Towards an Improved Classification System of Dry-Fallen Intertidal Flats in the German Bight Using Microwave and Optical Remote Sensing Data

Remote sensing techniques allow for relatively cheap surveillance of large coastal areas and, thus, for meeting the requirements of coastal monitoring along the German North Sea coast. Optical sensors have already been successfully used for sediment classification purposes on intertidal flats, and promising results have been achieved through the classification of different sediment types, vegetation, and mussel beds. However, useful optical data from the German North Sea coast acquired by satellite-borne optical sensors at low tide are sparse, because of the strong dependence on daylight and cloud conditions. A classification system based on spaceborne remote sensing data would therefore strongly benefit from the utilization of synthetic aperture radar (SAR) data.
We have used different spaceborne SAR sensors to investigate the sensitivity of the radar backscattering to surface roughness variations on dry-fallen intertidal flats in the German Bight of the North Sea. We demonstrate that the radar backscattering from (wet) intertidal flats depends on their surface roughness properties, which vary due to, e.g., current-induced ripple formation, which in turn depends on the grainsize composition of the sediment. Further, benthic fauna such as blue mussels or (Pacific) oysters may cause an increase in surface roughness, when they are not covered by sediments. Thus, they may also be detected, and classified, by a SAR-based classification system.
An optimum coastal classification system, however, will use data from both optical and microwave sensors, which may be accompanied by a-priori knowledge and in-situ data. The combined use of optical and SAR sensors allows for gaining information about the optical properties and surface roughness properties of the different surface types quasi-simultaneously. This information, together with in-situ observations and time series of habitat maps, will yield an improved classification of different sediment types, as well as of mussel beds and seagrass. It is the aim of the studies presented herein to provide the basics for such an improved classification system.
The work is supported by the German national project DeMarine (50 EE 0817)