A sensor network for long-term monitoring of sediment transport in the coastal region

We introduce a sensor network suitable for measuring pressure and light transmission in coastal seas over time periods of several months. The purpose is to use the collected data for the investigation of sediment transport processes in the tidal flats of the North Sea in particular during extreme events like storm surges and high tides. The main development objectives were to conceive a modular system that is mobile and can be easily relocated in order to be able to respond quickly to emerging scientific demands. The system design builds up a controller system with CAN bus, which proved to be robust and reliable. Design considerations have been spent on what type of cable could be used under the expected adverse conditions and alternative designs using single buoy deployments have been evaluated as well. The electronic is designed to accommodate a variety of interfaces to be able to integrate the type of sensors that are important for the process study. To that end CAN bus is offering commercially available adapters that facilitate the interfacing process significantly. Special care has been taken to exclude a malfunction of the network if only one sensor fails. The pressure sensor has been selected based on considerations in regard to cost/performance ratio. Minimum requirements exist because of the needed fast and precise resolution in pressure to resolve wave heights. With the selected pressure sensor a prototype underwater device has been constructed and lab and field tests were carried out. A newly developed hyperspectral transmissometer allows deriving data on suspended matter concentration and size distribution. The sensor casing has been specifically designed to protect against biofouling. The capsulated sensors get only in contact with seawater on demand, e.g. prior to a storm surge. The sensor system will result in a better understanding of the dynamics of sediment transport and changes of seafloor morphology which cannot be observed with conventional met- ods during storm events. First results from deployments and algorithms for data processing algorithms are presented.