Current Assessment and Future Prediction of Forest Cover Change in Cumberland and Morgan Counties, Tennessee: A Modeling Technique

Determining the relationship between human disturbance of the environment and natural forest change is critical for sound natural resource planning. Improved land cover modeling techniques that incorporate geographic information systems and statistical models are needed to assist in this analysis. Continued forest fragmentation due to increasing population and urbanization has created a growing interest in forest protection for the Cumberland Plateau of Tennessee. Specifically, Cumberland and Morgan Counties have seen unprecedented population growth over the last two decades, resulting in fragmentation of forestland. This study developed a model to determine the probability of exurbia development and its resulting forest fragmentation. Geographic data used in the research included satellite imagery from 1992 and 2000, U.S. Census population and demographic estimates, and road and water coverages for the two counties. The first objective of this study was to develop an accurate and efficient procedure for the development of a land cover map for use in a forest change detection system for Cumberland and Morgan Counties, Tennessee. A unique method was developed to generate this procedure by combining post-classification and image differencing. The second objective of the study was to determine the relationship between urbanization and forest loss in Cumberland and Morgan Counties, Tennessee, and to predict current and future land cover patterns. Logistic regression analysis suggested that demographic variables such as education and population along with spatial factors such as slope, distance to water, distance to interstate junctions, and gravity index factors of nearby urban retail centers, significantly influenced the transition of forest to urban cover. Of these parameters, a high gravity index, a suburban designation, and unsloped terrain had the greatest impact on forest to urban conversion. In addition, spatial factors such as parcel distance to water, and parcel distance to interstate junctions significantly influenced the probability of development. Finally, using population density predictions, the model identified the probability that forest land would be urbanized by 2010