According to the Department of Economic and Social affairs of the
United Nations, the world population is likely to grow by 2.4 billion between
2013 and 2050. While the population in the developed countries will
remain largely unchanged, the population in the less developed countries
rises from 5.9 billion to 8.3 billion. In the 49 least developed countries the
fastest population growth is recorded.
Besides the high population pressure, the least developed countries are the
most vulnerable ones to natural hazards. Only 11% of the population live
exposed to hazards, but 53% of all victims of natural hazards are documented
in these countries. The data, required for disaster risk reduction, is
often of poor quality or lacking.
In case of a natural disaster a proper post disaster management is depending
on the knowledge about the quantity and distribution of population. The
information on population distribution provided by statistical agencies, is
commonly aggregated to administrative units. However, this level of detail
is mostly not sufficient enough in case of a disaster. To provide a more
detailed population distribution, several models, disaggregating population
counts on settlement areas, were developed. The resolution of the existing
models Gridded Population of the World, Global Rural Urban Mapping
Project, LandScan, and WorldPop is mostly to coarse to facilitate a precise
statement for affected population during a disaster. Therefore, a new disaggregation
model with a high spatial resolution and worldwide applicability
is necessary.
In this research work a population distribution algorithm was developed,
based on Census data and Urban Footprint data. The high resolution of
the Urban Footprint facilitates a precise population disaggregation within a
pixel size of 12 m. The algorithm was developed in Bavaria, Germany and
later transferred to Namibia. Exemplary, a population distribution model
of the Cuvelai-Etohsa, a region in Northern Namibia, is generated and combined
with a floodmask of the flood of 2009 to show the applicability of the
new model for flood loss estimation.