Fuzzy Cognitive Modeling in Coastal Geomorphology

Abstract

Coastal systems are inherently complex and involve multi-scale feedbacks and interactions that are difficult to quantify. Coastal geomorphologists organize their understanding of these interactions and feedbacks through conceptual models based on their time in the field or laboratory and research networks. While coastal geomorphology is rich in theory and concepts related to time, process, systems and landforms, there have been few attempts to quantitatively evaluate this conceptual knowledge and predict complex spatial patterns. This article reviews the potential for conceptual modeling using Fuzzy Cognitive Maps (FCM), a form of artificial intelligence that merges fuzzy logic and neural computing, to be applied in coastal geomorphology. FCM development, testing, and interpretation is demonstrated using examples for marsh response to sea level rise, eolian transport from the beach to foredune, the establishment of dune-building vegetation in the backshore, barrier island washover and resiliency, and controls on rip current velocity. Additionally, further integration with Cellular-automata (CA) modeling is shown to be an opportunity to develop evolutionary FCMs to predict landscape change. Based on the ability to intuitively incorporate concepts related to physical and social dimensions of coastal change means that FCM can be applied broadly for integrated coastal zone management (ICZM), and as an innovative way to train the next generation of coastal geomorphologists.