Save Climate Refugees using Policy Recommender Systems and Informal Settlement Mapping via Machine Learning:
We build a climate justice tool to identify, map and classify regions of high climate risk and informal settlements to monitor and track their geospatial, socio-economic and policy characteristics for global, national and local intervention.
Protect Climate Refugees
Climate refugees are people displaced due to extreme weather events caused due to climate change, which could push an additional 100 million people into poverty by 2030. Currently, extreme weather events and natural disasters contribute to the displacement of more than 200 million people per year. Informal settlements have an estimated growth rate of 9.85%. As of 2021, over 1 billion people who live in slums, are more vulnerable if they live near high-disaster-risk areas. Climate justice recognises that climate change can have a variety of social, economic, public health, and other negative consequences for impoverished populations. Therefore, tracking informal settlements which are high risk is important for climate justice to ensure timely aid and policy intervention.
Our solution is two-fold, first involves identification and tracking of informal settlements, and second is to recommend policy interventions using machine learning. How are we going to achieve that? A web-based dashboard for all stakeholders to engage with. Our pipeline begins with data collection as we collect and process satellite data using open-access Copernicus services. We build a supervised machine learning model with labels of regions of high disaster risk and existing informal settlements using the locations known from the literature. Using the model, we identify regions of informal settlements. We infer the size of informal settlements via satellite data and the spatial resolution of the satellite used. Over time, we aim to use the dashboard to visualise and reflect the impact of climate-based extreme weather events on the size, shape, location, migration and estimated population density of informal settlements. As we deploy the dashboard, we aim to collect data about policy interventions in the regions of interest via our stakeholders over time. We measure the effectiveness of policy intervention first by visualising the changes in informal settlement sizes and second by conducting qualitative surveys and asking climate refugees about their feedback. We incorporate the policy effectiveness and recommendation information on our web-based dashboard as a pop-up text as one hovers over a particular location.
Our solution is aimed to maximize positive impact on people at risk, and efficiency of disaster response. First, sharing the location and migration of vulnerable communities via our dashboard means international, national or local aid can increase the quality of life and protect human rights. Second, it fosters collaboration between NGOs, governments, the private sector and academia to create better intervention strategies to help displaced communities in extreme climate events. Every leader in the public or private sector has a limited time to make quick and efficient decisions. Our tool compresses huge amounts of information to recommend policies and allocate optimal resources for climate refugees. Third, we provide an ethical mapping of regions to prevent data misuse and enable access to geographical data visualisations to only relevant stakeholders. We mitigate any security, privacy or social risks as we do not reveal the exact location of any specific individual. Lastly, our tools can help smart construction and rebuilding long-term housing for displaced minorities in newer and low-risk areas. With our solution, we support not only SDG 13 (Climate Action) but also SDG 11 (Sustainable Cities And Communities) as well. We innovate responsibly by creating new societal value for the poorest of the poor, often ignored in sustainable development goals.