Sources and Methods
County level Data about population and land use for the U.S. Great Plains from 1870 to the present.
Our county-level social, demographic, and agricultural data come from:
- The United States Census of Population
- The United States Census of Agriculture
- Vital statistics records from the United States Department of Health
- The Inter-university Consortium for Political and Social Research
In addition to using these data ourselves, we have also made this valuable data collection available to the broader scientific community through ICPSR and on the web (link to data page).
Environmental data for the U.S. Great Plains from 1895 to the present.
To the social, demographic, and agricultural data we have added environmental data, including:
- Weather
- Elevation
- Soil quality
- Natural amenities
- Distance between counties
Environmental information comes from:
- PRISM precipitation and temperature data
- VEMAP wind and humidity data
- SSURGO and STATSGO soils databases
We have interpolated these data to historical county boundaries, allowing us to link population dynamics and land use to environmental features and thus examine the impact of environment on land use and population change through both family formation and migration at the county level.
Great Plains Farm Family Survey.
A set of roughly 150 interviews with farm families in five areas of the Great Plains, conducted from 1997 to 1999, which provide valuable information about:
- Farm family incomes
- Responses to conservation measures
- The ways in which farmers see and use government programs
Through these interviews we attempt to demonstrate how families make demographic and land-use decisions in response to social, economic, political, and environmental forces. Linking these interviews to county- and individual-level census data will allow us to deal specifically with regional, environmental, and ethnic differences in the study population.
Century and DayCent Ecosystem models.
These generalized ecosystem models simulate carbon and nitrogen dynamics and greenhouse gas fluxes in grassland, forest, savanna, and crop systems. Our historical land use data provide parameters for these models, allowing us to simulate the outcomes of land use management practices on Great Plains agroecosystems. As land use is tied both directly and indirectly to population change, these models will reveal the effects of population on the ecosystem itself, in turn producing data about the environment of the Great Plains, which we can use in conjunction with our social and demographic data to analyze the interactions between environmental and social processes.
By examining the impact of agricultural practices on trace gas fluxes and soil carbon levels, we have developed carbon and nitrogen budgets for the counties of the Great Plains and determined best management standards for Great Plains grassland systems. Our data allow us to use the Century and DayCent models to determine the environmental consequences of Great Plains agriculture and of the recent and ongoing conversion of agricultural land to residential lawns and other non-agricultural uses.
Field-level land use data from 1930 to the present.
The interpretation of aerial photographs and satellite images by the Environmental Spatial Analysis Laboratory at the University of Michigan will produce data about land use practices at the field level for a sample of fifty-two counties representing the shortgrass, mixed grass, and tallgrass plains. Within each of the fifty-two sample counties, we will sample an average of eight sites over five time points, dating back to the 1930s.
These field-level data will allow us to test our hypotheses about the limits of human impact on the environment and about the stability of agricultural land use in the Great Plains by revealing land-use changes at the field level that may have been masked by county-level aggregation. For example, the county-level data analyzed by Geoff Cunfer in On the Great Plains demonstrate that land in crops reached a maximum in the 1930s and has changed little since. The field-level data will reveal whether individual fields and parcels were converted from one use to another If we find that land use changed often in many places, then human choices were more important than environment in driving land use; if land use remained generally stable across many decades, then environmental drivers come to the fore.