Ground Heat Flux Estimation: What's the Best Approach?

Abstract:

Ground heat flux (G) is typically the smallest term of the surface energy balance (i.e., R_n=LE+H+G), but high uncertainty in G means that evaluation of the other terms remains problematic. There has been a large range of approaches to estimating G for large-scale applications, with a subsequent wide range of accuracies/errors. We provide the largest review of these approaches to date (n=10), evaluating modeled G against measured G from 78 FLUXNET sites. Empirical relationships constrained by vegetation indices and R_n better capture daily variability of instantaneous G, while a physically based heat transfer model exhibits lower bias and error across all sites. Global relationships calibrated to in situ measurements also reveal low bias and error in addition to capturing the variability comparable to the other empirical relationships. We produced global decadal datasets of G for all models to illustrate regional sensitivities by model. This work has potential to improve energy balance estimates of LE and provides a new dataset to assimilate into or assess land surface models.