Swan Island Restoration

Engineering With Nature (EWN) Principles In Practice

Why Nature-Based Solutions?

Sea level rise, coastal flooding, increased storm frequency, and other hazards pose increasing risks to our nation’s communities and natural resources. Nature-based solutions to infrastructure are those that incorporate natural features (e.g. dunes, reefs, marshes, and islands) for the purpose of flood and coastal storm risk reduction. Compared to conventional hardened infrastructure, nature-based solutions are more capable of adapting to a changing environment and can be self-sustaining. They also provide socio-economic and environmental benefits.

The National Centers for Coastal Ocean Science and The US Army Corps of Engineers, Engineering With Nature® (EWN®) Initiative work collaboratively at places like Swan Island, Maryland, to conduct research on the use of nature-based solutions to reduce risks from coastal hazards. Together, NCCOS and USACE are providing the, science, engineering and guidance to help coastal communities know how, where, and when to best construct nature-based solutions.

The Swan Island restoration is an example EWN 'principles in action' because it emphasizes the use of dredged sediment, natural processes, and ecosystems to achieve a broader suite of benefits for people and the environment.

Shown above are examples of natural and nature-based features, also known as natural infrastructure that provide the 'triple-win' benefits that align with the Engineering With Nature philosophy. Clockwise from left are an island, beach, seagrasses, intertidal marsh, oyster reef, and dune system.


Background

Coastal islands and marshes of the Chesapeake Bay, USA are disappearing at an alarming rate due to the combined impacts of sea level rise, subsidence, and inadequate sediment supply.

The last house on Holland Island in the Chesapeake Bay from Photo Credit Michael F. Young, https://sometimes-interesting.com/last-house-on-holland-island/.

At Swan Island, Maryland a 25 acre island within the Martin National Wildlife Refuge in Tangier Sound, Chesapeake Bay, high rates of shoreline erosion and subsidence have deteriorated the island’s natural habitat and its ability to shelter the nearby town of Ewell, Maryland from wave energy.

The red polygon indicates the areas of erosion for Swan Island (far left) and the Martin Wildlife Refuge since 1942.

At Swan Island, the strongest winds tend to blow from the northwest, making the Island a natural wave break for the nearby town of Ewell, MD.

Above is a compass rose showing the predominant wind directions and intensities near Swan Island from 2015 to 2020.

Restoring degraded islands can help to reduce erosion to nearby shorelines and communities while increasing habitat and recreational value.

The infographic on the right is an idealized rendering of the benefits of a restored island vs. a degraded island. The restored island in the bottom panel has a greater range of elevations that support a greater diversity of vegetative habitats and has greater capacity to shield the adjacent developed shoreline from wave energy.

Given the potential benefits of restored islands, we want to better understand the range of physical settings in which Islands like Swan can function effectively.

Island Erosion and Restoration from ERDC Corporate Communications on Vimeo


The Restoration

To counter island loss in the Chesapeake Bay, in 2019 the USACE Baltimore District placed 60,000 cubic yards of sediments, dredged from a nearby navigation channel, to restore the footprint of Swan Island. The restoration plan included creation of dunes and high and low marsh, and the installation of 200,000 plants. The island was transformed from a low and highly fragmented marsh to one with a wider range of habitats that sit higher in the tidal frame.

Sediment was dredged from a nearby navigation channel by a cutter head dredge and placed onto Swan Island via a pipe as shown in the photo above.

High marsh near the western edge of Swan Island just after planting in 2019.

High marsh on Swan Island in 2021, two years post-construction and planting.


The photo on the left shows Swan Island in 2017 prior to sediment placement and on the right in 2019 after sediment placement and planting.


The Partners

The Swan Island restoration provides the research team with an opportunity to quantify and document island performance and benefits to inform the future planning, design and adaptive management of islands and natural features. In collaboration with USACE’s Engineering With Nature® Program, staff members from USACE Engineer Research and Development Center (ERDC), USACE Baltimore District, NOAA’s National Centers for Coastal Ocean Science (NCCOS), Maryland Department of Natural Resources (MDDNR) and the U.S Fish and Wildlife Service (USFWS) came together to conduct the pre- and post-placement restoration monitoring required to quantify project success.


The Research

The restoration of Swan Island is expected to increase the long-term resilience of the Island to sea level rise and erosion, shelter nearby communities and shorelines from waves, and serve as habitat for a range of wildlife. Currently, the lack of quantitative data on the resilience and ecological benefits of natural features like islands are often cited as a barrier to their widespread use for coastal protection.


During a series of USACE ERDC facilitated workshops, the project team identified objectives, defined monitoring parameters, and developed a conceptual model to guide the research team in monitoring and the adaptive management efforts. The conceptual model reflects the primary objectives which are to evaluate the capacity of Swan Island to reduce waves and erosion to nearby shorelines (i.e. engineering performance) and the resilience of the island to storms, sea level rise and other environmental stressors. For example the project team will address questions such as:

1. How have the restoration actions influenced the capacity of Swan Island to provide protection from wave energy to the town of Ewell? and

2. How will the protective capacity and ecosystem services provided by Swan Island be influenced by sea level rise?

Dr. Brook Herman leads an ecological modeling discussion in the first of a series of workshops with the project team and other participants.

In a follow up meeting the project team came together to refine objectives and identify the primary monitoring parameters which resulted in the conceptual model shown on the right.


To meet project objectives, the team identified three dominant monitoring parameters as shown in the diagram at the right. These include waves and currents, vegetation biomass, Island profile (elevation) and available natural sediment within the water column.


Measuring Hydrodynamic, Ecological, Topographic, and Sediment Characteristics

The partners have monitored Swan Island annually for four years (2018-2021) to collect the hydrodynamic, topographic, ecological and sediment parameters necessary to evaluate project performance and benefits and inform adaptive management actions. Specific details of the monitoring and adaptive management are described in the Swan Island Monitoring and Adaptive Management Plan (in review, see below). These data will also inform the development and evaluation of integrated hydrodynamic and ecological simulation models.

The photo on the left was taken in 2019. The photo in the center is from a similar vantage point in 2021, two years post-construction. On the right shows the Spartina alterniflora (low marsh), that is nearly two meters tall in this area of the island.


Hydrodynamic and Water Quality parameters such as waves, currents, water level and total suspended solids (i.e. sediment particles) are collected from sensors stationed at four platforms around Swan Island. Locations are shown by the brown squares in the figure on the right.

Four platforms were installed and instrumented with sensors to measures waves, currents, water level and available sediment in the water column. (platform locations shown in map to the right)

Ecological and topographic surveys are conducted to document changes in elevation, vegetation and sediment over time as the site matures. Data are collected annually from fixed sampling locations indicated by circles on the map to the right.

Using a meter squared quadrat, researchers collect data on plant species, percent cover, canopy height and elevation, at predefined points as shown in the figure at the right.

At a subset of these sites, sediment cores are also obtained to better understand the change in sediment composition over time.

RTK or Real-Time Kinematic GPS is used to collect accurate location (lattitude, longitude) and elevation data (< 2 centimeter accuracy) at each of the points on the island.


Measuring Sediment Accretion

The roots and rhizomes of vegetation have the capacity to stabilize sediments while their blades have a baffling effect on waves and currents which help to trap sediment and reduce erosion at the sediment-water interface. Below, feldspar clay is placed on the sediment surface (left). Two years later, a core collected in that same location (middle) reveals the amount of new sediment that has been deposited on top of the feldspar layer (right), providing a measure of sediment accretion over time.

Photo on the left shows Dr. Jenny Davis placing feldspar in three 0.25 meter squared quadrats in 2019 to later measure the amount of sediment accretion via coring. In 2021, two years later, the amount of sediment accretion ranged from 10.4 to 41.3 mm in the high marsh zone.


Drone-based data collection

Drone-based imagery georeferenced with RTK-GPS is collected annually to document change over time. Sturcture from Motion image processing is used to generate whole-island Digital Elevation Models (DEMs), Habitat Classification Maps, and to measure shoreline change. These data are critical to detecting change over time on the island and foundational to development and evaluation of predictive simulation models (see below).

The NCCOS drone pilot Ryan Giannelli, flies the drone at altitudes up to 120 meters to collect the island imagery. Data are used to create habitat classification maps and digital elevation models that are key to evaluating change over time and developing and evaluating predictive simulation models.


From left to right these aerial images show Swan Island 2017 (pre-placement), 2019, 2020 and 2021 (post-placement).


Elevation change between 2019 and 2021 are shown in the surface model on the right. Note the greatest elevation change, shown in green is due to vegetation growth. Subsequent analysis will correct for the height of vegetation to understand the change in sediment surface elevation.

Habitat classification maps such as the one shown on the right, created with 2020 imagery, will be used to detect and quantify change over time.

 


The Models

The monitoring data are being used to develop predictive simulation models such as the Marsh Transect Model and the ADCIRC+STWAVE models to test hypotheses and answer questions about island performance, resilience and ecological services under a variety of sea level rise and storm conditions.

Marsh Transect Model

The marsh transect model simulates the processes of vegetative growth (above and below ground biomass), sediment accretion and erosion under various sea level rise and tide scenarios to predict island evolution. The inputs, outputs, and preliminary results are shown below.

This diagram shows the inputs and outputs from the Marsh Transect Model.

Analysis of elevation, plant growth, and water level data from 2019 on the left and the predicted (in 10 years) elevations on the right from the Marsh Transect Model. These preliminary data show decadal-scale marsh persistence of the island.


Coastal STORM Modeling System - CSTORM

ADCIRC (Advanced CIRCulation) was coupled with the STWAVE (Steady-state Wave) Model to simulate water elevations and wave heights during storm events. Preliminary results from a Hurricane Sandy simulation model run are shown below. The results from this modeling effort will allow the project team to quantify the capacity of Swan Island to reduce storm risk to nearby shorelines.

Using the CSTORM model currents, waves, and water levels are calculated at each of the red points in the figure shown on the right under a variety of sea level rise and storm scenarios.

This diagram shows the inputs and outputs for the CSTORM Model.

The results from a simulated Hurricane Sandy model run show the effect of Swan Island in reducing water level which is most pronounced just south and east of the island.

The image on the right shows the reduction in water level due to the presence of Swan Island as depicted by the cooler colors shown in blue.


Report is currently in review.

Adaptive Management

The project team developed a Monitoring and Adaptive Management Plan (MAMP) to track progress and serve as a blueprint for the monitoring (e.g. What, How, and How often data is collected) and the adaptive management approach. This includes reporting, data management, roles and responsibilities, performance metrics and decision thresholds for adaptive management.

As part of the adaptive management strategy an additional 22,000 Spartina alterniflora plants were installed in June 2021 by USFWS. In the photo on the right Dr. Jenny Davis observes the experimental 'clumped' plantings where 9 plants were installed on 1 meter centers as opposed to one plant on 18 inch centers. Survival of each planting type will be compared to better understand best practices for planting success.

The Takeaway

The results of this research effort will address current information gaps and uncertainties surrounding the protective benefits and impacts of island restoration activities on surrounding ecosystems, and will be used to inform adaptive management of the Island as it matures over time. The approach developed here will be applicable to other island systems.

Multi-disciplinary partnerships are critical to leverage both the expertise and the resources required to monitor, evaluate, and manage island or other natural systems.

Long-Term Function of Coastal Islands Derived from Engineering With Nature® Efforts from ERDC Corporate Communications on Vimeo

Credits

Jenny Davis, Paula Whitfield, Ryan Giannelli, Michael Greene, Leanne Poussard, NOAA NCCOS

Amanda Tritinger, Jeff King, Joe Gailani, USACE ERDC

Rebecca Golden, MDDNR

Matt Whibeck, USFWS

Danielle Szimanski, USACE Baltimore District

Marsh Transect Model

Doug Krafft USACE ERDC

CSTORM Model

Amanda Tritinger, S. Catie Dillon, USACE ERDC

This diagram shows the inputs and outputs from the Marsh Transect Model.

Report is currently in review.

The red polygon indicates the areas of erosion for Swan Island (far left) and the Martin Wildlife Refuge since 1942.

Above is a compass rose showing the predominant wind directions and intensities near Swan Island from 2015 to 2020.

Sediment was dredged from a nearby navigation channel by a cutter head dredge and placed onto Swan Island via a pipe as shown in the photo above.

High marsh near the western edge of Swan Island just after planting in 2019.

High marsh on Swan Island in 2021, two years post-construction and planting.

Dr. Brook Herman leads an ecological modeling discussion in the first of a series of workshops with the project team and other participants.

In a follow up meeting the project team came together to refine objectives and identify the primary monitoring parameters which resulted in the conceptual model shown on the right.

Four platforms were installed and instrumented with sensors to measures waves, currents, water level and available sediment in the water column. (platform locations shown in map to the right)

Using a meter squared quadrat, researchers collect data on plant species, percent cover, canopy height and elevation, at predefined points as shown in the figure at the right.

At a subset of these sites, sediment cores are also obtained to better understand the change in sediment composition over time.

RTK or Real-Time Kinematic GPS is used to collect accurate location (lattitude, longitude) and elevation data (< 2 centimeter accuracy) at each of the points on the island.

This diagram shows the inputs and outputs for the CSTORM Model.