can accurately and efficiently predict the outflow temperature for various stratified water intake schemes.