It is a common practice to disaggregate population totals and to compare the results with
actual small area counts for validation. This is difficult where there are no published census
figures, such as in Nigeria. This study evaluates dasymetric surfaces generated using different
satellite data with differing resolutions and scales of analysis in a location where small area
census data is available, in order to infer the most appropriate methods for areas where it is
not. Initial results show that land cover derived from high-resolution satellite data and disaggregations
over fine scale target grids provides better population dis-aggregations for use as
demand surfaces in siting service facilities. This research addresses problems in population
estimation in locations where validation data is not available.