Traditionally, coastline-migration maps, showing erosion and accretion at the fringes of marine and lacustrine water bodies, have been made using field-monitoring data, supplemented locally by analyses of aerial photography. The EMODnet-Geology coastline-migration map is a good example. It provides an overview of our current knowledge and is useful in transnational coastal-zone management. At the same time, it is far from perfect.
First, it combines monitoring series from different time periods, necessitated by an overall scarcity of data. For some regions, the only information available may be decades old. Second, it includes data gaps as many remote regions have never been surveyed. Quite a few of those gaps have been filled using expert knowledge from coastal geologists, but there are extensive stretches that must be left unfilled for the time being. Third, the coastline is not a single line that is easily and consistently recognized and used. Along sandy coasts, one can map the low-tide line, the high-tide line and the dune foot, and each of these indicators may behave differently through time. Along bluff and cliff coasts, one can map the bluff/cliff top and base, or the actual low- or high-water line some distance away. When part of a bluff or cliff collapses and thus recedes landward, its rubble accumulates at the base, moving the water line seaward.
The public availability of satellite data and new analytical tools for processing big data, such as the Google Earth Engine, enable us to look at coastline migration in a new way. Scripts for automated detection of the land-water boundary generate numerous data points for each part of the European coastline. When averaged by year and analyzed for a decadal period, these data points form the basis for a new pan-European coastline-migration map that covers a consistent time period relevant for present-day coastal-zone management, eliminates data gaps, and portrays a single coastline indicator that is assumed to correspond to the mid-tide land-water boundary. As part of EMODnet-Geology, Gerben Hagenaars at Deltares performed an analysis for tens of thousands of transects with a spacing of 500 meters, giving a map resolution of 1:1,000,000.
The need for validation
As the pixel resolution of individual satellite images is about 10 meters, the precision of the method is still limited. Calculating annual averages from multiple measurements within a year reduces this uncertainty, and validation for sandy coastlines with beaches shows the method’s accuracy. The usefulness of the resulting data is discussed in ‘The state of the world’s beaches’, an article published in Scientific Reports. For bluff and cliff coasts, such validation and further analysis still need to take place. Once published as an EMODnet Web Feature Service, we are keen to invite all users to compare our satellite-based output with corresponding field-monitoring data.
For further information, please contact Sytze van Heteren.