Using Geospatial Data Science to Identify Vulnerable Communities to Climate Change

Keywords: Climate change, big data, Twitter, flooding, vulnerability
Session Type: Virtual Paper Abstract
Day: Tuesday
Session Start / End Time: 3/1/2022 05:20 PM (Eastern Time (US & Canada)) - 3/1/2022 06:40 PM (Eastern Time (US & Canada))
Room: Virtual 18

Abstract

Communities face unprecedented risks due to climate change including coastal and inland flooding due to rising sea levels and precipitation. Studies demonstrate that some communities are especially vulnerable due to socio-economic and demographic status. Another crucial but understudied risk is climate change denialism, in that some individuals and communities are rendered more vulnerable because they choose not to take necessary steps to adapt. We use state-of-art spatial data science and big data to assess these risks and to create a climate change vulnerability index for the State of Michigan. The index will combine three pillars: 1) flooding vulnerability; 2) building condition; and 3) climate change opinion (belief vs skepticism). We use millions of tweets reflecting opinion of climate change and spatially integrate it with flooding risk and maintenance conditions for residential properties in Michigan (~5 million). This prototype model explores how to optimally integrate all these heterogeneous data using machine learning techniques and language recognition. The resulting integrated vulnerability map is expected to have sufficient spatial resolution to enable state and local municipalities to identify vulnerable communities and to foster resilience to climate change.