CONUS-scale implementation of NeuralFAO56 v1.0: data, model outputs, and supporting materials for GMD submission

This repository supports the continental United States (CONUS)-scale implementation of NeuralFAO56 v1.0 presented in the associated Geoscientific Model Development (GMD) manuscript. NeuralFAO56 was executed in Mode 2 across CONUS, integrating physics-based FAO-56 reference evapotranspiration (ETo) computation with deep learning (DL)-based ETo forecasting at each meteorological station. The archive includes: Historical, real-time, and forecast meteorological datasets retrieved for each station, along with FAO-56-based daily ETo computations, merged and stored under Station_Data/With_etref. DL-based ETo forecasts at daily temporal resolution for all forecast horizons using both LSTM and Transformer models, stored under NeuralFAO56_DL_models/predictions. Model performance metrics for both DL models, stored under NeuralFAO56_DL_models/metrics. Simulation plots generated during model execution, stored under NeuralFAO56_DL_models/plots. Scripts used to reproduce all results and figures presented in the GMD manuscript stored under Code_for_figures_in_paper. All materials are organized to enable full reproducibility of the results presented in the associated publication.