A geophysical study of palaeochannels on the Somerset Levels coastal plain and wetland to explore river landscape evolution.

Palaeochannels offer a glimpse into the history of a landscape. In the context of shifting perspectives from traditional hard engineering to soft nature-based measures, modern flood risk management could benefit from an understanding of the natural processes and features preserved within palaeochannels, which have otherwise been hidden by a legacy of engineering and land management on the river and floodplain. This study uses geophysical surveying techniques to bridge the gap between surface topography LiDAR data and sediment core data, in order to investigate the evolution of past rivers and tidal inlets in the Somerset Levels coastal plain and inland wetlands in southwest England. Case studies from a range of palaeochannels across the Somerset Levels are presented to identify the advantages and limitations of applying the methodology to a coastal plain and wetland dominated by Holocene alluvium and increasing human influence over the past several centuries. Four river systems represent both tidally dominated and inland freshwater conditions: a large tidal creek system within predominantly clay sediment; an inland river system traversing a peat wetland which was the former course of a major drainage network before intentional diversion; and two systems at the transition between tidal and freshwater influence.

Two-dimensional subsurface profiles derived from electrical resistivity tomography (ERT), shallow seismic refraction traverses, and ground penetrating radar (GPR) are used to laterally connect one-dimensional vertical sediment core data, and then integrated with the surface topography LiDAR data to construct channel and floodplain cross-sectional models. Past geomorphological processes – such as lateral migration, channel adjustment, and avulsion – are revealed in the preserved channel sediments, indicating responses to the contemporary climatic and anthropogenic conditions. Geophysical survey designs for identifying fluvial-geomorphological processes and features within palaeochannels are discussed, along with the need to adapt survey design for best resolution and depth in different, peat-dominated or clay-dominated, sedimentary settings. ERT is shown to consistently provide excellent depth penetration and estimations of channel extent. High resolution GPR data at the near-surface can be used in tandem with available core data to delineate the channel fill and bank geometry and calibrate depth estimations.

Flow conditions are reconstructed quantitatively using palaeohydrological drainage equations based on cross-sectional area values derived from the geophysical profiles. This avoids reliance on oversimplified cross-section estimates based upon surface parameters such as width and meander length, and one-dimensional depth estimates from core profiles. Existing hydraulic and drainage regime equations are tested against flow gauge data and channel measurements from active rivers to obtain optimal parameters for palaeohydrological calculations. These parameter estimates also benefit from on-the-ground channel parameter measurements in tandem with topographic remote-sensing. Hence, this study proposes a novel methodology that integrates geophysical surveying within palaeohydrological estimation techniques to improve models over long timescales of past fluvial environments that have been modified by humans.