National GAP Analysis
US Geographical Survey

Remote Sensing and GIS Laboratory at Utah State University

Mapping Natural and Semi-Natural Land Cover

The mapping of natural and semi-natural land cover for Utah and the other participating states is a multi-year effort. This initial year will focus on mapping protocol development, the acquisition and interpretation of ancillary data layers, and the integration of The Nature Conservancy (TNC) alliance level classification for land cover in Utah. This work will also support the regional effort by collaborating with individual state investigators to integrate the TNC alliance scheme. This preliminary work will be focused at the regional level. The methods used by different state GAP projects to map land cover / land use in the past have to date been variable due to a "state oriented" focus. By modifying the focus of the mapping effort to a regional, multi-state program, the ability to evaluate biodiversity and identify management gaps on a regional basis increases.

Land cover / Land use mapping will use newly acquired Landsat 7 Thematic Mapper Data composited into biogeographically unique mapping units that span political boundaries. The development of the individual mapping units is nearing completion. These mapping units, as described above in the progress-to-date section, uses a combination of topography and soils information to identify relatively large geographic areas that have similar topographic and biologic (soils, vegetation, etc.) characteristics. By segmenting the 5-state region into mapping zones we reduce the confusion and work required to interpret imagery over a diverse environment. Furthermore, since these boundaries are biogeographic in nature and not political, we increase the ability to edge match these units into a 5-state mosaic of vegetation. This process was used in Utah, Nevada and Southern Idaho vegetation mapping projects with success. 

Data from Landsat 7 will not be completely available until the summer of 2000. During this initial year, USU will collaborate with the USGS to develop the protocols to mosaic these images in a seamless, radiometrically correct composites. This work will follow from research and experience gained from the initial Utah Gap effort and the recently completed Mojave Desert Ecosystem Initiative which produced a four-date set of radiometrically corrected image mosaics for the 80,000 sq. mile Mojave Desert Ecoregion. 

The USGS, EROS Data Center will be responsible this initial year to develop the imagery database from Landsat 7 as images become available. As part of the data development, all images will be corrected for system errors, terrain displacement, and geometrically corrected to an average root mean square error of 1 pixel. All images will be radiometrically corrected and combined into mapping unit mosaics by the EROS Data Center. With the current budget at EROS Data Center 2 mapping zone mosaics per state will be made available as imagery is collected. Additional mosaics will be done on a negotiated cost reimbursable basis by USGS. USU will work closely with the USGS to develop the mosaic protocols. USGS will have the primary responsibility to generate the mosaics. The imagery database will consist of all 7 Thematic Mapper bands in addition to the 15 meter panchromatic layer. 

Ancillary data layers processed by USU and the USGS for this year will consist of a minimum of the 30 meter elevation database for the 5-state area (USGS) to assist in the vegetation and topographic modeling phase. This elevation database will be processed to generate slope, aspect, and topographic position layers. We will compile all NOAA-AVHRR 10-day composites generated since 1990 to help identify phenological fluxes within the 5-state area. This data layer can provide insight on yearly growth patterns of general cover types. Transportation networks to aid in field visits, drainages to help identify potential riparian areas, additional interpretation of the soils data layers, and as many cartographic data layers as possible. 

The development of ancillary data layers, and image mosaic and mapping protocols will be focused on the entire region. It is estimated that approximately 20% of this effort will be focused on Utah. For Utah specifically, work will center around the further collection and analysis of existing training sites, cover-type development (from TNC) and field sampling planning and method development. These activities will have influence on the entire region, but is primarily focused on the state of Utah.