Authors: David Retchless*, Texas A&M University at Galveston, Nicholas Wellbrock, Texas A&M University at Galveston
Topics: Cartography, Coastal and Marine, Global Change
Keywords: cartographic generalization, coastal change, climate change
Session Type: Virtual Paper
Start / End Time: 4:40 PM / 5:55 PM
Room: Virtual 45
Presentation File: No File Uploaded
We present a method to identify coastal embayments – and quantify coastal change therein – through repeated application of positive and negative buffers to the coastline. The method builds upon and automates the approach of Perkal (1958): epsilon-convex areas produced using a buffer series are used to find (and eliminate) bends along a coastal boundary. We use this epsilon-convex approach to identify coastal embayments. Using Google Earth Engine (GEE) and ArcGIS Model Builder, a buffer series is applied to Landsat-derived coastal surface waters to trim away un-embayed surface waters, leaving only embayed areas. The resulting polygons enclose surface water landward of the concave arc stretching across the mouth of each embayment. Larger buffer sizes (25km) typically produce a larger polygon with wider arcs, while smaller buffer sizes (5km) produce smaller polygons with narrower arcs. By applying an iterative series of buffer sizes, a nested hierarchy of embayments is developed: large-mouthed embayments identified via larger buffers contain many smaller-mouthed “sub- embayments” identified via repeated application of smaller buffers. The method supports development of “embayment order” datasets analogous to Shreve stream order (Shreve, 1966) as well as the further classification of embayed areas based on other criteria, e.g. fresh-water input. To identify coastal change hotspots – including effects of sea level rise and coastal development – comparisons are made across embayments of different orders and types as well as with coastal waters as a whole. The method is validated – and potential use-cases explored – through application to South Korean coastal waters.