Using 1-D models to interpret the reflectance anisotropy of 3-D canopy targets: Issues and caveats

This paper evaluates 1) to what extent one-dimensional (1-D) models can be used to represent the magnitude and directionality of the surface reflectance field of heterogeneous canopy targets at different spatial resolutions, and 2) whether this usage results in significant biases in the estimation of the corresponding state variables. It will be seen that when both the I-D and three-dimensional (3-D) models account for all features of the measured radiation field, then-in the absence of further information regarding the nature and structure of the target-the use of a 3-D model may amount to an over-interpretation of the available data. The simplified surface structure formulation contained within the I-D model, on the other hand, may affect the values of the state variables that such models will retrieve. This is because the shape of the reflectance anisotropy of the 3-D target is almost always different from that of a structurally homogeneous (1-D) canopy with the same state variable values but no foliage clumping. By consequence the 1-D canopies that are capable of mimicking the bell (or bowl) shaped reflectance anisotropy of 3-D targets will tend to feature lower leaf area index, higher soil albedo and, in particular, predominantly erectophile (or plagiophile) leaf normal distributions.