Robust estimates of soil moisture and latent heat flux coupling strength obtained from triple collocation

Land surfacemodels (LSMs) are often applied to predict the one-way coupling strength between surface soil moisture (SM) and latent heat (LH) flux. However, the ability of LSMs to accurately represent such coupling has not been adequately established. Likewise, the estimation of SM/LH coupling strength using ground-based observational data is potentially compromised by the impact of independent SM and LH measurements errors. Here we apply a new statistical technique to acquire estimates of one-way SM/LH coupling strength which are nonbiased in the presence of random error using a triple collocation approach based on leveraging the simultaneous availability of independent SM and LH estimates acquired from (1) LSMs, (2) satellite remote sensing, and (3) ground-based observations. Results suggest that LSMs do not generally overestimate the strength of one-way surface SM/LH coupling.