Update under way for the five-state region encompassing Arizona, Colorado, Nevada, New Mexico, and Utah. State coordination for the project is facilitated through the SWReGAP website (http://leopold.nmsu.edu/fwscoop/swregap/default.htm).
The RS/GIS Lab at Utah State University is the regional land cover mapping lab for the five-state southwest region. The regional project focused on 4 major objectives in 2004: 1) completing land cover modeling activities; 2) model validation; 3) producing a regional mosaic of the mapping zones; and 4) developing a data delivery system for the provisional land cover product.
The majority of natural and semi-natural land cover classes were modeled using a decision tree (DT) classifier. Advantages of DT include the ability to use both continuous and categorical predictor datasets with different measurement scales, good computational efficiency and an intuitive hierarchical representation of discrimination rules. Decision tree models were validated by generating initial models using 80% of available samples, while withholding 20% of samples. Withheld samples were randomly selected and stratified by cover class. Withheld sample polygons were intersected through the land cover map to create an error matrix, presenting users, producers and overall “accuracies.” Using the 4 km overlap region between mapping zones a “cutline” was used to edge-match adjacent mapping areas where land cover discontinuities resulted from the modeling process. The resulting 5-state region mosaic was qualitatively reviewed by the five state teams and NatureServe. Following review a limited number of errors were “flagged” for final editing. The “edits” determined to be relatively easy to correct with localized recoding, or a simple conditional model, were made to the regional map.
The SWReGAP land cover dataset is currently available to the public with ‘provisional’ status from http://earth.gis.usu.edu/swregap/ . See Figure 1 below. Because the dataset encompasses such a large region, the website allows users to download specific geographic segments of the region, such as individual states, counties, or ecoregions. Additionally the website offers an internet map server from which users can interactively clip a specified rectangle in the region. The clipped dataset is subsequently bundled with metadata and made available for download.
.
The regional project focused on six objectives during 2004. These objectives included: 1) collection of habitat modeling attributes; 2) creating region-wide modeling datasets; 3) creating a database to facilitate association compilation, expert review and modification, and potential end user application; 4) internal and expert review; 5) accuracy assessment; and 6) conducting a regional animal habitat modeling workshop in Las Vegas, Nevada in March 2004.
The region is working with a MS Access database to facilitate data collection and to compile taxa specific information for modeling purposes. The intent was to create a dataset that manages information and was used to construct each taxon’s wildlife habitat relationship model. Included within the database is a method for defining range limits using the 8-digit hydrologic unit code (HUC). The database also incorporates the core data layers the region had identified to be minimally addressed in each wildlife habitat relationship model. These core data layers are land cover, elevation (minimum and maximum), slope, aspect, soils, hydrology (distance to and association with permanent water), and patch size. Species were allocated to each state based on expertise and species distribution. These states were responsible for creating the habitat models for those species. The individual databases were then combined and currently any modification to the database is done through online connection to the master database. All species data collection is complete as of this report with modifications occurring as internal and external reviews are completed.
The region has undergone an internal review process to check consistency within the models and to provide the framework for an external review. The internal review is complete as of 14 February 2005. Expert review is beginning as of 1 February 2005 and is scheduled to be completed by 30 April 2005. The region will complete the standard gap analysis habitat modeling measure of agreement as well as a measure of agreement with existing species occurrence records. States are currently identifying qualified species list for the standard measure of agreement. These lists will then be provided to the regional laboratory. The Arizona project, in coordination with the regional laboratory, is identifying a procedure to use existing data to measure the degree of concordance between habitat models and species occurrence records. This analysis will be done as models are completed, and is scheduled to be finished by July 2005.
The final regional stewardship and management status map is expected to be complete by June 2005. External review of stewardship mapping products began in December 2004. Nevada stewardship and management status maps have undergone the external review process resulting in updates to internal parcel boundaries and refinements to the GAP status codes. The draft maps for Colorado are also complete and the external review is scheduled for March 2005. External reviews are also scheduled for Arizona, New Mexico, and Utah this spring. As a conservative estimate the regional stewardship lab has collected over 300 management planning documents from various federal, state, and county entities. The process of reviewing current management plans, interviewing various land stewards, and assignment of the GAP Status Codes is complete. In addition, most of the digital boundary information for all five states has been collected, and cooperators have been generous with providing digital parcel data layers. Currently, the regional stewardship lab is in the process of assembling the GIS database using the geodatabase format to maintain data integrity. In an effort to keep the stewardship mapping effort consistent across the region, the regional stewardship lab digitizes additional internal information when digital information is unavailable from local sources. This effort hopes to provide a consistent product across the region in both the level of mapping detail and assignment of the GAP status codes. In addition, the detail and refinements of the stewardship product will create a better assessment in the final gap analysis.
Analysis for SWReGAP will take place when all mapping tasks are completed. Land cover analysis and animal habitat modeling analysis will begin in May 2005.
All products derived from the Southwest Regional Gap Analysis Project are scheduled to be complete by approximately October 2005.