Mapping recent built-up area changes in the city of Harare with high resolution satellite imagery

Worldwide, more people currently live in cities than in rural areas. The appeal of the city remains strong in particular in Africa, where currently the highest growth rates are observed. Population growth of urban areas is relatively well monitored by international organizations, but there is often little geographic information available on where, inside an urban area, this population growth occurs. Built-up area is a good proxy for population. Earth observation (EO) data can help to locate and identify settlement features and to measure the growth of a city over time. Monitoring changes is important for urban planning and assessments relating to the quality of life and human security. This is especially relevant for cities in developing countries where built-up area extension is not always planned or monitored. EO data are particularly relevant for these areas also because they are an independent information source. In the current study we apply a workflow for multi-temporal analysis of high resolution satellite images aiming at characterising changes in built-up area in the city of Harare, Zimbabwe, between 2004 and 2010. The workflow was designed such that it can be applied to other urban areas using other input data. To achieve this, the level of manual intervention is kept low. The workflow combines an automatic feature extraction method with a grid based automatic change analysis. Built-up area was extracted from high resolution Spot images with accuracies between 78% and 84% and changes were computed as percentage change in a grid using an object-based hierarchical method. The derived built-up area and changes were compared to population figures from the last two censuses at sub-district level. Both parameters reflect urban growth that occurs mainly in the urban fringes. New built-up areas appear either as new construction sites or as extension to existing high density areas. Most of those are formal housing constructions. Large extensions in informal settlements were not observed. The study demonstrates the added value of remote sensing data for providing refined geographic information in a cost efficient and quick way. The applied workflow is efficient in providing approximate information on the location of population in areas where such information is missing or only available at very coarse resolution (e.g. administrative units).

ANNETT Wania;  KEMPER Thomas;  TIEDE Dirk;  ZEIL Peter; 

2013-11-21

ELSEVIER SCI LTD

JRC85771

0143-6228,   

http://www.sciencedirect.com/science/article/pii/S0143622813002440,    https://publications.jrc.ec.europa.eu/repository/handle/JRC85771,   

10.1016/j.apgeog.2013.10.005,