An evaluation of global urban growth via comparison of DCW and DMSP-OLS satellite data

Whilst urbanised areas account for the smallest land cover class, it is arguably the most important in that knowledge of the size and spatial distribution of our cities is a key parameter on a number of fronts from resource management to economic development planning. An evaluation has been performed comparing satellite assessments of urban land area extent at two different times. The 1:1,000,000 Digital Chart of the World (DCW) dataset originally collected in the 1960s and 1970s mapped many physical, social and cultural features. A populated places layer is included in this. This has been quantitatively compared using a geographic information system (GIS) to the 1994-5 Defense Meteorological Satellite Programme Operational Linescan (DMSP-OLS) city lights dataset at 1 km prepared by Chris Elvidge and others at the National Oceanic and Atmospheric Administration National Geophysical Data Center (NOAA-NGDC). The city-lights dataset is a temporal composite of nocturnal VIS-NIR emissions. Once processed into a “stable lights” product a map of electrified human settlements is derived. The city-lights are observed to envelop the DCW polygons suggesting many of these areas have merged into larger conurbations. By using the International Geosphere-Biosphere Programme (IGBP) global land cover map (derived from 1 km satellite data) it has also been possible to assess which land cover types have been misidentified in the IGBP land cover map. This raises important questions regarding the diminishing extent of agricultural land and its human carrying capacity