← Past projects
Dominican Republic · 2025
Quality of life in Greater Santo Domingo
Predicted household quality of life (ICV) for every populated 100 m cell in the metropolitan area. The predictions combine the ENHOGAR 2022 survey with building footprints, night-time lights, vegetation and road networks.
Predicted ICV, quintiles of grid cells
ICV runs from 1 to 5, and higher means better living conditions. The barrio map uses the same class breaks as the grid, so the two views can be compared directly.
How these maps were made
We trained a random forest model on average ICV values from the ENHOGAR 2022 survey, linked to 2010 census segments. The predictors come from geospatial data: Google Open Buildings (count, size and height of buildings, and change from 2016 to 2022), VIIRS night-time lights, NDVI and OpenStreetMap roads. Cells with fewer than 5 buildings are left out.
The model ranks areas better than it predicts exact values. That is why we show the cells in five relative classes rather than as a continuous scale.
0.58Rank correlation (cross-validated)
0.48RMSE on the 1–5 scale (cross-validated)
46,100100 m grid cells
571Barrios & parajes
Client: Government of the Dominican Republic & the World Bank, in support of a climate-smart housing strategy for the Mancomunidad del Gran Santo Domingo.