Land-Cover Change in the East Central Texas Plains, 1973-2000

Project Background:
The Geographic Analysis and Monitoring (GAM) Program of the U.S. Geological Survey (USGS) Land Cover Trends project is focused on understanding the rates, trends, causes, and consequences of contemporary U.S. land-use and land-cover change. The objectives of the study are to: (1) develop a comprehensive methodology for using sampling and change analysis techniques and Landsat Multispectral Scanner (MSS) and Thematic Mapper (TM) data for measuring regional land-cover change across the United States, (2) characterize the types, rates and temporal variability of change for a 30-year period, (3) document regional driving forces and consequences of change, and (4) prepare a national synthesis of land-cover change (Loveland and others, 1999).

Using the 1999 Environmental Protection Agency (EPA) Level III ecoregions derived from Omernik (1987) as the geographic framework, geospatial data collected between 1973 and 2000 were processed and analyzed to characterize ecosystem responses to land-use changes. The 27-year study period was divided into five temporal periods: 1973-1980, 1980-1986, 1986-1992, 1992-2000, and 1973-2000. General land-cover classes such as water, developed, grassland/shrubland, and agriculture for these periods were interpreted from Landsat MSS, TM, and Enhanced Thematic Mapper Plus imagery to categorize land-cover change and evaluate using a modified Anderson Land-Use Land-Cover Classification System for image interpretation. The interpretation of these land-cover classes complement the program objective of looking at land-use change with cover serving as a surrogate for land use.

The land-cover change rates are estimated using a stratified, random sampling of 10-kilometer (km) by 10-km blocks allocated within each ecoregion. For each sample block, satellite images are used to interpret land-cover change for the five time periods previously mentioned. Additionally, historical aerial photographs from similar timeframes and other ancillary data such as census statistics and published literature are used. The sample block data are then incorporated into statistical analyses to generate an overall change matrix for the ecoregion. For example, the scalar statistics can show the spatial extent of change per cover type with time, as well as the land-cover transformations from one land-cover type to another type occurring with time.

Field data of the sample blocks include direct measurements of land cover, particularly ground-survey data collected for training and validation of image classifications (Loveland and others, 2002). The field experience allows for additional observations of the character and condition of the landscape, assistance in sample block interpretation, ground truthing of Landsat imagery, and helps determine the driving forces of change identified in an ecoregion. Management and maintenance of field data, beyond initial use for training and validation of image classifications, is important as improved methods for image classification are developed, and as present-day data become part of the historical legacy for which studies of land-cover change in the future will depend (Loveland and others, 2002). The results illustrate that there is no single profile of land-cover change; instead, there is significant geographic variability that results from land uses within ecoregions continuously adapting to the resource potential created by various environmental, technological, and socioeconomic factors.