Accretion, retreat and transgression of coastal wetlands experiencing sea-level rise

Abstract

The vulnerability of coastal wetlands to future sea-level rise (SLR) has been extensively studied in recent years, and models of coastal wetland evolution have been developed to assess and quantify the expected impacts. Coastal wetlands respond to SLR by vertical accretion and landward migration. Wetlands accrete due to their capacity to trap sediments and to incorporate dead leaves, branches, stems and roots into the soil, and they migrate driven by the preferred inundation conditions in terms of salinity and oxygen availability. Accretion and migration strongly interact, and they both depend on water flow and sediment distribution within the wetland, so wetlands under the same external flow and sediment forcing but with different configurations will respond differently to SLR. Analyses of wetland response to SLR that do not incorporate realistic consideration of flow and sediment distribution, like the bathtub approach, are likely to result in poor estimates of wetland resilience. Here, we investigate how accretion and migration processes affect wetland response to SLR using a computational framework that includes all relevant hydrodynamic and sediment transport mechanisms that affect vegetation and landscape dynamics, and it is efficient enough computationally to allow the simulation of long time periods. Our framework incorporates two vegetation species, mangrove and saltmarsh, and accounts for the effects of natural and manmade features like inner channels, embankments and flow constrictions due to culverts. We apply our model to simplified domains that represent four different settings found in coastal wetlands, including a case of a tidal flat free from obstructions or drainage features and three other cases incorporating an inner channel, an embankment with a culvert, and a combination of inner channel, embankment and culvert. We use conditions typical of south-eastern Australia in terms of vegetation, tidal range and sediment load, but we also analyse situations with 3 times the sediment load to assess the potential of biophysical feedbacks to produce increased accretion rates. We find that all wetland settings are unable to cope with SLR and disappear by the end of the century, even for the case of increased sediment load. Wetlands with good drainage that improves tidal flushing are more resilient than wetlands with obstacles that result in tidal attenuation and can delay wetland submergence by 20 years. Results from a bathtub model reveal systematic overprediction of wetland resilience to SLR: by the end of the century, half of the wetland survives with a typical sediment load, while the entire wetland survives with increased sediment load.