Answers are grounded in this paper and cite their sources.
Latent heat flux (LE) is a key component of land surface energy partitioning and plays a critical role in regulating terrestrial water–carbon interactions and land–atmosphere feedback. Accurate quantification of LE depends critically on reliable estimates of effective aerodynamic resistance (rae) and surface resistance (rs), which regulate turbulent transport and surface water vapor transfer. However, existing large-scale LE estimation frameworks often rely on empirical or prescribed resistance parameterizations, and most global products provide only final flux estimates while lacking long-term, spatially continuous datasets of rae and rs. To address these limitations, we produced a spatiotemporally continuous global dataset at 0.05° resolution for 2001–2020, including LE, rae, rs, the decoupling coefficient (Ω), and maximum surface relative humidity (hs,max) by an energy- resistance-coupling (ERC) iterative framework based on multi-source remote sensing, reanalysis, and flux observations. The framework integrates (i) an aerodynamic temperature (T0) calibration scheme, (ii) an XGBoost-based dynamic constraint on the hs,max and (iii) a Newton–Raphson solver to obtain a physically constrained nonlinear solution for the resistance network and latent heat flux. Validation against 244 flux sites showed good agreement between ERC LE and site observations, with an overall R2 of 0.76 and RMSE of 23.54 W m−2; the XGBoost-estimated hs,max also performed well, with a R2 of 0.91, and an RMSE of 0.06, while the derived rae, rs and Ω exhibited physically consistent spatial gradients and seasonal dynamics. This dataset provides a new global high-resolution source of resistance-related variables and can support drought monitoring, land–atmosphere coupling analyses, and the evaluation of Earth system models.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: