This study explores the potential of a synergistic approach combining Sentinel-2 data and Google Open Buildings (GOB) for mapping urban deprivation over large areas at a 100m spatial resolution. Urban deprivation, including slums, is a crucial aspect of Sustainable Development Goal 11 (SDG): Make cities and human settlements inclusive, safe, resilient and sustainable. Using a Convolutional Neural Network (CNN) and a VGG19 architecture, we experimented with pre-training, self-training, and post-processing using OpenStreetMap to improve classification accuracy. Our results show that combining outputs from both Sentinel-2 and GOB models improves the overall model performance with an F1 score of 81%. The incorporation of post-processing is useful for the final map creation, particularly in correcting for mis-classifications in areas with obvious morphological similarities to deprived areas, such as markets. This approach, which reduces known errors in the model, holds promise for advancing the precision and reliability of urban deprivation mapping over extensive geographical areas.

IDEAMAPS is developing a participatory data-modelling ecosystem to produce maps of deprived areas ("slums") routinely and accurately at scale across cities in lower- and middle-income countries (LMICs). The Ecosystem is co-designed with local stakeholders with the aim of supporting decision-making and evidence-based policymaking. Within the IDEAMAPS Ecosystem, deprivation (multi-dimensional poverty) is modelled via a large set of subdomains (e.g., irregular layouts, waste accumulation) relating to the domains of deprivation (e.g., unplannedness, contamination). Earth Observation (EO) and AI are used to build these models at scale in environments where data is commonly absent. Models are built within intense community engagements using a Participatory AI approach, where empowerment and transparency about inputs, models and outputs are essential. Open and free EO data (e.g., Copernicus) and data with research access (e.g., SDGSat-1) are combined with public, official, and community-generated datasets to produce gridded surface maps of deprivation across cities. These outcomes support bottom-up planning and the localization of SDGs.

Population living in deprived conditions continues to grow, highlighting the urgent need for accurate high-resolution maps and detailed statistics to plan interventions and monitor changes. Unfortunately, data on deprived areas or "slums" is often unavailable, incomplete, or outdated. Leveraging satellite imagery can offer timely, and consistent information on deprived areas over large area However, there are limited studies that use free and open source data that can be used to map deprived areas over large areas and across multiple cities. To address these challenges, this study examines a scalable and transferable modeling approach to map deprived areas using contextual features extracted from freely available Sentinel-2 data. Models were trained and tested on three Sub-Sahara cities: Lagos Nigeria, Accra Ghana, and Nairobi, Kenya. The results indicate that models in individual city achieved F1 scores from 0.78-0.95 for the three cities. Additionally, the results indicate that the proposed approach may allow for the ability to transfer models from city to city allowing for large area and across city mapping.

The number of people living in deprived urban areas within low and middle income countries (LMICs) is large and predicted to continue to grow. Mapping these areas over time and space in a consistent manner is important for monitoring the Sustainable Development Goals (SDGs) and degree of deprivation between cities. This work describes the development of a mapping ecosystem to do this. The first steps are to define urban extents and collect the data needed for developing deprived area models. The goal of the mapping ecosystem is to produce maps of ranges in deprived areas that are comparable between cities and can be used by local governments and researchers to map within city variations in deprivation. By developing and describing this ecosystem, we hope to replicate this in other cities and countries so that this type of work can be expanded.

A continuously growing number of urban inhabitants in Low- and Middle-income Country (LMIC) cities live in deprived areas. Such areas are under-serviced and characterized by poor living and environmental conditions, where the unplanned morphology interacts with physical hazards. While such areas proliferate, climate change has increasingly severe impacts on them. Deprived areas are often located in high-risk zones, e.g., flood zones. However, the absence of global databases on such areas is an obstacle to locate and prioritize hotspots of deprived communities exposed to climate change. Consequently, quantifying the numbers of exposed inhabitants is not possible, though it is required in support of the Sustainable Development Goals (SDGs) (e.g., 11, 13) and local adaptation strategies. Earth observation (EO) that allows producing such data fall short of providing city-level information due to computational constraints and unsolved methodological challenges related to scalability and transferability. This paper presents a framework to combine EO data with data on hazards (e.g., storms, floods) that impact urban areas and, in particular, deprived communities. First results in pilot cities show that deprived communities are systematically more exposed to physical hazards as compared to formal built-up areas. These problems are expected to intensify in the context of climate change, as most hazards will increase in their severity and frequency.