At-sensor radiance simulation for airborne imaging spectroscopy

摘要

Physically-based radiative transfer modeling is the key to remote sensing of forest ecosystems. To scale spectral information from the leaf to the sensor level, the canopy architecture of a forest, illumination conditions and the viewing geometry have to be taken into account. Therefore, a new airborne image simulation approach is being developed for the 3D radiative transfer model DART to model individual viewing angles for each pixel of a scene. A first comparison to actual imaging spectrometer data showed promising results, mainly because the atmosphere simulation could be improved compared to previous versions of the DART model.