Characterization of Seasonally Dependent Emergent Vegetation Variables for Coastal Impact Models

Abstract:

Emergent wetland vegetation has been shown to mitigate coastal inundation and erosion hazards by reducing wave energy through friction (Shepard et al., 2011), although its use in coastal protection planning is limited because predictive models require improved vegetation data. We isolated biophysical characteristics (biomass, stem density, rigidity, etc.) of plants using horizontal digital photographs (Side-On Photos) in conjunction with remote sensing and physical surveys. We studied the dominant salt-marsh species/assemblages in Port Susan Bay of Washington State, a vulnerable estuary that has experienced up to 1 kilometer of marsh retreat since the mid-1960s. We measured plant height, stem diameter, stem density (area available for flow) from fall to early spring (August 2014 through April 2015) using Side-On Photography and digital image processing techniques. Metrics from Side-On Photography were highly correlated to physical lab measurements. Vegetation rigidity was measured in-situ with a handheld digital scale with respect to measurement height and bending angle. Plant elasticity showed a strong correlation to stem diameter in two dominant bulrush species. We employed remote sensing supervised classifications techniques (Maximum-Likelihood and Decision Tree Classifiers) to hyperspectral imagery to map the spatial extent of vegetation assemblages with an overall accuracy of 86.7%. Combining these methods enabled us to extrapolate and validate vegetation characteristics across the study area and to estimate species-specific friction coefficients for input to cross-shore wave models. On-going studies include sensitivity analyses of wave models to seasonally-dependent vegetation parameters in the nearshore and ultimately wave impacts along the coast. By accounting for site-specific and spatiotemporal variability in vegetation data, we inform scientific understanding of the interactions of vegetation, waves, and sediment processes.