Semi-automated methods for analysing the parameters that controlling shore platform evolution: the example of Glamorgan, South Wales

This study developed a series of new, semi-automated methods for quantitatively analysing the contribution of rock weathering processes and geological control on shore platform evolution and boulder production. We used high resolution Digital Surface Models (DSMs) and orthoimages to map and classify a series of geomorphological and geological control and verify the main characteristics and evolution of the two analyzed layers (23 and 24) in Glamorgan, South Wales, UK. The workflow has three elements, firstly a complete geomorphic analysis of the entire exposure layers. Second, four 10x10 m polygons were created for each layer, spaced across the shore platform from upper to lower intertidal zones. Within these polygons we measured and classified two types of joints in relation to their depth, size and spatial continuity, pool areas, and several geomorphological parameters. Finally, we used DSAS software to measure platform edge erosion (1981-2018). In combination, these data allowed us to quantify: 1) erosion rates; 2) spatial variability in geomorphological features and geological control parameters; 3) PCA was used to quantify the importance of each variable; and 4) combine the platform data with the pyDGS data to quantitatively assess the contribution of shore platform into the rock coast sediment system.

Results show multi-decadal (1981-2018) erosion rates vary spatially and between layer, where layer 24 shows the higher erosion in seaward plot (-0.023 m yr^-1) while the layer 23 shows the higher erosion in the most landward (-0.027 m yr^-1) and seaward plots (-0.032 m yr^-1). High variability in landform processes and characteristics between the two layers and within one layer across the cliff – seaward edge gradient was also found. Layer 23 shows a decrease of roughness from cliff to seaward exposures of the layer, while in layer 24 the roughest zone is the seaward sector. PCA analysis shows that the three first components accounted for 78.9% of the variance. The most important parameters driving shore platform evolution (including both layers) in PC1 were distance from cliff and elevation with -0.6625 and 0.6615, respectively. There are relevant differences between both layers, in layer 24, distance from cliff have positive value (0.5875) and elevation negative (-0.5795), while  in layer 23 distance from cliff are negative (-0.5793) and elevation is positive (0.5825). These methods improve quantification of geological contingency, which can be applied to other rock coast systems.  This will help better model rock coast processes under different climate change scenarios.