Quantifying Uncertainty in the Protective Values of a Marsh Ecosystem and Ramifications for Marsh Restoration Projects

Coastal ecosystems such as marshes provide multiple protective benefits to communities affected by coastal hazards like hurricanes. While water level reduction by various processes is the most commonly cited protective benefit, vegetation also protects the coastline from erosion by stabilizing sediment. Most existing models for simulating storm impacts on vegetated shores use a fixed bathymetry and are not able to capture sediment stabilization. This means existing models may underestimate the risk mitigation benefit of marsh ecosystems over time. The goal of this study is to determine which model inputs and assumptions, including bathymetry, vegetation heterogeneity, and hazard intensity, are driving the uncertainty in estimated protective values.

We hypothesize that sediment stabilization is the more important protective benefit of marshes over time as compared to water level reduction. To test our hypothesis, we run a Monte Carlo simulation in which the marsh geometry, vegetation, and hazard are randomly generated. Marsh geometry is created using typical values for marsh platform slope, intertidal channel dimensions, and sinuosity. We use Sequential Gaussian Simulation to randomly generate heterogeneous vegetation fields from field data. A 2D shallow water equation solver then simulates hazards of various intensities, ranging from a tide to a hurricane surge, which are forced as an idealized sine wave at the boundary of the marsh domain. The amplitude and wavelength of the forcing is also randomized using water level data collected in the Chesapeake Bay. We then calculate the sensitivity of protective value, in terms of water surface elevation reduction, to the random inputs using Distance-Based Generalized Sensitivity Analysis.

Our preliminary results suggest that the reduction in water level provided by the marsh is most sensitive to marsh geometry. We calculate the sensitivity of mass flux into the marsh and find that the morphological consequences of marsh degradation such as channelization and erosion lead to noticeable reduction in the risk mitigation provided by the marsh. A potentially important ramification of this finding is that the value of marsh restoration might be underestimated by current assessments that do not account for associated changes in coastal morphology.