Evaluation of an improved version of SAIL model for simulating bidirectional reflectance of sugar beet canopies

The processing of remote-sensing data requires simple but accurate models of directional reflectance of the vegetation canopy. In this study, a reflectance model for a homogeneous canopy is evaluated over an extensive set of radiometric measurements performed on sugar beet canopies. The model corresponds to the Scattering by Arbitrary Inclined Leaves (SAIL) model (Verhoef, 1984) in which the term for first order scattering is corrected for hot-spot and leaf specular reflectance. Leaf optical properties are calculated using the PROSPECT model (Jacquemond and Baret, 1990). Experimental data correspond to a two-year experiment and express a large variability of leaf area index, chlorophyll concentration and soil background optical properties. In the first data set, reflectance was measured about midday under vertical viewing in five optical Thematic Mapper bands. In the second data set, both vertical and oblique measurements (zenith angle 45°, four azimuth angles) were performed from sunrise to sunset in the three SPOT bands. Except for leaf cuticle reflectance, structure and optical variables were measured in the field or adjusted to field measurement, independently of reflectance calculations. Although the structure of sugar beet canopies departs strongly from a turbid medium, a good agreement with measurements was obtained in the case of vertical, north and south view directions. However, the model underestimated the measurements close to the hot-spot direction. In the near infrared, there was also some underestimation of canopy reflectance in the opposite direction to the hot-spot. Possible reasons for these differences are discussed.