Spatially explicit modeling of coastal vegetation change associated with projected sea level rise: The Potomac estuary

Elmore, Andrew; Cadol, Daniel; Guinn, Steven; Sanders, Geoffrey; Engelhardt, Katia; Fitspatrick, Matthew

Coastal environments are expected to respond to rising sea levels through migration inland. This process is limited by the availability of corridors of sufficiently flat, undeveloped land to be converted to wetland. Developed land protected from tidal influence through the construction of bulkheads and levies and natural areas of elevated land will decrease the area available for conversion to marsh and wetland forest over time, leading to a loss of biodiversity. Because these processes depend strongly on the spatial configuration of vegetation-elevation relationships, they must be modeled within a framework that accounts for the specific elevation ranges over which different vegetation classes persist, and elevation change (accretion net of settling and compaction) across the full range of affected elevations. For the National Parks along the Potomac River Estuary, models must operate at both a high spatial resolution and over broad spatial extents to capture changes at scales relevant to park resource managers. This project produced a spatially explicit computational model (termed the Marsh Accretion and Inundation Model (MAIM)) and model results predicting the impact of user-defined sea-level rise scenarios on vegetation. The model takes as input detailed (1-m resolution) map layers representing elevation (generated from LiDAR) and initial vegetation classification. Vegetation classes are constrained to predetermined individual elevation ranges, based on NPS vegetation maps, plot inventory data, and digitalization of aerial photography. MAIM is not a dynamic model and therefore does not alter elevation change in response to accelerated sea-level rise or any other environmental conditions; MAIM assumes that the relationship between accretion and elevation observed in historical data is adequate for modeling future conditions. It also does not account for hydrologic interaction with the watershed that might be expected to increase inundation times in areas of high flow accumulation area. MAIM is not a sediment dynamics model, and therefore cannot respond to changes in suspended-sediment concentration in estuarine waters. Finally, MAIM does not model shoreline erosion, as this was not found to be a dominant predictable process over the majority of the study area. The combination of carefully parameterized modeled processes and simple run-time efficiency make MAIM an ideal tool for resource managers seeking to predict changes in biological resources over the next 50 to 100 years. The model output represents detailed vegetation change projections at 10-year intervals, including the probability of occurrence for each modeled grid cell. Output also includes tables and line graphs that can be used to analyze trends and divergence in probability results over time. All results are included for 3 sea-level rise scenarios, all based on the best available scientific projections of future conditions. Because the study region includes extensive areas of developed land within 5m of current sea level, we also provide two case-scenarios, one in which developed land is protected from the effects of sea-level rise through the construction of bulkheads and levies and a second in which developed land is abandoned as it becomes inundated, leading to conversion to marsh habitat. Overall, the results from each scenario provide a unique landscape-level perspective of the impacts of sea-level rise for which park resources managers must be prepared.

Spatially explicit modeling of coastal vegetation change associated with projected sea level rise: The Potomac estuary

Series

Collections