Date Thesis Awarded


Access Type

Honors Thesis -- Open Access

Degree Name

Bachelors of Science (BS)


Data Science


Dan Runfola

Committee Members

Robert Rose

Anthony Stefanidis


Previous studies have used Convolutional Neural Networks for regional detection of deforestation breaks. However, there is limited research into the capability of deep neural networks to identify sudden shifts in global forest cover from satellite imagery. Additionally, many deforestation detection models are trained on region specific data and need manual input thresholds. In this work, we develop a deep learning model to predict the percent of deforestation in a region between two points in time, trained on globally sourced data. Using the before and after satellite images of a deforestation event as inputs, we implemented a two input Convolutional Neural Network with ResNet transfer learning. The model yields a percent estimate of the deforestation that occurred within the region, achieving an error of 7.61%, using thousands of observations across a wide range of bio-climatic regions. These results illustrate the ability of a deep learning model to predict deforestation when it occurs, at a global scale. Our study, which uses random sampling from every continent, suggests the efficacy/possibility of moving from a limited and regional method to a global model.