Classification of soft cliff dynamics using remote sensing and data mining techniques

Coastal soft cliffs are subject to changes related to both marine and subaerial processes. It is imperative to comprehend the processes governing cliff erosion and develop predictive models for effective coastal protection. The primary objective of this study was to bridge the existing knowledge gap by elucidating the intricate relationship between changes in cliff system morphology and the driving forces behind these changes, all within the context of ongoing climate change. Therefore in this study, we employed various quantitative numerical methods to investigate the factors influencing coastal cliffs and the adjacent beaches. Our analysis involved the extraction of several morphological indicators, derived from terrestrial laser scanning data, which were then used to assess how cliffs respond to extreme weather events. The data span two winter storm seasons (2016-2018) and encompass three soft-cliff systems situated along the southern Baltic Sea, each characterized by distinct beach and cliff morphology. We conducted a detailed analysis of short-term cliff responses using various data mining techniques, revealing intricate mechanisms that govern beach and cliff changes. This comprehensive analysis has enabled the development of a classification system for soft cliff dynamics. Our statistical analysis highlights that each study area exhibits a unique conditional dependency between erosion processes and hydrometeorological conditions, both during and between storm events. Furthermore, our findings underscore the vulnerability of cliff coastlines to extreme water levels and episodes of intense precipitation.