Mapping erosion along CT shoreline is difficult due to the access challenges presented by the marshes. Additionally, the change detection is monitored in centimeters so very accurate – survey grade – models are required to identify changes in thin layer placement. Further, multiple flights annually over a large area of interest is required for useful change detection results.

Objective: Autonomously map and provide change detection metrics for elevations in the marshes. Track thin layer placement changes and identify the success rate of the manually increased sediment elevations for protecting identified assets such as Salt Marsh nesting sites.

Using machine learning, autonomously identify vegetation types like Phragmites, Spartina Patens, Spartina Atlernaflora, and Salicornia, then track changes in health and location over time.

Challenge: Gaining baseline data of existing thin layer sediment, the best known locations of Salt Marsh Sparrow nest sites, and mapping where and what types of sea grass are currently growing across the area of interest has in the past been a very manual task, which historically required human access to the marsh. Trending analysis needs data to be logged frequently over time; therefore, a more efficient approach of gathering the information on a recurring basis is required. Our challenge was to confirm that flying high-precision UAS to aerially collect data that had previously been collected solely from the ground could capture metrics of these marsh assets at an equivalent or even improved level.