This project, led by the Center for International Earth Science Information Network at Columbia University, contributed to filling the gaps in spatial coverage of population distribution and to increasing the visibility of off-grid, hard-to-reach populations to various development and adaptation actors. Such populations are often underrepresented and underserved in development projects, lack access to basic services including sanitation, health and education, and lag behind on achievement of the Sustainable Development Goals (SDGs).
This project addressed the following SDGs: SDG 1 - No poverty, SDG 3 - Good health & well-being, SDG 4 - Quality education, SDG 13 - Climate action, and SDG 17 - Partnership for the goals. To achieve its objective, the project built a publicly accessible portal--Mapping Hard to Reach Villages (MAHRV)--allowing anyone to contribute and access population data. The project focused on people living in mangrove areas of coastal West Africa. The project’s methodology innovatively combined the analysis of high-resolution satellite imagery, Volunteered Geographical Information (VGI) and statistical modeling. Villages and their extents were identified from the High-Resolution Settlement Layer (HRSL), taking advantage of machine learning approaches to remote sensing problems. Village extents and settlement-level population data, contributed through the VGI, were used in statistical modeling in conjunction with other covariates to further infer population counts in villages for which in-situ data were not available. Both processes were made publicly accessible through the MAHRV portal. Population estimates will be iteratively improved as more in-situ data become available through the VGI data collection process. Data validation and model updates run in the background. ArcGIS Online (AGOL) was selected as the primary technology for the Online Services, Web Applications and Mobile Applications, as it is supportive of easy technology transfer. for the Online Services, Web Applications and Mobile Applications, as it is supportive of easy technology transfer.
The framework of software and services allows for the update of the population model and public data product with a minimum of effort. Overall, the new interface allows an increase in spatial resolution of village extents, thus the identification of very small villages, and provides population estimates at a granularity never before achieved at regional scale (most of West Africa mangroves). While this project focused on villages within the mangroves in West Africa, the approach and the technologies are easily transferable to other geographies.
Disclaimer: As many of the links are external, some links may not work.