Rates and Drivers of Coastal Cliff Erosion in England from 2 Decades of Lidar Observations

Coastal cliff erosion affects communities worldwide. In England, cliffs make up approximately 54% of the coastline, a vast proportion of which is of weak to very weak rock resistance. Being able to adequately quantify rates and styles of retreat as well as relating these to their driving forces is key to making informed decisions on future coastal management strategies.

Nationwide change analysis, spanning more than 20 years in places, was carried out by comparing the earliest and most recent LiDAR Digital Elevation Models (DEMs) using raster differencing in a GIS environment. Erosion rates over the cliff face were derived by excluding any bias caused by the presence of vegetation from the change analysis. Uncertainty of erosion rates is affected predominantly by the precision of the LiDAR sensors (typically ±15cm in elevation), and was derived at each location by comparison with a network of ground control points. Limitations of a DEM differencing approach are widely discussed in the literature, particularly in areas of high relief where overhanging and undercutting features may be present. Nonetheless, this large-scale assessment provides a consistent approach to estimate erosion rates and was used as a basis for the selection of sites at which more extensive topographic analysis was carried out in 3D, overcoming some of the limitations of a raster differencing approach. Sites were selected to capture the variability in intrinsic (e.g., geology, slope, aspect) and extrinsic (e.g., waves, tides, rainfall) drivers of retreat. An open-source Python workflow was developed by integrating the M3C2 plugin for Cloud Compare to calculate distances between subsequent timesteps of point cloud data, after which the volume of meshed erosion clusters was computed. As a result, an inventory of cliff face erosion scars was compiled, with attributes relating to scar shape, area, volume and elevation relative to cliff top height. A comparison of the two approaches for calculating erosion rates highlighted the ability of the 3D workflow to capture cliff failure in greater detail and with higher accuracy, at the expense of increased computational costs. Findings also indicate spatio-temporal variations in erosion patterns which are related not only to the intrinsic nature of each site, but also to its climatology. It is expected that the broad spatial scale of this research will provide some insights into the main styles and drivers of retreat affecting specific stretches of coastline in England.