Evaluating the influence of surface soil moisture and soil surface roughness on optical directional reflectance factors

Fine‐scale information on soil surface roughness (SSR) is needed for calculating heat budgets, monitoring soil degradation and parameterizing surface runoff and sediment transfer models. Previous work has demonstrated the potential of using hyperspectral, hemispherical conical reflectance factors (HCRFs) to retrieve the SSR of different soil crusting states. However, this was achieved by using dry soil surfaces, generated in controlled laboratory conditions. The primary aim of this study was therefore to test the impact that in situ variations in surface soil moisture (SSM) content had on the ability of directional reflectance factors to characterize SSR conditions. Five soil plots (20 cm × 20 cm in area) representing different agricultural conditions were subjected to different durations of natural rainfall to produce a range of different levels of SSR. The values of SSM varied from 8.7 to 20.1% across all soil plots. Point laser data (4‐mm sample spacing) were geostatistically analysed to give a spatially‐distributed measure of SSR, giving sill variance values from 3.2 to 23.0. The HCRFs from each soil state were measured using a ground‐based hyperspectral spectroradiometer for a range of viewing zenith angles from extreme forward‐scatter (θ_r = ?60°) to extreme back‐scatter (θ_r = +60°) at a 10° sampling resolution in the solar principal plane. The results showed that despite a large range of SSM values, forward‐scattered reflectance factors exhibited a very strong relationship with SSR (R² = 0.84 at θ_r = ?60°). Our findings demonstrate the operational potential of HCRFs for providing spatially‐distributed SSR measurements, across spatial extents containing spatio‐temporal variations in SSM content.