Detecting Forest Cover Change and Agricultural Expansion in Western Kentucky Using Satellite Imagery
Academic Level at Time of Presentation
Sophomore
Major
Geology
List all Project Mentors & Advisor(s)
Dr. Haluk Cetin
Presentation Format
Poster Presentation
Abstract/Description
This project looked at the change of forest cover over time in the Western Kentucky region using Landsat satellite imagery. The main objective was to find and show where forested areas and regions with higher vegetation have been lost and ultimately changed over the past several decades. It also explained how these changes have been connected to agricultural expansion and other land use changes. In order to do this, older and more recent satellite imagery datasets were compared to see how this land cover has changed. Supervised classification methods, such as Support Vector Machines, Random Forest and Maximum likelihood, were used to sort the land into categories, such as forest, cropland, pasture, and developed areas. These methods relied on training the classification data to improve the overall accuracy of the data. After having classified these images, how the land has changed over time was analyzed, focusing on the areas affected by agricultural expansion and other land use changes. The spatial patterns of these were examined to see if they are related to the farming practices within the Western Kentucky region. Overall, this project aimed to show remote sensing could be used to track the environmental changes and to help understand the land use patterns in Western Kentucky.
Spring Scholars Week 2026
Sigma Xi Poster Competition
Detecting Forest Cover Change and Agricultural Expansion in Western Kentucky Using Satellite Imagery
This project looked at the change of forest cover over time in the Western Kentucky region using Landsat satellite imagery. The main objective was to find and show where forested areas and regions with higher vegetation have been lost and ultimately changed over the past several decades. It also explained how these changes have been connected to agricultural expansion and other land use changes. In order to do this, older and more recent satellite imagery datasets were compared to see how this land cover has changed. Supervised classification methods, such as Support Vector Machines, Random Forest and Maximum likelihood, were used to sort the land into categories, such as forest, cropland, pasture, and developed areas. These methods relied on training the classification data to improve the overall accuracy of the data. After having classified these images, how the land has changed over time was analyzed, focusing on the areas affected by agricultural expansion and other land use changes. The spatial patterns of these were examined to see if they are related to the farming practices within the Western Kentucky region. Overall, this project aimed to show remote sensing could be used to track the environmental changes and to help understand the land use patterns in Western Kentucky.