Dynamic response of coastal landforms to sea-level rise: representing lateral wetland processes in a probabilistic framework

Abstract:

Future coastal hazard assessments seldom account for differing ecological and geomorphic conditions that govern the landscape response to physical and anthropogenic drivers. For example, a barrier island with ample sediment supply will dynamically respond to sea-level rise by accreting sediment and transgressing landward, while a low elevation salt marsh with no external supply will not respond dynamically and will inundate. The Coastal Response Likelihood model incorporates spatially variable land cover, elevation, and relative sea-level rise at 30-m resolution in a Bayesian framework to predict the probability of dynamic landscape response to future sea-level rise. The response likelihood of wetland land classes assesses adaptive potential using elevation relative to sea level, with higher elevation wetlands having a greater likelihood of dynamic response, i.e. accretion and growth, when compared to lower elevation wetlands that are likely to inundate. We added a wetland-specific metric to incorporate sediment budgets and lateral marsh stability, via the unvegetated-vegetated marsh ratio (UVVR), to constrain the dynamic response of salt marsh environments. Using the Chesapeake Bay region as an example, we quantified the UVVR from Landsat imagery and used it as a proxy for the sediment budget, dynamic response likelihood, and lifespan of salt marshes. Adding this element provides a larger range of dynamic response by accounting for lateral processes in addition to vertical processes. For example, because of their greater sediment trapping potential and increased resistance to open-water conversion, low-elevation marshes with intact vegetated plains (i.e. low UVVR) demonstrate higher likelihoods of dynamic response. Similarly, high-elevation marshes with disintegrating marsh plains have lower likelihood of dynamic response, as the lack of vegetation will not allow for lateral stability or growth. The enhanced prediction demonstrates model flexibility in incorporating new data streams that account for both vertical and lateral marsh stability mechanisms. New UVVR and lifespan data can be combined with marsh-forest transition likelihood within the model to refine predictions with local conditions and provide a more nuanced picture of dynamic coastal response.