Progress Report

CICEET Progress Report for the period 9/01/08 Through 2/15/09

Project Title: Integrating Geospatial and Web-Based Technologies to Improve Land Use Planning in Coastal New Hampshire
Principal Investigator(s): Fay Rubin, Complex Systems Research Center, UNH
Project Start Date: September 1, 2007
Report Compiled By: Fay Rubin

Figures


Figure 1

Figure 1


Figure 2

Figure 2


Figure 3

Figure 3


Figure 4

Figure 4


Figure 5

Figure 5


Project Objectives for This Reporting Period
Objectives
The overall objective of this project is to disseminate a toolbox of land use planning geospatial resources in order to assist coastal communities in making decisions that protect environmental character and water quality. We are focusing on identifying, developing and/or assembling resources that yield information appropriate for incorporation in future master plans and conservation guidelines. Our initial plan was to disseminate these resources using a dual approach of relying on both industry standard software (ArcGIS) and open source tools. However, based on discussions with targeted end-users as well as internal capacity-building, we have opted to focus on open source tools as our primary dissemination mechanism.

Our specific project objectives for this period were as follows:
1. Develop impervious surface data for Rockingham County
2. Refine and run the SUBNET model (previously two objectives, but aggregated here for reporting purposes)
3. Select existing resource identification techniques to convert to ModelBuilder format
4. Release first iteration of the web site

Progress on Tasks
1. Develop impervious surface data for Rockingham County
Impervious surface (IS) coverage is an important water quality indicator that is presently unavailable to most coastal communities. We are developing this data set from high resolution aerial imagery, and will utilize it with contemporaneous land use mapping to generate county level IS coefficients. These coefficients will then be applied to the future land use projections to generate estimates of future "degree of imperviousness". The IS data will also be available to communities to utilize in conjunction with other planning and environmental data sets.

To complete the IS data set, staff from the Complex Systems Research Center (CSRC) planned to:

  • Finalize the technical approach to the data development task
  • Collect any required field data
  • Process the imagery
  • Conduct post-processing accuracy assessment

Following Definiens eCognition software training and development of the technical approach (i.e. ruleset development), we are now in the midst of the IS mapping. We find that the most manageable approach is to process the data in a town-by-town manner. As such, the 1-foot aerial photography is combined into a single 4-band data set for each town (3-band CIR and blue band from RGB). This image is then processed (in 16,000x16,000 pixel subsets) using the eCognition software to extract impervious surfaces. The automated processing is followed by data clean-up to correct any misclassifications.

To date, we have completed initial data development for over half of Rockingham County, and final data refinement for six (Atkinson, Auburn, Kingston, Newington, Portsmouth, and Seabrook) of the 37 towns. Figure 1 shows the completed town-wide IS data mapped for Seabrook. Figure 2 displays a close up view of the completed data mapped in downtown Portsmouth.

We have found numerous areas in the aerial photography where specific types of impervious surfaces, such as buildings, are obscured by trees and therefore produce confounding signatures. Without supplemental data (LIDAR for example), these situations present difficulties only because they require greater than expected cleanup efforts. We feel, however, that this data layer is important enough to warrant the additional effort.

2. Refine and run the SUBNET model
Staff from Applied GeoSolutions (AGS) continued to refine the SUBNET model output. With a set of spatial variables in place and a preliminary run of the model complete, we opted to pare down the number of variables in the model to determine the key factors driving development. Using a stepwise removal process we determined that the model was optimized (r of testing = 0.2, r of training = 0.4) using two variables: wetlands (National Wetlands Inventory) and distance to 1998 built areas (Rockingham County Land Use).

SUBNET model output now shows the probability that an area will convert before the year 2012, or within 7 years of the base modeling date. Figure 3 below illustrates this for Rockingham County, NH.

Development of the SUBNET model is now complete. Model results are currently available through a web mapping interface (see Figure 4 and Figure 5) at the website (http://subnet.homedns.org; user name is ‘guest' and password is ‘guest').

3. Select existing resource identification techniques to convert to ModelBuilder format
For GIS-savy planners, ModelBuilder has emerged as a key feature of ESRI's ArcGIS and ArcGIS Server platform. ModelBuilder provides an environment within ArcGIS where analytical procedures can be detailed, stored, re-run with new data, and shared with other ArcGIS users. While the majority of the coastal towns and planners in NH do not currently use ArcGIS to a great extent, key players in our target audience (Regional Planning Commissions, National Estuarine Research Reserves, several coastal towns) do possess sufficient skills and technology to greatly benefit from the availability of resource identification techniques in a ModelBuilder format. A variety of potential techniques have been discussed and evaluated over the course of the grant, but little direct progress has been made recently in this area due to a sabbatical leave taken by the lead on this task (Shane Bradt). When Shane returns to work in late April, his remaining project time will be dedicated to the coding of techniques in ModelBuilder. Once the techniques have been converted, planners will be trained on their use through the Getting to Know ModelBuilder workshop currently offered by the Geospatial Technology Training Center (next scheduled for September 10th at UNH in Durham, NH).

4. Release first iteration of the web site
An outgrowth of the SUBNET web-mapping application and other planning related online-tools proposed as part of this project is the concept of a web-GIS tool oriented towards local planning boards and conservation commissions. The tool is being developed jointly by GRANIT, UNHCE, and AGS.

With input and direction from our local partners, the project team developed a concept for the tool functionality. Additional input was then collected at a series of meetings with stakeholder communities, coordinated by the Natural Resources Outreach Coalition (NROC), with participation from UNHCE Geospatial Technologies staff, and AGS staff. The meetings had as their primary focus the land use planning needs of coastal communities and the types of data that will allow the towns to most effectively address these needs. At these meetings, input was solicited from diverse group of stakeholders in coastal communities, including planning boards, conservation commissions, selectboards, and city/town planning departments. Through this process, the tools developed will address the land use planning needs of the target constituencies as well as their counterparts throughout the coastal community.

Meetings were held with three towns, and attended by staff from NROC, UNHCE Geospatial Technologies, and AGS.
Raymond Monday December 1, 2008
Participants included representatives from the selectboard, conservation commission, planning board, and town planning department.

Milton Monday December 8, 2008
Participants included representatives from the conservation commission, planning board and selectboard.

Dover Thursday December 18, 2008
Participants included representatives from the Dover Conservation Commission, Open Lands Committee, Planning Board, and City Planning Department

In addition to the scheduled meetings, staff from AGS had independent discussions with numerous conservation commission members, planning board members, and professional planners to discuss project ideas and get confirmation of aspects of the planning tool.

The result of collaborative planning sessions and discussions with potential end-users is the design for a web-tool to analyze spatial data for planning board applications. The tool will be geared towards user with limited or no GIS experience. It will allow users to rapidly access and combine the data layers necessary to make decisions regarding a particular parcel/s. The tool will provide both a regulatory and a non-regulatory context for each tract being considered, so that the natural resource values of the surrounding landscape are taken into consideration during the decision-making process.

Given that most of New Hampshire's statewide spatial datasets are created at 1:24,000 scale and not meant for parcel-specific applications, the goal of the tool is to flag potential concerns associated with a subject tract (such as steep slopes, land regulated under Comprehensive Shoreland Protection Act, etc.), rather than provide precise statistical information. This will allow board members the opportunity to ask questions of applicants and ensure that they get the engineering-level data that they need.

Users will benefit by (1) having access to GIS data that they would not otherwise have access to and (2) having a geographically-linked checklist that will prompt them to ask the relevant questions to applicants. AGS will be responsible for the development and hosting of this tool, which will include the following components:

Tract boundary input
A user will have several options to identify the tract location (or approximate location), including: a) drawing a polygon on-screen of the tract boundary; b) uploading a shapefile of the tract; c) drawing a point on-screen to estimate the tract centroid; or d) entering the tract address.

Data analysis
The discussions with stakeholder groups helped to identify data layers that are most helpful for communities in land use planning decision-making. Generally, these data fall into two categories (see listing below). Group I data represent natural resources regulated under NH law and are relevant to application approvals. Group II data are important for long-term planning (like the master plan process) and conservation interests and provide the landscape context for the tract in question.

Group 1 Data ­ RegulatoryGroup 2 Data ­ Non-regulatory
Wetlands (NWI and hydric soils)Watershed setting (from stream catchment to larger basin)
Prime wetlands (where available)Soils derivatives (state/local agricultural soils, important forest soil groups, slope, depth)
Steep slopesLand use
Comprehensive Shoreland Protection ActConservation lands
AquifersCoastal Watershed Conservation Focus Areas
FloodplainsWildlife Action Plan habitats and focus areas
Natural Heritage Inventory1Natural Services Network
StreamsImpervious surfaces (where available)
Wellhead Protection AreasDevelopment pressure (to allow for identification of parcels that represent confluence of high-value resources and susceptibility to development)
Distance to town line (to assess whether a proposed development is likely to impact neighboring towns)Zoning districts with parcel data (if/where available)

1 This may ultimately be excluded, as the data are sensitive and a state-sponsored screening tool already exists. If excluded from our tool, we will incorporate a link to the state site.

Reporting
A short report in PDF format will be returned specifying the appropriate spatial statistic for the various datasets (for instance, the report might list maximum, minimum, and mean slope, wetland acreage for the tract, and stream frontage). The report will also include maps, if requested by the user, with each resource displayed.

  • To allow for the imprecision of tract boundaries or to account for the submission of a point, the submitted tract data will be buffered (distance yet to be determined)
  • The PDF report will include an automatically generated locus map (so that the reader can confirm the tract location as submitted).

Data Generated to date
Impervious surface data are still under development, and will be available for release during the next reporting period.

Project Objectives for Next Reporting Period

Objectives
1. Complete impervious surface data
2. Apply impervious surface data to land use categorization to derive IS coefficients
3. Develop and test online GIS planning tool

Work Plan to Meet Objectives
1. Complete impervious surface data

  1. Complete image processing and cleanup
  2. Complete accuracy assessment
  3. Develop metadata and post data to GRANIT web site for distribution
2. Estimate future degree of imperviousness
  1. Calculate impervious surface coefficients based on current land use data
  2. Apply coefficients to land use projections
3. Develop prototype online GIS planning tool using open source software
  1. Restructure web flow
  2. Implement Google Address Locator
  3. Add buffer around a point (PostGIS)
  4. Add data set selector w/ metadata and "last updated"
  5. Develop CGI for functionality (intersection in PostGIS)
  6. Add a pdf report creator
  7. Test the web tool with local partner(s) and revise, as necessary
  8. Test the tool with 3 additional pilot communities (TBD)
  9. Conduct publicity to make other towns in the coastal watershed aware of the tool

Expenditures
Expenditures are in the range anticipated for the work accomplished to date.