Generating Numerical Model Grids Using Light Detection and Ranging (LiDAR) Data

In this paper we focus on the development of a model grid for a numerical approach to study sediment transport processes in Delaware Bay marshes. Marshes feature a high degree of spatial variability on very small spatial scales. Integration of tidal channels with widths of only a few meters is required to provide enough volume for the tidal prism. Consequently, a high resolution numerical grid is necessary to resolve the modeled marsh area. Marshes also have a large geospatial extent, no exact boundaries, and are often difficult to access. Thus, remote sensing technologies such as LiDAR (Light Detection And Ranging) are a helpful tool to provide data for topography and the location of tidal channels and other structures. For our study area, we utilized high resolution LiDAR data (2x2m) for the Blackbird Creek reserve in Delaware, which is located in the western shore of the Delaware Bay. LiDAR measures the highest elevation of existing elements, including ground, structures, and vegetation. Tidal flats are mostly covered by dense vegetation, but the ground elevation is needed for the model set up. We will present an approach that adjusts the existing LiDAR data by subtracting the height of the respective vegetation cover using a season-adjusted correction, thus generating a Digital Terrain Model (DTM) that represents only the ground elevation. The DTM will be used to map the respective topography on the model grid. Additionally, the LiDAR data is used to extract the outline of the first, second, and third order channels as polygons, which are used to triangulate the model grid appropriately.