HR: 0800h
AN: B21A-0049    [Abstracts]
TI: Multi-temporal MODIS-Landsat Data Fusion for Relative Radiometric Normalization and gap Filling of Landsat Data
AU: Roy, D P
EM: david.roy@sdstate.edu
AF: Geographic Information Science Center of Excellence, South Dakota State University, 1021 Medary Ave, Brookings, SD 57007, United States
AU: * Ju, J
EM: junchang.ju@sdstate.edu
AF: Geographic Information Science Center of Excellence, South Dakota State University, 1021 Medary Ave, Brookings, SD 57007, United States
AB: The primary limitation to the utility of Landsat data, other than data cost, is the availability of cloud-free surface observations. As currently only degraded Landsat 5 and Landsat ETM+ systems are acquiring data, and only a single successor Landsat Data Continuity Mission sensor is scheduled, a potential alternative to provide more surface observations is the fusion of Landsat data with data from other remote sensing systems. A semi-physical fusion approach that uses the MODIS BRDF/Albedo surface anisotropy characterization product and Landsat ETM+ data to predict 30m Landsat reflectance on any date is presented. The methodology may be used for ETM+ cloud and SLC-off gap filling and for relative radiometric normalization. It does not require any tuning parameters and so may be automated, it is applied on a per-pixel basis and is unaffected by the presence of missing or contaminated neighboring Landsat pixels. The methodology and spatially explicit results are presented for two Landsat acquisitions at three Landsat scenes, one in Africa and two in the U.S., selected to encompass a range of land cover land use types, and temporal variations in solar illumination, land cover, and phenology. Summary statistics of the difference between the predicted and observed ETM+ reflectance (prediction residual) are compared with the difference between the ETM+ reflectance observed on the two dates (temporal residual) and with respect to the MODIS BRDF model parameter quality. For all three Landsat scenes, and for all bands, except one short wavelength band, the mean prediction residual is smaller than the mean temporal residual, typically by a factor two. This fusion methodology may be applied to any high spatial resolution satellite data with similar spectral bands as MODIS and where the sensor viewing and solar illumination geometry can be accurately derived.
DE: 0480 Remote sensing
SC: Biogeosciences [B]
MN: 2007 Fall Meeting