Remote Sensing of Temperature Profiles in Vegetation Canopies Using Multiple View Angles And Inversion Techniques

Most remote sensing applications in natural resource disciplines deal with vegetation or with substrate materials covered by a vegetation canopy. Remote sensing of vertical temperature profiles of the vegetation canopy (including the substrate) provide valuable information to resource managers. A mathematical approach is presented which theoretically allows the determination of a temperature profile from multiple, sensor view angles and a priori knowledge of vegetative geometric structure. The technique was evaluated on data from several wheat canopies at different stages of development. The technique was shown to be most applicable for the separation of vegetation and substrate temperatures. One can expect relatively accurate inferences of mean vegetation surface temperatures for intermediate and dense vegetation canopies, and relatively poor inferences of vegetation temperatures for sparse canopies. The converse is true for inferring substrate temperatures. The root-mean-square prediction accuracy of vegetation temperatures for intermediate-dense wheat canopies was 1.8 and 1.4°C for an exact and overdeterminate system, respectively. These findings have significant implications for any discipline concerned with the vegetation itself or the underlying substrate. For specific vegetation geometries this technique may provide a means for uniquely separating mean vegetation and substrate temperatures when a priori knowledge of vegetation geometry and two or more sensor view angle measurements can be obtained.