Freshwater & Oceans
Input data QA/QC
- Arkansas Multi-Agency Wetland Planning Team
- Ducks Unlimited
- Maryland Watershed Resources Registry
- Ducks Unlimited
- Maryland Watershed Resources Registry
- Michigan Tech Research Institute
- NHDES Wetland Restoration Assessment Model
- UMass Amherst Conservation Assessment and Prioritization System
- USACE Sunrise River Watershed-Based Mitigation Pilot
- Wisconsin Department of Natural Resources
Arkansas Multi-Agency Wetland Planning Team:1 Data layers assembled for the analysis are groundtruthed based on local knowledge, windshield surveys, and field visits. Layers evaluated for QA/QC generally include NRCS soils data, GAP analysis data, NHD data and other available hydrologic data. Ground-truthing of inputs is generally initiated when obvious discrepancies exist in the initial maps.
Ducks Unlimited (DU):2 DU conducted ground verification to verify the accuracy of input variables for two counties in the Mississippi Alluvial Valley (MAV). In doing so, they confirmed that areas classified as soil moisture index (SMI) classes 1 and 2 contained ground indicators of high surface soil moisture.
Maryland Watershed Resources Registry (WRR):3 The WRR is in the process of developing a method for the field validation of the input data sources. These on-the-ground assessments will likely be rapid and will seek to confirm whether factors are present as described by the input maps.
Ducks Unlimited (DU):2 DU completed a map agreement analysis in which they confirmed that Soil Survey Geographic (SSURGO) hydric soils data for five counties corresponded strongly with soil moisture index (SMI) classes 1 and 2.
Maryland Watershed Resources Registry (WRR):3 WRR completes a desktop review of the model outputs to ensure that they are being calculated correctly within the model.
Michigan Tech Research Institute:4 Prior to running the analysis, Michigan Department of Transportation (MDOT's) recommended workflow for the Wetland Mitigation Site Suitability Tool (WMSST) directs users to examine input and visualization data. By comparing input data with other information provided in the visualization data layers (e.g., NWI data, aerial photography, etc), users are able to evaluate the data inputs for QA/QC.
New Hampshire Department of Environmental Services (NHDES) Wetland Restoration Assessment Model:5 New Hampshire Department of Environmental Services (NHDES) applied comprehensive GIS data quality standards to add datasets used in the analysis. These included using only GIS data of known origin, obtaining the most updated version of each dataset from its original source, and using only datasets properly documented to Federal Geographic Data Committee (FGDC) standards.
UMass Amherst Conservation Assessment and Prioritization System (CAPS):6 Though the researchers put "considerable effort" into integrating data in ways that maximize accuracy, because CAPS input data come from a variety of sources, of variable quality, they expect that some amount of error will inevitably be present. They are unable to estimate the accuracy of the final dataset or effects that errors in the base map may have had on final results. They believe the effects of errors are negligible but plan to evaluate them in more detail in the future.
USACE Sunrise River Watershed-Based Mitigation Pilot:7 While validating the Spatial Decision Support System (SDSS) model, the Corps realized that its roads dataset was inaccurate. In response, it updated some roads data in some areas and removed some mapped roads that were no longer present. In addition, the Corps discovered that some sites identified as containing hydric soils were actually forested. In order to avoid inadvertently advocating conversion of forested areas to wetlands, the Corps excluded these areas from consideration as priority areas.
Wisconsin Department of Natural Resource (WDNR):8 In developing the potentially restorable wetlands (PRW) layer , WDNR worked to reconcile conflicting attribute classifications among the wetland, land use, and soils layers. For example, a feature may have been classified by Digital Wisconsin Wetland Inventory (DWWI) as upland but classified by Southeastern Wisconsin Regional Planning Commission (SEWRPC) as wetland. In such cases, WDNR determined a final wetland classification by "studying the input layers, comparing dates of the sources, and studying randomly selected features on aerial photographs." WDNR also cleaned up slivers and gaps in the final PRW dataset created as a result of overlaying and intersecting base layer feature polygons and eliminated all PRWs less than 0.5 acres in size.
1Email correspondence received on 10/13/2011 from Jennifer Sheehan, Arkansas Multi-Agency Wetland Planning Team Coordination Office.
2 Shankle, S, Brown, D, Holden, J. Site suitability modeling for the restoration of forested wetland in the Mississippi Alluvial Valley.
3 Interviews on 8/3/2011 with Ellen Bryson, USACE Baltimore District, and on 8/11/2011 with Ralph Spagnolo, USEPA Region III.
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 Interview on 8/19/2011 with Collis Adams and Lori Sommer, NHDES Wetlands Bureau.
6 Interview on 7/29/2011 with Scott Jackson, Program Director, UMass Extension's Natural Resources and Environmental Conservation Program, Department of Environmental Conservation, University of Massachusetts, Amherst.
7 Interview in 12/2011 with Tim Smith, Enforcement and Compliance Coordinator, U.S. Army Corps of Engineers, St. Paul District.
8 Kline J, Bernthal T, Burzynski M, Barrett K. 2006. Milwaukee River Basin Wetland Assessment Project: Developing Decision Support Tools for Effective Planning.