Remote sensing and spatial metrics for mapping and modeling of urban structures and growth dynamics.

摘要

In support of decisions towards sustainable and desirable future urban development, this research integrates remote sensing, spatial metrics, and urban modeling to map, monitor, and model spatial urban structures and spatio-temporal growth dynamics. The focus is on investigating remote sensing system capabilities and limitations in urban mapping and the evaluation of spatial metrics in mapping and analysis of urban form and processes towards a framework to support and improve urban modeling, management and planning efforts, and advance spatial urban theory.; From a remote sensing perspective, analyses of regional urban spectral libraries show that urban environments have unique characteristics and show a considerable amount of variability and spectral confusion between urban land cover types. Remote sensing data with higher spatial and spectral resolutions, and three-dimensional information from LIDAR allow for more detailed and accurate mapping of the urban environment. Distinct limitations in spectral land cover mapping are identified for broadband multispectral satellite sensors, like IKONOS, that can be partly resolved with object-oriented image processing techniques. Spatial metrics are a useful tool to describe and map spatial land cover characteristics. They provide sophisticated quantitative measurements of urban form (e.g. size, shape, density) and link urban land cover to urban land use. With multi-temporal remote sensing mapping products, spatial metrics are able to capture multi-faceted characteristics of dynamic urban growth processes on different spatial scales.; Empirical analysis of relating spatial urban structures and growth dynamics to urban development processes and growth factors revealed distinct regularities in urban phenomena. From a theoretical perspective, this approach using remote sensing and spatial metrics helps to link urban form to processes, empirical growth observations to urban theory, and connects patterns and processes among different spatio-temporal scales. With its inductive perspective "from structure to process," rather than "from process to structure" as adopted by most urban studies, this research emphasizes an underrepresented component in the urban research agenda. Urban management and planning could benefit from the resulting better data, knowledge, and information, if current issues of understanding, integration, communication of results, and general acceptance of new and innovative techniques can be resolved.