Hurricane Sandy impacts on coastal wetland resilience

Abstract

The goal of this research was to evaluate the impacts of Hurricane Sandy on surface elevation trends in estuarine marshes located across the northeast region of the United States from Virginia to Maine using data from an opportunistic (in other words, not strategic) and collaborative network (from here on, an opportunistic network) of surface elevation tablemarker horizon (SET-MH) stations. First, we built a database of metadata for 965 individual stations from 96 unique geographical locations that included the location, geomorphic setting, and wetland type for each SET-MH station. The dominant estuarine settings included in the analyses were back-barrier lagoonal marshes and emergent marshes along embayments and tidal tributaries. We then calculated prestorm elevation trends to compare to poststorm elevation measurements to determine the storm impact on each station trend. We hypothesized that the effect of Hurricane Sandy on marsh elevation trends would differ by position relative to landfall (right or left) and distance from landfall in southern New Jersey, as both of these variables influence the presence or absence of storm surge as a result of the physical characteristics of tropical cyclones (in other words, strongest winds typically occur to the right of landfall). Storm surge was spatially less extensive and less deep (~1 meter [m]) in marshes located to the left (in other words, south) of landfall compared to marshes 1U.S. Geological Survey. 2University of Minnesota, Duluth, Minn. 3U.S. Fish & Wildlife Service, Rachel Carson National Wildlife Refuge, Wells, Maine. 4Yale University, School of Forestry & Environmental Studies, New Haven, Conn. 5University of Maryland, College Park, Md. 6Natural Resources Conservation Service, Amherst, Mass. 7Fairleigh Dickinson University, Teaneck, N.J. 8University of Virginia, Charlottesville, Va. 9Jackson Estuarine Laboratory, University of New Hampshire, Durham, N.H. 10U.S. Fish and Wildlife Service, Eastern Virginia Rivers National Wildlife Refuge Complex, Warsaw, Va. 11Save the Bay, Narragansett, R.I. 12New York State, Department of Environmental Conservation, Hudson River National Estuarine Research Reserve, Staatsburg, N.Y. 13Wells National Estuarine Research Reserve, Wells, Maine; Center for Marine and Environmental Studies, University of the Virgin Islands, St. Thomas, V.I. 14Meadowlands Environmental Research Institute, Lyndhurst, N.J. 15New York City Department of Parks & Recreation, New York, N.Y. 16Partnership for the Delaware Estuary, Wilmington, Del. 17Chesapeake Bay National Estuarine Research Reserve, Gloucester Point, Va. 18National Park Service, Northeast Coastal and Barrier Network, Washington, D.C. 19 The Nature Conservancy, Cold Spring Harbor, N.Y. 20Barnegat Bay Partnership, Toms River, N.J. 21U.S. Fish and Wildlife Service Northeast Region, Bombay Hook National Wildlife Refuge, Smyrna, Del. 22Waquoit Bay National Estuarine Research Reserve, Waquoit, Mass. 23New England and Interstate Water Pollution Control Commission, New York State Department of Environmental Conservation, Long Island Sound Study, East Setauket, N.Y. 24Academy of Natural Sciences of Drexel University, Philadelphia, Pa. 25Narragansett Bay National Estuarine Research Reserve, Prudence Island, R.I. 26Delaware Department of Natural Resources and Environmental Control, Delaware Coastal Programs, Dover, Del. 27Chesapeake Bay National Estuarine Research Reserve in Maryland, Maryland Department of Natural Resources, Annapolis, Md. 28University of Maryland Center for Environmental Science, Horn Point Laboratory, Cambridge, Md. 29Edwin B. Forsythe National Wildlife Refuge, Oceanville, N.J. 2 Hurricane Sandy Impacts on Coastal Wetland Resilience located to the right (in other words, north) of landfall where storm surge covered a larger area and was deeper (3–4 m). About 63 percent of 223 eligible stations had a poststorm trend that was similar to the prestorm trend (in other words, less than ±5 millimeters [mm]), indicating little storm impact on elevation trends at those sites. The remaining 37 percent of stations exhibited significant poststorm deviations from the prestorm trend (in other words, greater than ±5 mm). Of these, stations located to the left of landfall had a significant and greater deviation in their elevation trend, and the deviation was more likely to be positive (elevation gain) compared to marshes located to the right of landfall, which had a significant deviation in their elevation trend that was more likely to be negative (elevation loss). This finding is directly related to storm surge impacts on marsh sediment deposition, where deep storm surge (3–4 m) results in sediment deposition in habitats inland of coastal marshes but less so in the marshes themselves. Substrate compaction by the storm surge overburden may have contributed to elevation loss, but this was not measured because sufficient marker horizon data were not available for analysis. In contrast, to the left of landfall the wind-driven flooding of sediment laden water pushed into the headwaters of rivers and small bays with an ~1 m surge, and resulted in more prevalent sediment deposition on the marsh surfaces and elevation gain. In general, the findings support previous research showing that the physical characteristics of the storm (for example, wind speed, storm surge height, impact angle of landfall) combined with the local wetland conditions (for example, marsh productivity, groundwater level, tide height) are important factors determining a storm’s impact on soil elevation, and that the soil elevation response can vary widely among multiple wetland sites impacted by the same storm and among different storms for the same wetland site. The final objective of this project was to create a framework using metadata from the opportunistic network of SETMH stations that could be used to develop a strategic monitoring network designed to address specific climate change impacts and related phenomena identified by land managers and stakeholders. We evaluated the spatial distribution and density of SET-MH stations in relation to geographic coverage, marsh setting, availability of public land, and historical storm surge footprints and hurricane return intervals in order to identify gaps in our understanding of risk and our ability to assess it. Analyses revealed that the general geographic coverage of SET-MH stations is limited given the low percentage of marsh patches with stations, low density of stations, the clumped distribution of stations, and the often limited and uneven distribution of stations in wetlands with a high historical frequency of hurricane strikes and storm surge impacts. These findings can be used by managers and planners to inform the creation of a strategic monitoring network that can, in turn, inform management and adaptation plans for coastal resources in the region. Final plan designs will need to consider financial and infrastructural support required for station maintenance, as well as data collection and management over the long term. Introduction On October 29, 2012, Hurricane Sandy made landfall as a large (1,611 kilometers [km] diameter) post-tropical cyclone near Brigantine, New Jersey, with maximum sustained winds of 130 kilometers per hour (km/h) (70 knots [kn]) and minimum pressure of 94.5 kilopascals (kPa) (945 millibars [mb]; Blake and others, 2013). The storm affected, in varying degrees, the entire Atlantic seaboard of the United States from Florida to Maine (Blake and others, 2013; Valle-Levinson and others, 2013). Understanding the ecological and physical impacts of hurricanes on coastal wetlands and their interaction with local conditions is important for identifying resilience of these marsh communities. In light of the projected increase in the frequency of intense storms and in storm intensity (Knutson and others, 2010; Bender and others, 2010; Peduzzi and others, 2012; Emanuel, 2013; Horton and others, 2011, 2014), the physical (for example, storm surge, sediment deposition) and chemical (for example, salinity, pollutants) impacts associated with hurricanes need to be understood to efficiently and effectively protect and restore these critical habitats. Some of the potential long-term impacts of low-frequency, highmagnitude storm events such as Hurricane Sandy on marsh surface elevation include sediment deposition and erosion, storm surge-related soil compaction, and altered community dynamics, which includes mortality of existing vegetation or promotion of growth through increased availability of nutrients from newly deposited sediment (Cahoon and others, 1995a; Cahoon, 2006). Death of wetland vegetation can lead to loss of elevation by increased root zone decomposition and erosion, whereas increased root growth can lead to elevation gain by root zone expansion (Cahoon and others, 2003). The goal of this study was to evaluate Hurricane Sandy’s shortterm impacts on marsh surface elevations within 1 year after the storm. To assess the impacts of Hurricane Sandy on marsh surface elevation dynamics, we evaluated elevation and accretion data collected with the surface elevation table-marker horizon (SET-MH) method (Cahoon and others, 1995a) throughout marshes of the northeastern region of the United States both before and after the storm. The SET-MH method directly measures both surface elevation change (SET) and vertical accretion (MH). These two measurements can be used to calculate subsurface process influences on elevation change, namely shallow subsidence or shallow expansion (for example, root zone expansion from enhanced root growth; Cahoon and others, 1995b, 2002a,b; Callaway and others, 2013). A more detailed description of the SET device and the marker horizon method, including photographs and a diagram, is provided in appendix 1. Previous research using the SETMH method has shown that hurricanes are powerful agents of geomorphic change (Cahoon, 2006), resul