The effect of solar illumination angle and sensor view angle on observed patterns of spatial structure in tallgrass prairie

While it has long been recognized that the anisotropic reflectance properties of a natural surface affect the intensity and spectral distribution of radiance received by a remote sensing instrument, the effects of canopy reflectance geometry on the observed spatial structure of canopy reflectance have not adequately been evaluated. In this paper, near-surface spectrometers were used as part of two experiments to evaluate the systematic variations in the sun-target-sensor geometry on semivariogram metrics (range, sill+nugget variance) summarizing the spatial structure observed in a tallgrass prairie canopy. In the first experiment, reflectance measurements and normalized difference vegetation index (NDVI) values were collected at five sensor viewing angles (-50/spl deg/, -25/spl deg/, 0/spl deg/, 25/spl deg/, and 50/spl deg/) from six measurement grids representing three burn treatments and two slope/aspect situations. In the second experiment, data were collected at 2-h intervals, beginning at /spl ap/0800 LST from the same grids with the radiometer at nadir, allowing the spatial structure of reflectance and NDVI to be observed under naturally changing illumination. Results of the geostatistical analysis show that both the range and sill+nugget variance values change with viewing angle. These effects were consistent across all treatments and slope/aspect combinations. However, when viewed from nadir, the sill+nugget variance values of the canopy changed with solar illumination angle and the range values remained nearly constant. These relationships were also observed across all treatments and slope/aspect combinations. The results suggest that sill+nugget values for the same surface may not be directly comparable if not acquired under very similar view angle and illumination conditions. Range values are comparable if the nadir view is used, but not under off-nadir viewing conditions. The implications of these findings point to the need for caution in interpreting spatial structure derived from close-range radiometry or from satellite/aircraft instruments with cross-track or off-nadir pointing capabilities, and in the comparison of images acquired under varying illumination conditions.