Urbanization---both the shift in population to urban areas and the physical transformation of land---has accelerated in recent decades in nearly all parts of the world. Urban growth is known to affect local environments through microclimatic changes, pollution, and loss of natural and agricultural lands, and has the potential to influence global circulation, biogeochemistry, and energy balance. To assess these impacts, rates and patterns of urban expansion and their relation to socioeconomic drivers must be determined. A combination of multiple sources of coarse resolution remotely sensed data and a Bayesian data fusion approach result in global maps of urban areas that show reasonable accuracy in terms of the size and location of cities, and are suitable for use in regional-global climate models.; To understand changes in urban form locally, medium-resolution remote sensing data, pattern metrics, and local census data are combined to provide a suite of urban growth indicators that facilitate comparison of trends across cities from different regions and different levels of economic development. Chengdu, a rapidly developing city in Western China, provides an ideal case study to test these methods. Results show that the city has increased in area over 350% from 1978 to 2002, while expanding primarily along transportation routes with fragmentation in periurban areas. Comparison of growth metrics for 25 cities from 13 countries for the period 1990-2000 reveals four city types: low-growth cities characterized by modest rates of infilling-type expansion, high-growth cities with rapid, fragmented development, expansive-growth cities with extensive dispersion at low population densities, and frantic-growth cities such as Chengdu, that exhibit extraordinary rates of growth at high population densities. Although all 25 cities are expanding at the urban-rural boundary, cities outside of the U.S. do not show evidence of the large, dispersed spatial forms characteristic of American urban sprawl. Empirical analysis of the linkages between urban growth and population, household, and labor data shows a limited relationship between urban expansion and population change, a factor conventionally thought to play a role in land conversion. Local economic restructuring to industry and service-oriented activities shows some importance in urban growth and dispersion.
展开▼