CICEET Progress Report for the period 9/01/08 Through 2/28/09

Project Title: Vegetation, Impervious Surfaces, Soils, and Topographic Analysis Tools: Geospatial Technology to Promote Coastal Water Quality
Principal Investigator(s): Thomas R. Allen
Project Start Date: 1 September 2007
Report Compiled By: Thomas R. Allen

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Our non-technical objective was to incorporate feedback from practitioners, which we obtained at a presentation to the NC ArcUser Group and meetings in Currituck and Dare Counties. In addition, we exchanged input with other CICEET land use projects and will share results with the broader research community next quarter. To bolster end-user participation, Currituck County has expressed strong interest and participation in the project. The following quote in correspondence indicates the positive progress in our overall objective to provide a tool and transformational information that planners can use to guide policy and action:

January 27, 2009
Tom,
The preliminary information you are putting together is great. The idea of delineating subwatersheds at a subdivision scale would be invaluable for the Planning Department. Not only could we use the information for the Currituck Goes Green effort, but it would assist us with our land use planning efforts (subdivision approvals, small area planning, etc...). In a perfect world, we would even develop regulations or policies that are unique to the attributes of a particular subwatershed(s). From a water quality perspective, for example, this type of information could help us to determine where a bioretention area makes sense versus a reconstructed wetland.
Please let us know how we can help.
Ben E. Woody, AICP
Planning Director
Currituck County

Tasks to meet objectives
1. DEM data assimilation and watershed delineation
The limited availability of high-resolution watershed delineations is a serious limitation to fine-scale stormwater runoff analyses. Even with North Carolina's relative richness of extensive high-resolution LiDAR-derived DEM, few coastal counties have watersheds mapped to resolution better than 1:24,000 or the USGS' 14-digit hydrologic unit codes (HUC). While the LiDAR DEMs are instead produced primarily for floodplain delineation and hazard mitigation, we are exploiting this data source for water quality and land use planning. Our project began with two jurisdictions at different capacities to collaborate, Camden County which lacked any subwatershed delineation below HUC-14 scale, and the Town of Southern Shores (Dare County) which had available fine-scale subwatersheds delineated by engineering consultants. In order to address the applicability of our model to Camden County (as well as transferability of our model among other jurisdictions), we aggressively explored the opportunity to automate watershed delineation, as an optional component of the VISSTA tool.

In addition, in response to a solicitation to work with Currituck County and a volunteer environmental steering committee, we have selected a county with a composite situation of our two initial study areas (and which is geographically situated between the other two.) Currituck includes both barrier island and coastal mainland topography, beachfront and rural communities, and also lacks the availability of fine-scale watershed layers. However, the county has a full-time GIS analyst and planning staff and administration receptive to applied research participation.

2. Soil database
We expanded our USDA Natural Resources Conservation Service (NRCS) Soil Survey GIS database (SSURGO) for agricultural and other non-point source pollution to include Currituck County, including the calculation of infiltration capacity, hydric groups, and K-factor values.

3. Soil moisture model
The summary index we previously developed, adapted from Louis Iverson's (1998) Integrated Moisture Index, is now complete. The modelbuilder application runs on our server. This geoprocessing tool can be enabled to run via a server, remotely. We demonstrated the output to visualize soil moisture for a calculation county-wide. We also have established a protocol for precipitation data analysis using PRISM products.

4. NSPECT and ArcCN Runoff Evaluation
As a benchmark for our model, we have developed the above datasets to also run with NOAA Coastal Services Center Nonpoint Source Pollutant and Erosion Comparison Tool (NSPECT). NSPECT was installed and all necessary soil, DEM, land use, and climatological data assimilated (Zhan and Huang 2004.)

We also initiated a comparison of our approach to ArcCN Runoff, a curve-number based runoff model, also developed as an extension for ArcGIS.

Progress on Tasks
1. DEM and watershed delineations
We completed acquisition of LiDAR DEM data, delineation and initial evaluation of watersheds the 20-foot resolution DEM. We evaluated ArcHydro and GRASS GIS r.watershed algorithm, variable areal threshold watersheds, and contrast of pit-filling and explicit sink effects on resulting delineations. We also have attempted using PCSA, a software package developed for hydrologic modeling in depressional topography (e.g., glacial landscape) but encountered difficulty with Fortran source code on our systems. The basic algorithm may yet be rewritten in Python and included in our ArcGIS toolbox, but for now, we are using the GRASS r.watershed tool.

Further review with county participants is planned, to ensure the watershed scale is adequate and accurate for their prospective needs at parcel to subdivision scale of analysis. The project may re-run the delineations subsequent to this review and comparison in the field.

This task completes (for a third jurisdiction), the goal of developing a a seamless, high-resolution watershed delineation for subsequent modeling.

2. Soils data
The soils data compilation and attribute calculations with SSURGO data for Currituck County are complete. The NSPECT soil preprocessing routines were primarily utilized to derive K-factor erodibility values. We also completed the modelbuilder script to sort and reclassify hydric groups with a master list of soil types for coastal North Carolina.

3. Soil moisture potential
The Integrated Moisture Index (IMI) previously adapted from Iverson has been written in a modelbuilder geoprocessing script. Runs have been completed and tested in all three counties. Utility of this individual layer has not been determined yet by our end-users.

4. NSPECT and ArcCN Runoff model evaluation
ArcCN Runoff model was obtained and installed from the ESRI ArcScripts online repository. The curve-number and hydric group method is analogous to the VISSTA approach, yet the coefficients used are not the same, with hydric groups treated slightly differently.

NOAA CSC's NSPECT model has also been installed and all data for Currituck prepared for analysis. The preprocessing routine used by NSPECT for watershed delineation appears to falter in several locations. It is not feasible to alter the algorithm embedded in this tool for our project. Most likely, the underlying ArcHydro routines are filling "pits" in the coastal LiDAR DEMs that are in actuality real sinks and depressions in the coastal beach ridge topography of the area. Hence, the preference is very strong for an alternative source of watershed delineation (or else substantial manual editing.) NSPECT and ArcCN Runoff will next be run using our developed data from GRASS GIS' r.watershed algorithm, which allows flow direction routing with depressional topography.

5. Non-technical Stakeholder engagement
Setbacks have been encountered but also overcome in this phase of the project with stakeholders. We were dismayed at the job move of a collaborator in Camden but later enthused to have an adjoining county enthusiastically able to accommodate our project. Currituck County has been transitioned to the focus of this project. The county planners, GIS staff, and a handful of other volunteers have joined to establish an environmental steering committee for the county. This project joined the "mapping committee" and provides input to the group's sustainability and "green" building and development strategy. Data products and tools from our CICEET work are to be featured in this group's activities, beginning with the mapping of subwatersheds necessary for subsequent corresponding VISSTA tools. We have participated in two stakeholder meetings since December and established regular correspondence. The county and volunteers are now reviewing the first data product. A demonstration area has also been selected, the Town of Moyock.

In addition, our coverage in Currituck County includes the Currituck Banks NCNERR. This project and a graduate student class are applying the same data analysis for the county to the requisite areas of the CB NERR. With the assistance of the NCNERR GIS coordinator, we also can provide the NERR with the same subwatershed and nonpoint source and runoff data we will provide to the county stakeholders.

Difficulties
The greatest impediment to progress has been data and software assimilation and the shift required in focus area to adjoining Currituck County. Acquisition and processing of soils databases and DEMs are time-intensive tasks. In addition, we attempted to run the PCSA watershed program but have not completed it successfully, requiring our resorting to GRASS GIS (and accompanying data conversions.) The move of a collaborating GIS analyst in Camden was not caught until several weeks, and a replacement had not been immediately available. We have taken the opportunity to work more closely with Currituck but fully intend to reconnect with Camden and the Town of Southern Shores (now as validation and demonstration areas at the end of the project.) Although time demanding, these difficulties have been surmounted. We will further work on the watershed delineation issue, and attempt to either integrate the PCSA algorithm in ArcGIS or to very extensively apply the high quality GRASS r.watershed algorithm throughout several counties.

Data Generated to date
A sample of products are provided in the graphics enclosed. These data include subwatersheds, NRCS soils hydric groups, K-factor erodibility, integrated moisture index (IMI) and input variables.

See Figure 2. NRCS soils hydric groups for Currituck County showing hydric groups used in ArcCN Runoff model evaluation.

See Figure 3. Integrated moisture index (IMI) model output for Currituck County, incorporating soil infiltration capacity, hydric group, and DEM.

See Figure 4. Currituck County subwatershed delineations derived with GRASS GIS r.watershed algorithm.

Project Objectives for Next Reporting Period

Technical objectives
a. Integrate the modelbuilder tools into an ArcToolbox with documentation, tutorial, and sample data for a specific watershed (e.g., NCNERR site/s.)

b. Develop output query functionality, such as parcel-scale or variable watersheds and areal thresholds, and match these to Currituck BMP demonstration projects.

c. Evaluate preset or default precipitation regimes (e.g., design storm or recurrence-interval based values) versus composite, annual "potential" values.

d. Port the geoprocessing models to ArcGIS server-based implementation. Test and evaluate user interface and performance.

Non-technical objectives
a. Test the model on various use cases in conjunction with end-users in Currituck County

b. Document this prototype and start-to-finish example for at least one use case scenario within Currituck

c. Identify an additional end-user to evaluate the result and plan the dissemination via workshop, large audience presentation (e.g., NC ArcUser Group, ESRI User Conference, Coastal GeoTools, American Planning Association or other large venue) and provide online access.

d. Mockup and evaluate end-user feedback and preferences for GIS server-based or desktop application environment.

Expenditures
Approximate expenditures during the semi-annual reporting period include:
Principal investigator $ 4,500.
Graduate student assistants $ 7,000.
Travel (NCAUG conference and local coordination) $ 1,200.
Data acquisition $ 500.

References
Iverson, L.R.; Dale, M.E.; Scott, C.T.; Prasad, A. 1997. A GIS-derived integrated moisture index to predict forest composition and productivity in Ohio forests. Landscape Ecology. 12:331-348.

Zhan, X. and M-L. Huang. 2004. ArcCN-Runoff: an ArcGIS tool for generating curve-number and runoff maps. Environmental Modelling and Software. 19(10):875-879.