The Repository @ St. Cloud State

Open Access Knowledge and Scholarship

Date of Award


Culminating Project Type


Degree Name

Geography - Geographic Information Science: M.S.


Geography and Planning


School of Public Affairs

First Advisor


Second Advisor


Third Advisor


Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Keywords and Subject Headings

Deforestation, Predictive Modelling, Cross River National Park, GIS, Remote Sensing, Machine Learning


Population growth, urban sprawl, agricultural expansion, and illegal logging has led to losses in forested land in most parts of the world, especially in a highly populated country like Nigeria. The Cross River National Park (CRNP) in southeastern Nigeria with an area just above 4000km2 is designated a biodiverse hotspot and one of the oldest rainforests in Africa. As with all other tropical forests spread across the globe the CRNP is not immune to these factors that threaten its existence. The focus of this study is to analyze the change of forest cover at the Oban division of the Cross River National Park using multi-temporal remotely sensed data to predict and model the future probability of deforestation within the area of interest. This study made use of the Landsat West Africa Land Use/Land Cover Time Series dataset for the years 1975, 2000 and 2013 and Landsat 8 operational land imager (OLI) imagery for the year 2020 in a post classification change detection model to determine the extent of change in forest cover classes. Random forest decision tree machine learning algorithm was used to predict the future risk of forest cover loss using the datasets produced from the post classification change detection. The model related deforestation probabilities with several physical and anthropogenic factors such as elevation, slope angle, solar radiation, aspect, topographic roughness, soil type, distance from roads, distance from towns, distance from rivers, distance from plantations and population density. The results from the change detection analysis showed that from 1975 to 2020 the forest cover declined by 1909km2 a rate of 42km2 per year. The random forest regression analysis predicted areas of the forest with modest to high deforestation probabilities and indicated that socio-economic factors are major drivers of deforestation in the region rather than physical factors.


I would like to thank the members of my thesis committee:

Dr. Kirk Stueve (Chair), Dr. Mikhail Blinnikov and Dr. William Cook, their invaluable advice and mentoring was crucial in completing this paper

My eternal gratitude to my family who stood by me throughout my studies in SCSU, Ngozi, Darren and Dora this is for you.