The Bi-directional Reflectance Distribution Function (BRDF) is a description of the anisotropic scattering of electro-magnetic radiation by a surface. Mathematical models of the BRDF have the potential to provide estimates of various key environmental variables such as Leaf Area Index (LAI) and Earth surface albedo, by inversion against Earth Observation data. A suite of linear semi-empirical kernel-driven BRDF models, known collectively as AM-BRALS, has been devised by NASA for use with data from the TERRA/MODIS sensor. These have yet to be tested in any comprehensive, operational fashion. This thesis examines the potential of the AMBRALS models (specifically the Ross-Thick Li-Sparse kernel combination), to provide information from directional reflectances acquired by the CNES sensor SPOT-4 VEG-ETATION (VGT). VGT is comparable with MODIS in terms of angular sampling and spatial resolution and thus provides an ideal opportunity to test the AMBRALS suite. Potential weaknesses in the pre-processing of VGT are identified and examined in terms of their impact on the inversion of AMBRALS. These were namely; cloud-masking and atmospheric correction. A new cloud-mask is suggested for VGT data that are to be used with BRDF models. Aerosol optical depth is identified as the atmospheric parameter which has the greatest impact on BRDF model inversion. The information content of VGT acquired directional reflectances and resulting BRDF model products are examined statistically. The complementarity of this information with that already provided by the VGT spectral channels is explored.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:638596 |
| Date | January 2000 |
| Creators | Quaife, T. L. |
| Publisher | Swansea University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0068 seconds