Conditioning a multiple-patch SVAT model using uncertain time-space estimates of latent heat fluxes as inferred from remotely sensed data

It has been shown that the calibration of soil vegetation-atmosphere transfer (SVAT) models is inherently uncertain, even when data are available over a relatively limited homogeneous area. The representation of subgrid-scale variability of fluxes is not easily achieved because of the lack of information available about appropriate parameter distributions and their covariance. However, remote sensing of thermal surface responses offers the possibility of obtaining distributed estimates of surface fluxes. In this paper, multiple Landsat-Thematic Mapper (TM) images of the First International Satellite Land Surface Climatology Project (ISLSCP) Field Experiment (FIFE) site are used to derive uncertain estimates of the land surface-atmosphere sensible and latent fluxes over a period of time. Employing a framework based on fuzzy set theory, the parameter space representing all feasible parameterizations of a SVAT model are examined with respect to these image estimates. Areal weightings for a number of functional types of flux behavior are then derived through which the temporal evolution of surface fluxes can be estimated.