Structure
The concept of the project is based on the general notion that the intensity of extreme climate events is not the only factor of relevance to climate change impact, but the social component is equally important. Therefore, our methodology will combine the physical, social and institutional components in climate vulnerability assessment. In this study, a deep learning approach that integrates EO and CS to map slums and understand their climate vulnerability at the neighbourhood level will be designed, implemented and tested at six large and secondary cities in Africa. The targeted large cities include Accra in Ghana, Lagos in Nigeria and Nairobi in Kenya. Secondary cities include Tema in Ghana, Kisumu in Kenya and Akure in Nigeria. These cities are selected because of our strong collaborations with existing local CS groups such as YouthMappers, Missing map, Local Humanitarian OpenStreetMap Team (HOT), NGOs (e.g., Lagos (JEI), Nairobi/Kisumu (CommunityMappers), Accra (People’s Dialogue)). These collaborations will facilitate our interactions with local slum communities. In addition, these cities are threatened by flooding [3], [4]. The methodology is divided into five work packages (WPs), as shown in Fig. 3. Two PhD students will carry out the work in collaboration, one focussing on the EO deep learning components and one on the climate vulnerability modelling.
The project will be structured into five main Working Packages (TP):