Sea Level Rise Effects on Roads & Marshes

The Sea Level Rise Effects on Roads and Marshes projects is a project between CT DEEP and UConn CLEAR.

To better understand how Connecticut’s coastal area marshes and roads may respond to sea level rise (SLR), Sea Level Affecting Marshes Model (SLAMM) was applied to Connecticut’s shoreline by Warren Pinnacle Consulting, Inc. This viewer displays the model’s results for Connecticut’s 21 largest marshes and all coastal area roads. For marshes, SLAMM was run two hundred times using alternative model inputs for key model parameters, including various SLR scenarios, to produce probability based results of likely future marsh conditions. For roads, the model uses a single SLR scenario of about 4.2 feet by 2100 from the base year of 2002.

Limitations

The Sea Level Affecting Marshes Model (SLAMM) is a widely adopted and effective model to predict wetland response to long-term sea-level rise and has been applied in every coastal  state in the U.S. However, as with any model’s attempt to predict change within a complex natural system, it’s important to understand the model’s limitations.

  • One important potential limitation is the uncertainty associated with some of model input parameters, such as sea level rise rates. SLAMM used the best available regionally adjusted range of rates available at the time the model was run in 2016.
  • SLAMM is not a hydrodynamic model that considers rates of water flow or ecological feedback mechanisms.
  • Land cover data used to distinguish developed versus undeveloped land, factors used to identify areas capable of supporting marsh migration onto uplands, was coarser than the model’s 5-meter resolution land cover change results.
  • SLAMM likely overestimates the area of new marsh as it does not account for future development of currently undeveloped land that the model assumes will be available to accommodate marsh migration.
  • Water level information used to identify road flooding is based upon local National Oceanic and Atmospheric Administration (NOAA) tide gauge data and 2013 Federal Emergency Management Agency (FEMA) still water coastal storm surge data.
  • SLAMM's road flooding module does not currently indicate future road flooding water depths. Therefore predicted future road flooding cannot be equated with road access/egress restrictions.
  • SLAMM is a general planning tool to identify areas to focus further coastal wetland and road flooding management investigations and is not intended for use for regulatory purposes
  • The model only generally estimates the effect of tidal restrictions (e.g. tide gates, culverts, etc.).
  • Restrictions to the flow of tidal water in and out of marshes by structures such as tide gates and culverts ‘mutes’ the tide’s amplitude which can influence the type and extent of coastal marsh inland of these restrictions. See the SLAMM Final Report sections 2.11.2 and 3.5 for more on how SLAMM considered the effects of tidal restrictions on marsh modeling results.

How SLAMM Works

How SLAMM Works

The Sea Level Affecting Marshes Model (SLAMM) is a widely adopted computer model used to predict long term change in shoreline habitats and flooding frequencies as a function of land elevation, tide range, sea level rise (SLR), and other environmental factors.

Two Types of Results

The model is capable of generating results in two ways. The first is by using specific model input values. The second employs SLAMMs uncertainty module, which generates results in terms of their likelihood of occurring using alternative model input values. For example, one key model input factor is SLR. Instead of selecting a specific SLR scenario value, SLAMM’s uncertainty module generates results that consider all five possible scenarios with mid-range SLR values having greater influence over model results than low or high scenarios. For more on SLAMM’s uncertainty module, see Warren Pinnacle Consulting’s Advancing Existing Assessment of Connecticut Marshes’ Response to SLR-Final Report (Section  2.13 Uncertainty Analysis Setup).

The viewer displays instances of both types of SLAMM results.

Uncertainty

The model used the best available data at the time of its development. It inherently considers uncertainty in key model input factors such as marsh surface sediment accretion rates, tide range variation, SLR, etc. To calculate and include uncertainty, the model is run several hundred times using alternative model input values randomly drawn from data input distributions tables. Each modeling result represents one possible future state for the studied area. All model results were then assembled into probability distributions of wetland coverage to reflect the effects of data uncertainty on prediction results.  Model results were then statistically analyzed to show the likelihood of occurrence of different outcomes. This viewer displays the ‘likelihood’ of areas of new marsh, high marsh and marsh loss expressed in terms of their occurring in the future.

SLAMM's High-Medium Sea Level Rise Rate for Road Flooding

To predict future road flooding frequencies, the viewer uses SLAMM’s High-Medium SLR rate of about 4.2 feet by 2100, depicted by the orange line in the following graph. This line includes a point approximating the Connecticut Institute Resilience of Climate Adaptation’s upper bound SLR planning threshold recommendation of approximately 20 inches of SLR by 2050 recently adopted by the State of Connecticut.

Scenarios based upon New York City Panel on Climate Change, 2013: Climate Risk Information 2013: Observations, Climate Change Projections, and Maps. Prepared for use by the City of New York Special Initiative on Rebuilding and Resiliency, New York, New York. Mayor’s Office of Long-Term Planning & Sustainability.

Sea Level Rise and Coastal Marshes

Sea Level Rise and Coastal Marshes

What are Coastal Marshes

Coastal marshes are the waving green fields of grasses along Long Island Sound’s shoreline and its saltwater tidal rivers, mostly within low-energy environments such as embayments. Here, protected from waves and strong currents, sediment accumulates providing conditions for salt marsh plants to grow. Located at the margin between land and sea, largely between the elevation of the average height of all tides and the highest tides of the year (see figure below), coastal marshes are among Connecticut's most susceptible ecosystems to the effects of SLR.

Connecticut's Coastal Marshes

Adapted from Warren Pinnacle Consulting

Coastal Marshes and Tides

Connecticut’s coastal marshes depend on a delicate balance between regular salt water inundation (high tide) and drying periods when tidal waters recede from the marsh surface (low tide). The heights reached by rising and falling tides and the marsh surface topography control the frequency and duration of saltwater inundation that define the marsh’s distinctive plant communities.

For centuries, Connecticut’s marshes have adapted to increases in SLR in two ways. The first is by increasing the elevation of the marsh surface. This happens through the accumulation mineral and organic material deposited on the marsh by tidal and riverine flooding as well as complex subsurface marsh expansion processes. The second is by migrating landward or upland onto undeveloped land not previously inundated by regular saltwater flooding.

As SLR Out-paces Marsh Adaption

In some areas, as the rate of SLR out-paces a marsh’s ability to increase its surface elevation, the lowest part of the marsh flooded daily (called ‘low marsh’), becomes too wet for marsh plants to survive resulting in their conversion to tidal mud flats or open water. Similarly, as the less frequently flooded parts of the marsh (called 'high marsh') is flooded more frequently, this part of the marsh converts to low marsh which supports a distinctly different plant community. In some cases, the highest elevation areas of marsh, called transitional marsh or marsh border, transitions to high marsh and the transitional marsh migrates to previously dry upland.

However, where barriers to marsh migration such as steep slopes or developed land prevent migration, the total area of marsh may decline or change from a high marsh to a low marsh dominated plant community.

Although SLAMM assumes that existing undeveloped upland areas regularly inundated by saltwater will support new marsh, the model recognizes that developed upland areas regularly flooded in the future are not suitable to accommodate marsh migration. Therefore the model does not project new marsh within newly flooded development landscapes.. One of SLAMM’s limitations is that it does not recognize that some of the undeveloped lands that will become regularly flooded that it predicts will become new marsh will be developed, filled or otherwise fortified over time prevent the right soil conditions or regular tidal water inundation needed to sustain marsh migration . Given these limitations, SLAMM likely overestimates the total amount of future new marsh.

Barriers to Marsh Migration

Scientists are just beginning to understand all of the implications of these kind of changes to the ecosystem services coastal marshes provide, such as nutrient sequestration, shoreline erosion control. and breeding habitat for marsh-dependent wildlife species, such as the Saltmarsh sparrow that breeds only in the high marsh.

Sea Level Rise and Coastal Roads

Sea Level Rise and Coastal Roads

Low elevation coastal area roads are those just a few feet above the elevation of normal high tides. They are already vulnerable to coastal flooding. SLAMM’s road flooding results indicate that in some areas the extent and frequency of coastal area road flooding will increase significantly at a sea level rise rate of about 16 inches by mid-century, slightly less than the 20 inch rise by mid-century recently adopted for planning purposes pursuant by the State of Connecticut pursuant to Public Act 18-8.

SLAMM Model and Roads

SLAMM identifies future SLR-influenced road flooding frequencies by individually assessing 5 meter road segments. This is advantageous because even small areas of road flooding can potentially limit access and egress to neighborhoods served only by a single road. It is important to note that SLAMM currently does not provide road flooding depths. In some cases, road segments projected to be ‘flooded’ will only be wet, but passable, and in other cases, road flooding depths could preclude safe vehicular access and egress. The report A Vulnerability-Assessment Tool that Will Quantify the Future Risk of Inundation for the Road Segments and Critical Infrastructure contains more detail on how SLAMM’s road flooding results can be used to assess future road flooding risks. Additional investigation is required before likely approximate road flooding depths can be determined.

In some areas, like the neighborhood shown below, extreme high tide flooding not influenced by storms will increasingly limit dry access to areas served by low elevation roads.

Existing Road Flooding Road Flooding by 2055 with 20" SLR

Although only small sections of the road shown below already flood as often as every 30 to 60 days, by mid century, all of the roadway is expected to flood at least once every 30 days. However, because SLAMM does not currently identify road flooding depths, the degree to which SLR will increasingly limit safe access and egress to neighborhoods served by a single low elevation flood prone road requires additional investigation

See Public Act No. 18-82, An Act Concerning Climate Change Planning and Resiliency.

Connecticut-Specific Road Flooding Methods

The Connecticut Department of Emergency Services and Public Protection (DESPP) statewide road network was the starting dataset for roads. Warren Pinnacle Consulting, Inc., ran SLAMM’s road flooding module using 5-meter road segments to assign the following flood values for each road segment at four future time steps:

no flooding
floods at east every 30 days
floods between every 30 and 60 days
floods between every 60 and 90 days
floods between every 90 days and 10 years
flood between every 10 and 100 years

The model provides five possible SLR scenarios: high, high-medium, medium, low-medium and low. These SLR scenarios, shown in the following table, are relative a base sea level elevation in 2002:

Connecticut SLAMM Sea Level Rise Scenarios

Time Interval

Low Projection (feet)

Low-Medium Projection (feet)

Medium Projection (feet)

High-Medium Projection (feet)

High Projection (feet)

2002

0

0

0

0

0

2025

0.167

0.335

0.499

0.666

0.833

2055

0.666

0.915

1.332

1.749

2.50

2085

1.083

1.499

2.418

3.251

4.833

2100

1.250

1.834

3.00

4.167

6.25

Source: Final Report - Advancing Existing Assessment of Connecticut Marshes’ Response to SLR, by Warren Pinnacle Consulting (2016)

Because it best fits with State of Connecticut's recently adopted policy of planning for 20 inches of SLR by mid-century (see Public Act No. 18-82, An Act Concerning Climate Change Planning and Resiliency) the high-medium scenario was chosen for analysis and display in the viewer. The years included in the viewer are 2010, 2025, 2040, 2055 and 2085. To simplify the GIS layer primarily for drawing speed, road segments that are not expected to experience coasting flood in the future were removed from each year's dataset and neighboring like-coded segments were combined. For example, 4 linear segments, each with the same flooding value in the same year are be combined to be a single, 20 meter length of road. To provide context to the flooded road segments, an All Roads layer is included in the viewer. It is the same road layer mentioned at the start before it was segmented.

State Roads

The original DESPP roads contain a lot of useful attribution but they do not indicate which roads are state roads and which are local. The Connecticut Department of Transportation (DOT) provided state roads but they were not a perfect geographic match with the DESPP road data. For ONLY roads in the coastal area, UConn CLEAR did some GIS analysis to populate the road segments with a state road attribute. We are confident in the results, although this data has not been approved by any state agency and is not considered authoritative.

The State Roads layer in the viewer is the DESPP roads that were identified as state roads after the analysis described above.

Soils Suitable for Marsh Migration

Soils Suitable for Marsh Migration

As coastal saltwater moves inland and stays longer on the ground, understanding how salts are retained in terrestrial soils is key to establishing a rulebase for marsh migration as part of the Sea Level Affecting Marshes Model (SLAMM). In soils, the retention of salts is influenced by soil texture, water table depth, cation exchange capacity and extractable cation exchange capacity.

Below are the physical and chemical characteristics of soils that influence salt retention due to coastal saltwater inundation for SLAMM as identified and deemed significant by the USDA’s Natural Resources Conservation Service. This soil survey interpretation can be used with SLAMM to help predict marsh migration and plan reasonable alternatives for the use and management of these soils. This interpretation does not include such factors as elevation, slope, landcover, or landscape position which may be incorporated into SLAMM.

Soil Rating Classes

High Suitability: These soils have the best combination of soil properties and characteristics for marsh migration.

Moderate Suitability: These soils have an average combination of soil properties and characteristics for marsh migration.

Low Suitability: These soils have an adverse combination of soil properties and characteristics for marsh migration.

Not Rated – Areas labeled Not Rated have characteristics that show extreme variability from one location to another which may require an on-site investigation to determine soil conditions present at the site, are subaqueous soils, or are miscellaneous soil map units such as beaches.

Use and Explanation of Soil Interpretations

Soil survey interpretations are predictions of soil behavior based on soil properties. Soil survey interpretations allow users of soil surveys to plan reasonable alternatives for the use and management of soils. They are used to plan both broad categories of land use such as cropland, woodland, or tidal marshes, as well as specific elements of those land uses (for example, marsh migration).

When soil interpretations are used in connection with delineated areas on soil maps, the information pertains to the soil for which the soil area is named. This is called the major soil component of the soil map unit. Other soils too small to map but observed within the delineated area are called minor soil components or inclusions. Soil interpretations do not eliminate the need for onsite study and testing of specific sites. Interpretations should be used as a guide to plan more detailed investigations and for avoiding undesirable sites for an intended use. The soil map and interpretations can be used along with other information, such as elevation and landcover, to select sites that have a certain potential for an intended use.

This soil interpretation does not include the minor soil components or inclusions. More detailed studies are required if small, specific sites are to be developed or used within a given soil delineation. No consideration was given in this interpretation to the size and shape of soil delineations, soil landscape position, elevation, slope, or to the pattern formed with other soils on the landscape. Although not considered in the interpretations, these items may influence the site.

Disclaimer: In obtaining this data from the Connecticut Department of Energy and Environmental Protection and the U.S. Department of Agriculture’s Natural Resources Conservation Service (NRCS), it is understood that you and your organization have the right to use it for any internal purpose. This data is not designed for use as a primary tool but may be used as a reference source. This data is not suitable for site-specific studies or litigation.

Web Soil Survey

Web Soil Survey provides soil data and information produced by the National Cooperative Soil Survey. It is operated by the U.S. Department of Agriculture’s (USDA) Natural Resources Conservation Service and provides access to the largest natural resource information system in the world. The site is updated and maintained online as the single authoritative source of soil survey information.

Interpretation by

Jacob Isleib, Resource Soil Scientist, USDA NRCS
Debbie Surabian, State Soil Scientist CT/RI, USDA NRCS

References

Anisfeld, Shimon C., Katherine R. Cooper, and Andrew C. Kemp. 2016. Upslope development of a tidal marsh as a function of upland land use. Global Change Biology, doi: 10.1111/gcb.13398

Broome, S.W., Seneca, E.D. and Woodhouse, W.W. Jr., 1998. Tidal salt marsh restoration. Aquatic Botany, 32: 1-22.

Hussein, A.H. 2009, Modeling of Sea-Level Rise and Deforestation in Submerging Coastal Ultisols of Chesapeake Bay.

Hussein and Rabenhorst. 2001. Tidal Inundation of transgressive coastal areas: Pedogenesis of salinization and alkalinization. Soil Sci. Soc. Am. J. 65:536–544.

Aerial Photographs in the Sea Level Rise Effects on Roads & Large Marshes Viewer

Aerial Photographs in the Sea Level Rise Effects on Roads & Large Marshes Viewer

Low altitude aerial photographs were captured using a custom built hexacopter modified to hold an Olympus camera. Aerial imagery was captured at Great Meadows (Stratford), East River (Guilford/Madison) and Plum Bank/Oyster River (Old Saybrook). The imagery provides information at a meso-scale uniquely capable of describing details of different parts of the marsh with enough context to understand the relationship between tidal water bodies, marsh plant communities and adjacent upland marsh migration areas that conventional high altitude aerial photography and ground photography cannot. The imagery also provides a tool to visualize the response of the marshscape and coastal area roads to more frequent flooding from sea level rise. In flight on-board sensors and GPS record photo position, heading, and altitude.

Three types of imagery are provided: snapshots, landscape views created by stitching photos and panoramas. All of the photographs are available as linked point features within the SLAMM viewer and individually here.