Understanding Long-Term Changes of Coastal Saltmarshes from Satellite Remote Sensing

Saltmarshes are known to be very important coastal ecosystems. They provide crucial functions for flora and fauna, as well as valuable ecosystem services for humankind. Many methods that are used to investigate these ecosystems are limited in space and time. Long time series of global satellite data enable to observe changes in the extent of saltmarshes on a large scale and over a long time period. We developed an unsupervised decision tree classification method in Google Earth Engine that automatically classifies satellite images into saltmarsh vegetation, mudflats, and open water. We applied the method using Landsat 5 TM data between 1985 and 2011. With this, we are able to detect trends in the seaward extent of saltmarshes globally. We reveal transitions between saltmarsh, mudflat and open water. Furthermore, we put saltmarsh habitat changes in a spatial context and couple trends in saltmarsh dynamics to environmental drivers, such as sea level rise, tidal forces, waves, and sediment availability.