Tools for evaluating feasibility of restoration

Wetland Prioritization Study Main Page

 

Single-objective tools

 

Multi-objective tools

 


 

Ducks Unlimited Forested Wetland Restoration Suitability Model:1,2 This model evaluates site suitability for wetland restoration using Knowledge Engineer in ERDAS Imagine. The first step in Knowledge Engineer involves classifying each of five primary datasets (natural flood probability (NFP), soil moisture index (SMI), depressional sinks, stream buffers, and forest change) in either a binary fashion (high/low) or a non-binary fashion. The datasets classified in a non-binary fashion are NFP and SMI, with NFP separated into four classes and SMI into three. In the second step, a decision tree developed in Knowledge Engineer sorts particular pixels based on the combination of classifications for each individual dataset at that pixel's location. Another unique characteristic of Knowledge Engineer is that, based on the particular combination of classifications that result in a particular priority ranking for a parcel, confidence values are assigned for each pixel's categorization. The final output from Knowledge Engineer is that all pixels in the MAV study area are classified into one of the four priority classes (high, medium-high, medium-low, and low) and assigned a confidence value.

Factor used in analysis

Data source(s)

Suitability factors

Natural flood probability (NFP) maps

Period of record stream gauge data (US Army Corps of Engineers, Vicksburg District); Landsat TM imagery (after 1982)

Soil moisture index (SMI)

Landsat 7 ETM+ satellite imagery (acquired winter 1999)

Depressional sinks

USGS National Elevation Dataset

Forest change (1940's to 2001)

Landsat Multispectral Scanner (MSS) imagery; Landsat TM imagery;

Landsat ETM+ imagery

Stream buffers

USGS National Hydrography Dataset

Exclusion factors

2001 forest cover

Landsat Multispectral Scanner (MSS) imagery; Landsat TM imagery;

Landsat ETM+ imagery

Permanent water bodies

Generated internally using summer imagery and low river levels

Roads

ESRI dataset

Urban areas

ESRI dataset

Wildlife management areas

Data acquired from individual state agencies

National Wildlife Refuges

USFWS National Cadestral Database

National forests

Forest Service GeoData Clearinghouse

Farm and home administration lands

Data provided by either NRCS or FWS

Wetland Reserve Program project sites

NRCS geospatial database

DU conservation easements

DU conservation easement geospatial database

DU private landowner projects

DU private lands geospatial database

Private landowner conservation projects of other conservation partners

Vector data acquired from conservation partners

Acronym definitions

 

Back to top

 

The Nature Conservancy and Environmental Law Institute Duck-Pensaukee Watershed Approach Pilot Project (DPWAP )Potentially Restorable Wetlands (PRW) Tool:3 A planning team identified reestablishment opportunities as sites that formerly supported wetland hydrology and vegetation and which currently had appropriate land cover for restoration. These sites were termed Potentially Restorable Wetlands (PRWs). PRWs were identified based on factors such as presence of hydric soils and appropriate current land cover and were integrated into the TNC-ELI DPWAP Wildlife Tool , Flood Abatement Tool, Surface Water Supply Tool, Water Quality Protection Tool, Carbon Storage Tool, Shoreline Protection Tool, and Fish Habitat Tool.

Factor used in analysis

Data source

Hydric soils

NRCS SSURGO soils data

Is not within an existing wetland

WDNR WWI wetland map data, converted to NWI format by USFWS

Land use reflecting opportunity (e.g., agriculture) or lack of opportunity (e.g., urban, or existing natural cover) for successful wetland restoration

C-CAP (2001) land cover data

Acronym definitions

 

Back to top

 

Michigan Tech Research Institute (MTRI) Wetland Mitigation Site Suitability Tool (WMSST):4 MTRI combined eight statewide raster datasets in ArcGIS to score the feasibility of wetland restoration or creation for each pixel. Two of these layers, Topographic Wetness Index (TWI) and Soil Moisture Index (SMI), required additional processing by MTRI before they were integrated into the WMSST. The TWI was calculated based on slope and elevation data and served as a measure of the wetness at each site. The SMI was calculated based on the relationship between surface temperature and vegetation and served as a measure of soil moisture levels.

All eight data layers were first reclassified, based on feedback from Michigan Department of Transportation environmental personnel and published literature, to convert the original data values of each layer to values of 0, 25, 50, 75, and 100. For example, tables showing how original values were reclassified for the TWI and USDA SSURGO soils data are provided below:

MTRI Topographic Wetness Index‹00
0.5-2525
25.5-5050
50.5-7575
75.5-100100

 

USDA SSURGO Drainage ClassExcessively Drained0
Well Drained25
Poorly Drained50
Very Poorly Drained100


Reclassification ensured that all data layers, including those containing both nominal and interval data, used a standard scale. Once reclassified, each layer was clipped in order to divide each statewide layer into smaller areas that matched watershed and ecoregional boundaries to minimize the run-time of the final WMSST.

Factor used in analysis

Data source

Soil drainage class

USDA SSURGO drainage class data

Soil flood frequency

USDA SSURGO flood frequency data

Presence of hydric soils

USDA SSURGO hydric classification data

Soil ponding frequency

USDA SSURGO ponding frequency data

MTRI Topographic Wetness Index (TWI)

Slope

USGS Percent Slope Classification

Elevation

USGS DEM

MTRI Soil Moisture Index

Relationships between surface temperature and vegetation levels

Landsat satellite imagery

Percent slope

USGS Percent Slope Classification

Land cover

MNFI presettlement land cover data

Acronym definitions

 

Back to top

 

The New Hampshire Department of Environmental ServicesWetland Restoration Assessment Model(NHDES WRAM) Site Identification Model:5 Starting with a composite layer of existing wetlands, the Site ID Model identified a set of 951 candidate sites for input into the Site Prioritization Model using a simple screening process involving the following steps:

  1. Some portion of the wetland must be assigned one or more of the following Cowardin classifications in the NWI: "partially drained/ditched," "diked/impounded," or "excavated."
  2. Any portion of the wetland must intersect any of the following land cover classifications from New Hampshire Land Cover Classification data: "barren lands," "orchard," "other agricultural," "hay/pasture or row crop," "disturbed land," or "other cleared lands."
  3. Candidate sites less than five acres in size were excluded because the literature suggests that restoration is most likely to be successful when working in wetlands at least five acres in size. The Technical Advisory Group, composed of state agency, regional planning commission, and nonprofit personnel, refined the candidate site list from here, removing and adding sites at their discretion.

Factor used in analysis

Data source

Must have a Cowardin classification of "partially drained/ditched," "diked/impounded," or "excavated"

NWI

Must have a land cover classification of "barren lands," "orchard," "other agricultural," "hay/pasture or row crop," "disturbed land," or "other cleared lands"

NHLCC (2001)

Must be larger than five acres in size

NWI

Acronym definitions

 

Back to top


Kaufmann-Axelrod and Steinberg (2010) Scranton (2004) Tool for Identifying Restoration Consideration Areas:6,7 Scranton (2004) identified Restoration Consideration Areas (RCAs) as part of a process to delineate and classify tidal wetlands throughout Oregon by drawing upon a variety of datasets such as NWI data, head-of-tide locations, and other data. RCAs are "upland or non-tidal areas that might deserve closer scrutiny as possible candidates for restoration of tidal circulation, pending landowner involvement... These are generally lands that are diked or may have been partially filled or ditched for agricultural or commercial purposes." In total, Scranton identified 2264 RCAs throughout the state of Oregon.

Factor used in analysis

Data source(s)

Aerial photography

Various imagery sources and years

Wetland land cover

National Wetland Inventory

Local Wetland Inventories

Oregon Dept. of State Lands

Hydric Soils data

NRCS SSURGO

Estuarine land cover

Oregon Dept. of Land Conservation and Development Estuary Plan Book

Digital elevation models

USGS

Head-of-tide locations

Oregon Dept. of State Lands

Additional Resources

Fieldwork, wetland specialists, publications

Acronym definitions

 

Back to top

 
Strager et al. (2011) Wetland banking site selection model:8 Strager et al. (2011) identified potential wetland mitigation sites within HUC-10 watersheds by combining wetland soil type data from digital soil maps with a map of wetlands from the National Wetlands Inventory (NWI). The resulting wetland areas were then examined using aerial photos to remove areas from consideration in which development or conversion to impervious surface had occurred. Potential wetland areas identified in the aerial photos to contain natural land cover or permeable surface were selected for further evaluation in the subsequent rapid assessment analysis.

Factor used in analysis

Data source

Historical wetlands on or adjacent to site

NWI wetland coverage

Hydric or somewhat poorly drained soils

SSURGO soil coverage

Non-permeable surfaces or natural land only

NLCD land use coverage

Evidence of development or conversion to impermeable surface

2006 color infra-red 1:12,000 Digital Ortho Quarter Quads (DOQQ) aerial photos with one meter resolution (flown during leaf-off)

Potential wetlands with natural land cover or permeable surface

Acronym definitions

 

Back to top


The Virginia Institute of Marine Science(VIMS)Wetland Mitigation Targeting Tool (WMTT):9 The WMTT applies a hierarchical approach in which GIS layers representing different criteria for wetland restoration/creation are applied successively, in order of importance. As an area on the landscape satisfies more of these spatial criteria its suitability rank for the development and persistence of wetlands increases. This process begins with a raster land cover dataset, from which forested and agricultural land use types - considered to have a high probability of successful conversion to wetlands - are identified for further consideration as potential sites for wetland restoration or creation.

Within these areas, those containing hydric soils larger than 0.25 acres are ranked as "potential" restoration sites. Of these areas, those interecting streams (an indicator of hydrologic connectivity) are ranked as "moderate." Moderate areas adjacent to wetlands (an indicator of the likelihood that restoration will be successful) are then ranked "good" if they are forested and "high" if they are agricultural. The tool's higher preference for agricultural over forested land reflects the higher ecological values of maintaining forest buffers as well as the fact that agricultural lands meeting hydrology and hydric soil criteria are likely to be prior converted wetlands. Sites that were ranked "good" and are adjacent to conservation areas are elevated to "high," while those that were ranked as "high" and are adjacent to conservation areas are elevated to "excellent."

There was no formal collaboration with stakeholders during development of the tool - Center for Coastal Resources Management (CCRM) staff members were centrally responsible for developing the criteria and scoring system upon which the tool was based. However, because VIMS staff members are actively involved in wetland restoration and creation projects with area non-profits (e.g., the Elizabeth River Project) and government agencies (e.g., Department of Defense wetland restoration and creation), input from area stakeholders may have informally influenced the tool's development.

Factor used in analysis

Data source(s)

Forested and agricultural land cover

NLCD

Hydric soils

Virginia Tech GIS (VIRGIS); SSURGO

Hydrologic connectivity

DLG hydrography

Wetland polygons

NWI

Conservation Lands

Virginia Department of Conservation/Natural Heritage Division

Acronym definitions

 

Back to top


The Wisconsin Department of Natural Resources (WDNR) Potential Opportunity Tool:10 This tool evaluates the opportunity for wetland functional improvement through restoration across HUC-12 watersheds as indicated by the relative amount of Potentially Restorable Wetlands (PRWs) and the original percentage of wetlands in the watershed. WDNR assesses potential opportunities using the equation below.

Factor used in analysis

Data source(s)

Hydric soils

SSURGO soils data

Not within an existing wetland

DWWI wetland mapping data; other sources of wetland mapping data (e.g., county-level sources, if available)

Land use reflecting opportunity (e.g., agriculture) or lack of opportunity (e.g., urban) for successful wetland restoration

Milwaukee study used SEWRPC augmented by WISCLAND land cover and US Census TIGER/Line road data; Rock River study used NASS cropland GIS data

Acronym definitions

 

Back to top

 

The Wisconsin Department of Natural Resources (WDNR) Potentially Restorable Wetlands (PRWs) Tool:11 This tool identifies restoration opportunities across the landscape and is composed of the combined boundaries of hydric soils, existing wetlands, and land use datasets. WDNR classifies areas as potentially restorable if they: 1) contain hydric soils, 2) do not overlap the location of existing wetlands, and 3) are not developed.

Factor used in analysis

Data source(s)

Hydric soils

SSURGO soils data

Not within an existing wetland

DWWI wetland mapping data; other sources of wetland mapping data (e.g., county-level sources, if available)

Land use reflecting opportunity (e.g., agriculture) or lack of opportunity (e.g., urban) for successful wetland restoration

Milwaukee study used SEWRPC augmented by WISCLAND land cover and US Census TIGER/Line road data; Rock River study used NASS cropland GIS data

Acronym definitions

 

Back to top

 

Wetland Prioritization Study Main Page

 


 

1Shankle, S, Brown, D, Holden, J. Site suitability modeling for the restoration of forested wetland in the Mississippi Alluvial Valley.
2Feedback received on 5/10/2012 from Dale James, Manager of Conservation Planning, Ducks Unlimited.
3 Miller, N., T. Bernthal, J. Wagner, M. Grimm, G. Casper, and J. Kline. 2012. The Duck-Pensaukee Watershed Approach: Mapping Wetland Services, Meeting Watershed Needs. The Nature Conservancy and Environmental Law Institute, Madison, Wisconsin.
4 Brooks C, Powell R, Shuchman R, Leonard G. Developing and applying a geospatial decision support tool for efficient identification of wetlands mitigation sites.
5 Vanasse Hangen Brustlin, Inc. 2009. Merrimack River Watershed Restoration Strategy. Prepared for New Hampshire Department of Environmental Services.
6 Interview on 12/19/2011 with Jennifer Axelrod, FLO Data and GIS.
7 The application of geographic information systems for delineation and classification of tidal wetlands for restource management of Oregon's coastal watersheds. Masters Thesis, Marine Resource Management Program, Oregon State University.
8 Strager MP, Anderson JT, Osbourne JD, and Fortney R. 2011. A three-tiered framework to selection, prioritize, and evaluate potential wetland and stream mitigation banking sites. Wetlands Ecology and Management 19:1-18.
9 Berman MR, Rudnicky T, Berquist H, and Hershner C. 2002. Protocols for implementation of a GIS-based model for the selection of potential wetlands restoration sites in southeastern Virginia. Center for Coastal Resources Management, Virginia Institute of Marine Science, College of William and Mary. Gloucester Point, Virginia.
10 Hatch B, Bernthal T. 2008. Mapping Potentially Restorable Wetland in the Rock River Basin.
11 Kline J, Bernthal T, Burzynski M, Barrett K. 2006. Milwaukee River Basin Wetland Assessment Project: Developing Decision Support Tools for Effective Planning.