Contribution of leaf specular reflection to canopy reflectance under black soil case using stochastic radiative transfer model

Numerous canopy radiative transfer models have been proposed based on the assumption of “ideal bi-Lambertian leaves” with the aim of simplifying the interactions between photons and vegetation canopies. This assumption may cause discrepancy between the simulated and measured canopy bidirectional reflectance factor (BRF). Few studies have been devoted to evaluate the impacts of such assumption on simulation of canopy BRF at a high-to-medium spatial resolution (∼30 m). This paper focuses on quantifying the contribution of leaf specular reflection on the estimation of canopy BRF under a black soil case using one of the most efficient radiative transfer models, the stochastic radiative transfer model. Analyses of field and satellite data collected over the boreal Hyytiälä forest in Finland show that leaf specular reflection may lead to errors of up to 33.1% at 550 nm and 32.8% at 650 nm in terms of relative root mean square error. The results suggest that, in order to minimize these errors, leaf specular reflection should be accounted for in modeling BRF.