Identification of climate change-related hazards in informal communities through the application of machine learning to satellite images.

We describe a new Artificial Intelligence (AI) tool, specifically a Machine Learning (ML) tool that can, through an integrated analysis of a large variety of data including satellite or aerial images, provide a rapid and low-cost identification of exposure to climate-related hazards for informal settlements. The tool can identify exposure to these hazards more economically, quickly, frequently, and transparently than current approaches, and can facilitate visual accessibility to otherwise inaccessible areas. With wide potential availability to public officials and the public, it will add value to planning practices, municipalities and communities, leading to more inclusive and equitable cities, and to governance
processes that address the joint impacts of rapid informal urban expansion and climate change. Prototyping will be based on data from Tegucigalpa, Honduras. A second stage will expand the tool into a wide-range planning function by adding scenario analysis and enhancing ML capabilities.