Nick Van Driel and Tom Loveland

The U.S. Geological Survey's Land Cover Characterization Program

Abstract

The U. S. Geological Survey's (USGS) National Mapping Division has initiated a Land Cover Characterization Program in response to the need for land cover and vegetation data for inventory, monitoring, modeling, and management in the public and private sectors. The Program's general goal is a multi-scale, multi-purpose land characteristics data base. Customer needs, source data, and analytic techniques, form the basis of the program's four components. The small scale component uses satellite imagery acquired by the National Oceanic and Atmospheric Administration's (NOAA) advanced very high resolution radiometer (AVHRR). Multidate satellite imagery have been analyzed and combined with ancillary data to produce a multilevel, geographically referenced land cover data base for global change research. Landsat thematic mapper data are the source for the intermediate-scale component that delivers processed satellite data and ancillary data sets to cooperators. The USGS will synthesize a national land cover characterization from cooperators' results. Current activity in the large-scale component uses digital orthophotoquadrangles (DOQ) as a source for land cover and land use interpretations in selected urban areas for the National Water Quality Assessment (NAWQA) Program. A special projects component will accommodate cooperative activities in areas where standard products are not satisfactory. The USGS Land Cover Characterization Program is compatible with current concepts of government operations, the changing needs of the land use and land cover data users, and the technological tools with which the data are applied.


Historical Perspective

The USGS has a long heritage of leadership and innovation in land use and land cover mapping. The USGS Anderson system (Anderson, and others, 1976) defined the paradigm for land use and land cover mapping that has been the model both nationally and internationally for over 20 years. USGS scientists successfully applied these principles by mapping the United States. Recently, the USGS demonstrated the utility of multi- purpose land cover characteristics data bases, which build on the previous USGS efforts (Loveland, and others, 1991; Reed and others, 1994). The land characterization approach involves the interpretation of multi-resolution and multi-temporal data into flexible data bases describing landscape types, processes, and conditions. The results of the research phase have been broadly accepted within the national and international environmental assessment community (Steyaert, and others, 1994).

The land cover characterization program (LCCP) is founded on the premise that the nation's needs for land cover and land use data are diverse and increasingly sophisticated. The range of projects, programs, and organizations that use land cover data to meet their planning, management, development, and assessment objectives have expanded significantly. The reasons for this are numerous, and include the improved capabilities provided by geographic information systems (GIS), better and more data intensive analytic models, and increasing requirements for improved information for decision making.

Guiding Principles

The LCCP builds on the heritage and success of previous USGS land use and land cover programs and projects. It will be compatible with current concepts of government operations, the changing needs of the land use and land cover data users, and the technological tools with which the data are applied. The program is founded on the following guiding principles:

Land characteristics data bases that are compatible with previous USGS land use and land cover products and compatible with Federal Geographic Data Committee (FGDC) standards. This is essential so that there is continuity that allows analysis of land use and cover change.

Inter-agency partnerships. The USGS will actively solicit the involvement of other federal and state agencies in all aspects of the program, including funding, planning, mapping, and applications. This program will assist federal agencies in combining their agency efforts to characterize the landscape. No single organization can afford the cost of a national land characterization program.

There is no one product that satisfies all users or their applications. Therefore, the program will be based on a flexible data base strategy that facilitates both standardized categorized land use, land cover, and land characteristics describing landscape processes and dynamics.

Multi-resolution products will be designed for use at the local, regional, and national levels.

User driven and implemented on a cooperative basis. This will assure allocation of efforts to pressing national priorities.

Provide cyclic data that permit analysis of landscape change.

The LCCP is based on an integrated strategy of research, applications, production, and data management.

Project Components

One land cover product for the Nation would simply be inadequate. The LCCP thus consists of four components, each with unique but complementary land cover, land use, and ancillary data products which are:

1.Small scale (1:2,000,000) national and global land cover characteristics data produced periodically from coarse resolution (e.g., 1-km) remotely sensed data.

2.Intermediate scale (1:100,000) national land cover characteristics data produced on a cyclic basis from intermediate scale remotely sensed data such as data from the Landsat thematic mapper (TM).

3.Large scale (1:24,000) land use/land cover data for metropolitan areas and other areas experiencing rapid growth. Generally, this component will use the digital orthophotoquadrangle data as a mapping source.

4.Special projects will be conducted, as needed, to produce unique land cover characteristics data in situations in which the first three standard products do not satisfy user requirements.

As stated in the guiding principles, all four components will be developed through partnerships that include public and private participants. The following sections summarize the key aspects of each component. Note, however, that product definitions, timing, and other specific features will be thoroughly reviewed and refined during the first year of the program.

Small - Scale Component

The small - scale product will be a land cover characteristics raster data base with a grid cell size of 1-km2. The product will be designed for use in ongoing national operational programs with requirements for coarse resolution, nationally consistent data. This data set will be produced on a ten-year cycle corresponding to national population census periods. An experimental 1990 data base produced by the USGS is currently available (Loveland and others, 1993). The next national update will be in the year 2000.

Initially, the data set will be produced from advanced very high resolution radiometer (AVHRR) satellite data. Data from advanced sensors, such as the Moderate resolution imaging spectrometer (MODIS) may be used in the future. This product includes:

Seasonal land cover regions as the spatial component. They represent common mosaics of land cover, phenology, and landscape biophysical processes. See Loveland and others, 1995 for a description of seasonal land cover regions.

Attributes describing each region, including land cover types, vegetation components, phenology, spectral measures, and site characteristics (political boundaries, hydrology, elevation, general soils, climate, and ecoregions).

Derived thematic maps that are based on a translation of the seasonal land cover regions into common land cover classification legends; these include USGS Anderson System (Anderson and others, 1976), National Terrestrial Land Cover (Jennings, 1995), Biosphere-Atmosphere Transfer scheme (Dickenson, 1986), and Simple Biosphere Model (Sellers and others, 1986).

Source data used in the development of the data base, including all non proprietary satellite data and related earth science data sets.

A global land cover characteristics data base will be also produced as part of the LCCP. This effort is underway; a global 1-km data base produced from 1992-1993 AVHRR data will be completed by 1997. This product is intended for use in continental to global scale studies.

Intermediate-Scale Component

This product will be the primary detailed national land cover product. The data base, developed from Landsat TM satellite data or an equivalent source, will be a raster product with 30-m resolution. The data set is intended for use in regional (e.g., state, multi-county, ecoregions) land management, planning, and environmental assessment applications. As currently planned, this component will be implemented in two phases:

Phase 1: Every five years, a national preprocessed Landsat TM data set will be released by the USGS to United States Government cooperators. This data set will be georeferenced and will include appropriate ancillary data.

Phase 2: Every 10 years, the USGS will synthesize a national land cover characteristics data base. The process for the generation of this product will be based on interpretations made from a consistent, preprocessed Landsat TM data base. The primary source of these interpretations will be projects conducted by state and federal agencies, and the private sector. For example, the National Biological Service Gap analysis program, with state-level land cover mapping projects in most states, will be a key source of project-level land cover data. While the USGS will not dictate a standardized land cover legend for these projects, a set of specifications will be developed regarding minimum documentation standards. To facilitate access to the project-level data sets, the USGS will maintain an archive of land cover products developed from the national TM data base. As national land cover interpretations are completed, the USGS will synthesize the project-level interpretations into a nationally consistent classification legend, add appropriate attributes, and distribute the data base to anyone. This component is based on the current interagency land cover mapping initiative coordinated by the Multi-Resolution Land Characterization (MRLC) consortium (Loveland and Shaw, 1995).

Large-scale component

The large-scale component will emphasize land use and land cover. Keeping with the approach developed by Anderson and others (1976), land cover will be used as a surrogate for land use. The large scale data set will be vector format and will be developed from the interpretation of digital orthophoto quadrangle data. The data set will be consistent with the 1:24,000-scale USGS quadrangle maps. Unless special circumstances arise, source data will be used that can be released with the land use and land cover interpretations. This data set will be produced for metropolitan areas and other rapidly growing parts of the country. The target is to produce land use and land cover data every 10 years for all Standard Metropolitan Statistical Areas. The classification legend will be similar to the USGS Anderson System so that comparisons with the previous USGS land use and land cover data set are possible. However, refinement of the Anderson System is likely. The LCCP will work closely with the proposed FGDC Land Cover Subcommittee and with the NAWQA Program to define the final classification legend.

Special Projects

In cases where standard products will not satisfy user requirements, cooperative activities will be initiated to generate the necessary products. The modes in which special projects are conducted will be determined on a case-by-case basis.

Coordination

Many organizations and individuals conduct land cover and vegetation mapping, often without knowledge of each other's work. However, many of these activities are known to analysts and managers in the LCCP, and several initiatives are underway to provide coordination at state, regional, and national levels. Specifically, the Department of the Interior has created a Land Cover Working Group under the Interior Geographic Data Committee (IGDC). This working group, with membership from all DOI bureaus, will gather and publish information on each bureau's land cover and vegetation mapping plans, current activities, and related information sources. This information will be made available through the World Wide Web. In FY97, the IGDC Land Cover Working Group plans to expand its membership to other agencies under the FGDC. Because the FGDC includes 14 agencies, this working group will provide a coordination forum for all federal land cover and vegetation mapping activities. Some of these federal activities include coordination components at the national, regional, and state levels; these components will be used and expanded where appropriate. In addition, the LCCP will follow the successful MRLC Program example of coordination at the project level among agency programs with requirements for land cover data. MRLC participants include the USGS, NBS, EPA, NOAA, and USFS.

Outreach Activities

Communication of the LCCP's activities and goals will be the program's outreach focus in fiscal year 96. Information will be delivered through existing land cover and vegetation mapping projects and cooperatives, on the WWW under the USGS National Spatial Data Infrastructure node, through committees at bureau, department and agency levels, and by presentations at professional meetings and conferences. These activities will announce the program's products and schedules, and will encourage partnerships among individuals and groups with similar goals.

Standards

The USGS will work with the FGDC to establish a land cover working group. The initial charges of the working group will be to share information about federal plans and activities, develop a national land cover legend for the intermediate scale synthesized product, assist the FGDC land cover classification subcommittee and to write documentation standards for federal land cover projects.

The USGS will lead an effort to develop a crosswalk or conversion process for existing land cover classifications. With this conversion capability, valuable maps and data can be used as historic references in landscape monitoring activities.

Conclusion

Land cover, land use, and vegetation mapping data are essential elements in a wide array of government and private sector activities, including inventory, management, monitoring, and modeling. The time and expense required to develop land cover and vegetation data demand efficient collection and analysis procedures that allow multiple use of a single product. The LCCP reflects the U.S. Geological Survey's commitment to providing high quality land cover data for effective stewardship of the nation's resources.

REFERENCES

Anderson, J.R., Hardy, E.E., Roach J.T., and Witmer R.E. 1976. A Land Use and Land Cover Classification System for Use with Remote Sensor Data. U.S. Geological Survey Professional Paper 964, Reston, VA: U.S. Geological Survey.

Dickinson, R.E., A. Henderson­Sellers, P.J. Kennedy, and M.F. Wilson. 1986. Biosphere-Atmosphere Transfer Scheme (BATS) for the NCAR Community Climate Model. NCAR Technical Note NCAR/TN - 275+STR, Boulder, CO.

Jennings, M.D., 1995. Nomenclature and mapping units for gap analysis land cover data, in Technologies for Biodiversity Gap Analysis:Proceedings of the ASPRS/GAP Symposium, Charlotte, NC (in press).

Loveland, T.R., Merchant, J.W., Ohlen, D.O., and Brown, J.F., 1991. Development of a Land Cover Characteristics Data Base for the Conterminous U.S., Photogrammetric Engineering and Remote Sensing, v. 57, n. 11, p. 1453 - 1463.

Loveland, T.R., Ohlen, D.O., Brown, J.F., Reed, B.C., Merchant, J.W., and Steyaert, L.T., 1993. Prototype 1990 Conterminous United States Land Cover Characteristics Data Set CD-ROM, USGS CD-ROM Set 9307.

Loveland, T.R., and Shaw, D.M., 1995. Multiresolution Land Characterization:Building Collaborative Partnerships, in Technologies for Biodiversity Gap Analysis: Proceedings of the ASPRS/GAP Symposium, Charlotte, NC (in press).

Reed, B.C., Loveland, T.R., Steyaert, L.T., Brown, J.F., Merchant, J.W., and Ohlen, D.O., 1994. Designing Global Land Cover Databases to Maximize Utility, in Environmental Information Management and Analysis: Ecosystem to Global Scales, W.K. Michener, J.W. Brunt, and S.G. Stafford, editors: London, Francis and Taylor, p. 299 - 314.

Sellers, P.J., Y. Mintz, Y.C. Sud, and A. Dalcher. 1986. A simple biosphere model (SiB) for use within general circulation models. Journal of Atmospheric Science 43: 505-531.

Steyaert, L.T., Loveland, T.R., Brown, J.F., and Reed, B.C., 1994. Integration of Environmental Simulation Models with Satellite Remote Sensing and Geographic Information Systems Technologies: Case Studies, in Proceedings:Pecora 12 Symposium on Land Information from Space-Based Systems, American Society of Photogrammetry and Remote Sensing, Bethesda, MD, p. 407 - 417.


Nick Van Driel
Tom Loveland
EROS Data Center
Sioux Falls South Dakota, 57198 Phone: 605-594-6007
FAX: 605-594-6529