Assessment of moisture content in sandy beach environments from multispectral satellite imagery

A framework for estimating moisture content from satellite-based multispectral imagery of sandy beaches was tested under various site conditions and sensors. It utilizes the reflectance of dry soil and an empirical factor c relating reflectance and moisture content for a specific sediment. Here, c was derived two ways: first, from in situ measurements of moisture content and average NIR image reflectance; and second, from laboratory-based measurements of moisture content and spectrometer reflectance. The proposed method was tested at four sandy beaches: Duck, North Carolina; and Cannon Beach, Ocean Cape, and Point Carrew, Yakutat, Alaska. Both measured and estimated moisture content profiles were impacted by site geomorphology. For profiles with uniform slopes, moisture contents ranged from 3.0% to 8.0% (zone 1) and from 8.0% to 23.0% (zone 2). Compared to field measurements, the moisture contents estimated using c calibrated from in situ and laboratory data resulted in percent error of 3.6%-44.7% and 2.7%-58.6%, respectively. The highest percent error occurred at the transition from zone 1 to zone 2. Generally, moisture contents were overestimated in zone 1 and underestimated in zone 2, but followed the expected trends based on field measurements. When estimated moisture contents in zone 1 exceeded 10%, surface roughness, debris, geomorphology, and weather conditions were considered.