Effects of soil moisture content on reflectance anisotropy — Laboratory goniometer measurements and RPV model inversions

• Published in: Remote sensing of environment 2015-12, Vol.170, p.229-238
• Main Authors: Roosjen, P.P.J., Bartholomeus, H.M., Clevers, J.G.P.W.
• Format: Article
• Language: English
• Subjects: Anisotropy; Correlation; Goniometer; Goniometers; Laboratorium voor Geo-informatiekunde en remote sensing; Laboratory of Geo-information Science and Remote Sensing; Mathematical models; Multi-angular reflectance measurements; PE&RC; Reflectance; Reflectance anisotropy; Reflectivity; RPV model; Soil (material); Soil moisture
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2015.09.022

Optical methods to study soil moisture content (SMC) are often based on empirically or physically based models that relate changes in reflectance intensity to SMC. The effects of SMC on the reflectance anisotropy, however, have not received much attention. In this paper the effects of SMC on the anisotropic reflectance behaviour of soils were studied. Biconical reflectance factors (BCRFs) of five different soil samples were acquired at 60 positions covering the full hemisphere in the optical domain at different SMC levels using Wageningen University's laboratory goniometer facility. In addition, we inverted the Rahman–Pinty–Verstraete (RPV) model against the measured BCRFs in the principal plane. The results show that the anisotropic reflectance behaviour of soils is strongly influenced by the SMC. Dry soils displayed strong backward scattering behaviour, with a maximum reflectance close to the hotspot position. An increase of the SMC level up to the soil's saturation point caused the soils to scatter more in the forward direction and induced a weakening of the hotspot effect. Oversaturated soils displayed a strong sun-glint-like reflectance peak in the anti-solar direction. The RPV model fitted the measured BCRFs in the principal plane up to saturated SMC levels in general with an R2>0.9. It was not possible to fit the model through observations of oversaturated soils, since the RPV model is not capable of simulating specular reflectance. The asymmetry parameter (Θ) of the RPV model, which controls the proportion of forward and backward scattering, showed a strong correlation to SMC for individual samples. This correlation remained significant when we considered all samples together with a maximum R2 of 0.797 at 2123nm, indicating that reflectance anisotropy contained information on the water content of soils. The amplitude parameter (ρ0), which is closely related to the reflectance intensity, only had a maximum R2 of 0.622 at 1921nm. This indicated that reflectance anisotropy contained more information on SMC than the level of spectral reflectance as such. •Effects of SMC on reflectance anisotropy were measured and modelled.•Increasing SMC levels increased forward scattering intensity of light.•Increasing SMC levels decreased the backward scattering intensity and hotspot effect.•The RPV model could simulate anisotropy curves of soils at varying SMC levels.•The proportion of forward/backward scattering contained information on the SMC level.