Pre- and Post-Assimilation Net Radiation Simulation Database by SPAC model

This database comprises half-hourly and daily simulations of land surface net radiation (Rn) and associated energy flux and land surface temperature generated by the Surface Process and Energy Balance–Plant Atmosphere Continuum (SPAC) model for five typical ecosystems in the Heihe River Basin, China, spanning the period from 2018 to 2021. It includes both pre- and post-assimilation model simulations, where surface albedo data were assimilated using an Ensemble Kalman Filter (EnKF) algorithm (implemented in Kalman.py), along with corresponding in-situ observational data used to evaluate model performance. Key statistical metrics—such as R², RMSE, and slope of fit—are provided for each site and timescale to quantify the improvement in performance after assimilation. The accompanying data files also reference the specific meteorological and remote sensing input sources used to drive the model and assimilation system. This dataset supports the analysis of the effectiveness of data assimilation techniques in refining land surface process models, elucidating spatiotemporal dynamics of surface energy balance across heterogeneous landscapes, and reducing uncertainties in ecohydrological modeling. Such information is essential for enhancing the accuracy of climate impact assessments, supporting water resource management, and improving the predictive skill of land surface models in arid and semi-arid regions.