Comparison of agricultural roughness measurements from pin boards and terrestrial laser scanners: implications for SAR scattering models

English Abstract:

Surface roughness and soil moisture primarily influence C-band synthetic aperture RADAR (SAR) backscatter over unvegetated agricultural fields. Accurate estimation of soil moisture through inversion of SAR scattering models requires correct characterization of surface roughness relative to backscatter and soil moisture possibilities using look-up tables (LUT). Roughness is defined as the stochastic variance in height of a soil surface relative to the wavelength of interest. The traditional method of determining surface roughness for use in scattering model LUT development involves manual 1-m pin profilers. This equipment consists of a board holding a line of 100 vertically mobile pins separated by 1 cm each. Roughness profiles are obtained by aligning the pins vertically so that each pin rests on the underlying micro topography. A photograph is then taken of the pin profiler for digitization. Field specific roughness statistics from this technique are affected by the number of profile samples, profile length, profile direction, and photograph digitization error. Light detection and ranging (LiDAR) systems use reflected light to measure the distance between the instrument and an object and can be used to characterize a surface in three dimensions. LiDAR can yield an infinite number of omnidirectional surface profiles from a single surface model of a field. In this study, surface roughness parameters extracted from pin profilers are compared to estimates acquired from a terrestrial LiDAR scanner using data collected over 10 agricultural fields of varying roughness within the Elora Research Station (University of Guelph). Estimates from both instruments are applied to the Integral Equation Model to understand the sensitivity of soil moisture inversion accuracy to surface roughness measurement techniques.