Hierarchical Investigation of Socioeconomic Drivers of Decadal Scale Land-Cover Changes in the Upper Midwest.

Funding is from NASA’s Land-Cover and Land-Use Change Program (NAG5-6042.) 1997-2000.

Dan Brown,

Bryan Pijanowski and

Mike Vasievich.


This project examines changes in land cover patterns as a function of socioeconomic changes, dispersed development, and subsequent changes in the spatial patterns of land ownership in the Upper Midwest (defined to include forested regions of Michigan, Minnesota, and Wisconsin). Spatial patterns of land ownership (i.e., parcel size and pattern) and their trajectories through time provide a critical link in understanding the environmental implications (i.e., LCLUC) of societal and economic factors.

This project involves (a) mapping multi-temporal patterns of land cover from historical aerial photographs (at resolutions similar to the TRW/Lewis and CTA/Clark panchromatic data) and NASA Pathfinder data (i.e., North American Landscape Characterization--NALC) and (b) modeling rates and types of change in land cover patterns as functions of rates of change in land ownership patterns and socioeconomic factors. Spatial and temporal exploratory statistical analyses will be used to quantify the relationships between socioeconomic processes, changes in the patterns of land ownership and cover patterns observed for a set of 136 area frame samples throughout the region. The observed relationships are being used to parameterize a GIS and Artificial Neural Network (ANN) -based, spatial-temporal model of land use/cover and pattern change using the Land Transformation Model. The predictive ability of the spatial-temporal model is being be assessed at the pixel, site, county, and regional scales using NALC MSS triplets. The LTM has now been used to examine such effects as policy, economics, demographics, landscape features on land use change in a variety of locations in the Upper Great Lakes States.


The 136 area frame samples, shown in Figure 1, are stratified across the Upper great Lake States region. Stratification was based on county population, proximity to roads and state parts and federal forests. The NALC images selected for forest cover study are shown in red. The area frame samples shown as small squares on this figure are located in 17 counties (Table 1) in the three states.

Table 1. Counties located in the NAA study area.

Michigan      Luce
Grand Traverse     
Wisconsin Florence
Minnesota Lake of the Wood

Click on the image to enlarge


Location of the study areas (counties in purple) and Lnadsat TM images (red boxes)




Key Publications

Brown, D., B. Pijanowski and J. Duh. 2001. Modeling the Relationships between Land Use and Land Cover on Private Lands in the Upper Midwest. Journal of Environmental Management. 59:247-263.

Pijanowski, B.C., D. G. Brown, G. Manik and B. Shellito. 2002. Using Neural Nets and GIS to Forecast Land Use Changes: A Land Transformation Model. Computers, Environment and Urban Systems 26(6) 553-575.


Last updated by BCP on October 26, 2002.
Copyright by Purdue University