HR: 0800h
AN: IN31A-1142 [Abstracts]
TI: Synergy Between Ground Measurements and High Spatial Resolution Imagery to Validate Medium Spatial
Resolution Land Surface Product
AU: * Garrigues, S
EM: garrigue@pop600.gsfc.nasa.gov
AF: University of Maryland, NASA GSFC Mail Code 614.4, Greenbelt, MD 20771
United States
AU: Morisette, J
EM: jeff.morisette@nasa.gov
AF: NASA GSFC, Mail Code 614.5, Greenbelt, MD 20771
United States
AU: Baret, F
EM: baret@avignon.inra.fr
AF: INRA CSE, Agroparc-Domaine St Paul, Avignon, 84000
France, Metropolitan
AU: Privette, J
EM: jeff.privette@nasa.gov
AF: NASA GSFC, Mail Code 614.4, Greenbelt, MD 20771
United States
AB:
Currently, several biophysical variables such as the Leaf Area Index (LAI), the Fraction of Absorbed Photosynthetically
Active Radiation (fAPAR) and the surface albedo are derived from remote sensing observations acquired with medium spatial
resolution sensors (250m to 7km). Because of their high temporal frequency and their important spatial coverage, these
products are very useful to describe the mass and energy fluxes between the earth surface and the atmosphere. Since these
variables are required for an important range of investigations and applications, it is important to assess their accuracy.
Validation activity consists in evaluating by independent means the quality of the land surface products estimated from
coarse resolution sensors. Validation methods consist in generating a ground truth map of these products at high spatial
resolution. These maps are produced by using ground measurements of the biophysical variable and radiometric data from a high
spatial resolution sensor (10m-30m). The relationship between a biophysical variable and radiometric imagery, called the
transfer function, allows extending the local ground measurements to the entire high spatial resolution image. The resulting
biophysical variable map is aggregated to be compared with the medium spatial resolution satellite biophysical products.
Several geometrical issues influence the validation results:
- the registration accuracy of the local ground measurements relative to the high spatial resolution image
- the difference of spatial support between the ground measurement support and the high spatial resolution pixel
- the registration accuracy between the high and medium resolution images
- the point spread function (PSF) associated to the medium spatial resolution image
This work proposes a methodology to account for these sources of uncertainties within the validation process. First, these
problems are investigated at the field measurement scale. The registration accuracy between ground measurement located by GPS
and high spatial resolution pixel is modeled by a Gaussian random variable. For each possible relative position, the ground
measurement is related to the radiometric data for the surrounding area weighted by the ground measurement spatial support
area. A Monte Carlo simulation scheme accounting for the positional accuracy provides the probability distribution function
of the parameters defining the transfer function. Second, registration error is investigated when aggregated high spatial
resolution image is compared with coarse resolution image. The registration error is minimized by getting the best
geometrical match between the two images using correlation techniques. Finally, the importance of the PSF associated to the
medium spatial resolution biophysical products is evaluated using data from MODIS and VEGETATION sensors. This investigation
is applied to several validation sites expressing a range of spatial heterogeneity.
DE: 3200 MATHEMATICAL GEOPHYSICS (0500, 4400, 7833)
DE: 3252 Spatial analysis (0500)
DE: 3265 Stochastic processes (3235, 4468, 4475, 7857)
DE: 3275 Uncertainty quantification (1873)
DE: 3294 Instruments and techniques
SC: Earth and Space Science Informatics [IN]
MN: Fall Meeting 2005