PROJECT RATIONAL

The spatial response of a sensor is as important as its spectral and radiometric response. It is often assumed to be simply the spatial average over the sensor GIFOV However,the actual response function is often not square, and in fact for MODIS, is twice as wide cross-track as in-track because of time integration during scanning. Moreover, the profile of the EGIFOVis not a rectangle function, but a triangle in the cross-track direction. Although these differences between the GIFOV and EGIFOV are sufficient in themselves to warrant careful measurement of the spatial response, there is further justification in that the effect of the spatial response on derived quantities, such as certain MODIS Level 2 products, is nonlinear.

The significance of this discussion to EOS products is that the Level1 radiance data or Level 2 reflectance data should be corrected for the sensor spatial response before calculating geophysical parameters,such as NDVI, in order to obtain a truly absolute geophysical parameter for inter-sensor comparisons. A correction for the PSF at the level of the geophysical parameter may not be possible, as in the case of NDVI because of its nonlinear dependence on the sensor bands.