Sensible Heat and Latent Heat Flux Estimates in a Tall and Dense Forest Canopy under Unstable Conditions

A method to estimate the sensible heat flux (H) for unstable atmospheric condition requiring measurements taken in half-hourly basis as input and involving the land surface temperature (LST), H-LST, was tested over a tall and dense aspen stand. The method avoids the need to estimate the zero-plane displacement and the roughness length for momentum. The net radiation (Rn) and the latent heat flux (lambda E) dominated the surface energy balance (SEB). Therefore, lambda E was estimated applying the residual method using H-LST as input, lambda ER-LST. The sum of H and lambda E determined with the eddy covariance (EC) method led to a surface energy imbalance of 20% Rn. Thus, the reference taken for the comparisons were determined forcing the SEB using the EC Bowen ratio (BREB method). For clear sky days, H-LST performed close to H-BREB. Therefore, it showed potential in the framework of remote sensing because the input requirements are similar to current methods widely used. For cloudy days, H-LST scattered H-BREB and nearly matched the accumulated sensible hear flux. Regardless of the time basis and cloudiness, lambda ER-LST was close to lambda E-BREB. For all the data, both H-LST and lambda ER-LST were not biased and showed, respectively, a mean absolute relative error of 24.5% and 12.5% and an index of agreement of 68.5% and 80%.

Automated summary

This method estimates the sensible heat flux for unstable atmospheric conditions using half-hourly measurements as input, avoiding the need for estimating other parameters. It shows potential for remote sensing applications and demonstrates high accuracy without bias in all data.