Geotechnical Investigation of Exposed Intertidal Flats at the Great Bay Estuary Using Sediment Sampling and Satellite-Based Synthetic Aperture Radar

Abstract

Estuarine intertidal flats are exposed to a wide range of changing environmental conditions, but despite many years of field studies, little is known about their geotechnical characteristics. This study investigates (1) the variability of sediment grain size distributions, water contents, and Atterberg limits of three distinct exposed tidal flats; and (2) the potential pathways for remote characterization of those sites. Sediment samples were collected from three flats in the Great Bay Estuary located in the Gulf of Maine at the border of New Hampshire and Maine, United States: two sites composed of primarily fine-grained soils (Adams Point and Woody) and one composed of predominately coarse-grained soils (Mast Cove). Additionally, high-resolution X-band synthetic aperture radar (SAR) images were collected concurrently to field measurements from various satellites and over different incidence angles. In each image, the footprint area corresponding to the three exposed tidal flats was extracted, and statistical properties of the backscatter intensity histograms for each region were computed and compared. Moisture contents, fines contents, and grain size distributions varied between the three sites, with the USCS classifications ranging from elastic silts (Woody) and low-plasticity silts (Adams Point) with high water contents (83.2%-172.7%) and fines contents (83.4%-99.5%) to silty sand and well-graded sands with gravels (Mast Cove) with lower water contents (17.7%-48.3%) and fines contents (4.8%-44.3%). Across the sites, the statistical properties of the SAR backscatter exhibited distinct trends regarding mean backscatter, entropy, and uniformity in relation to the Unified Soil Classification Scheme (USCS) classification and fines content. These trends were consistent across all images collected. A potential sediment classification scheme to derive the USCS classification from the X-band SAR imagery is suggested utilizing the mean, entropy, and uniformity of the backscatter coefficient.