Shoreline Change Detection in Ghana through Satellite Images

• Tech Stack: QGIS, Jupyter Notebook, Colab
• Project Duration: 6 Weeks
• Github URL: Project Link

Ghana's 550 km coastline accommodates major cities, 46 towns, and numerous settlements. Reliant on fishing, smaller communities have traditionally settled close to the shoreline. However, rising sea levels, increased tidal action, and harmful practices like sandwinning have accelerated coastal erosion. Once buffered by wider coastlines during low tides, these areas now face heightened risks including property damage, tidal flooding, and potential loss of life.

Ghana's government has intervened in the past two decades to counter worsening coastal erosion through the construction of sea defense structures like revetments and breakwaters. However, the substantial cost, estimated at $8.25 billion for complete protection, is economically unfeasible. Consequently, prioritization focuses on major coastal communities and key areas. Nonetheless, this approach lacks comprehensive stakeholder input and timely decision-making, emphasizing the need for a shoreline assessment system to inform prioritization based on annual changes.

This project aims at creating a machine learning model for shoreline and coastline extraction from annual satellite images. Establishing a method to evaluate yearly shoreline changes, facilitating objectives like quantifying coastal soil erosion rates and to implement a web-based dashboard to visualize and communicate the observed shoreline changes effectively.

Members: Joseph Moturi, .

More information available on GitHub.