Fine-grained Population Mapping from Coarse Census Counts and Open Geodata
This addresses the need for accurate population mapping in domains like urban planning and public health, especially in regions with limited or outdated census data, representing a strong specific gain.
The paper tackles the problem of generating fine-grained population maps from coarse census data or no census data at all, achieving R2 values of 85-89% for disaggregation and 48-69% for unconstrained prediction in sub-Saharan Africa.
Fine-grained population maps are needed in several domains, like urban planning, environmental monitoring, public health, and humanitarian operations. Unfortunately, in many countries only aggregate census counts over large spatial units are collected, moreover, these are not always up-to-date. We present POMELO, a deep learning model that employs coarse census counts and open geodata to estimate fine-grained population maps with 100m ground sampling distance. Moreover, the model can also estimate population numbers when no census counts at all are available, by generalizing across countries. In a series of experiments for several countries in sub-Saharan Africa, the maps produced with POMELOare in good agreement with the most detailed available reference counts: disaggregation of coarse census counts reaches R2 values of 85-89%; unconstrained prediction in the absence of any counts reaches 48-69%.