The study is an empirical, on the construction of models of urban sprawl and fragmentation of 10 Latin American cases in a time horizon of 20 years: Asuncion, Bogota, Brasilia, Buenos Aires, Cordoba, La Paz, Lima, Montevideo, Santa Cruz de la Sierra and Santiago de Chile. Through the analysis of Landsat satellite images of high spatial resolution, the main parameters are quantified expansive and fragmented. An explanatory framework is developed by comparing the homologous parameters of urban sprawl patterns of European cities, on which there is a prolific series of similar studies. The range of Latin American cities can cover a variety of contexts and possible patterns of urban sprawl and fragmentation, love of the economic and geographical differences of the countries. These cases will generate a Latin American context of urban sprawl and fragmentation, with common parameters and empirical models.
For the image processing software is used ERDAS IMAGINE 9.1. The images are classified on the basis of land cover and use supervised classification. They will ensure that the categories established for each case are possible to compare and integrate later. This will be done on the basis of photo interpretation expert authors, site visits to establish the complex areas in cabinet and support the network of collaborators in situ.
Once processed the coverage vectors are exported to formats to be integrated into the GIS modeling with the other variables in the analysis, as eg census information. This part is developed with ArcGIS 9.0 software.
The project includes a round of field work with the network of partners to discuss and compare the preliminary results. On this visit to the cities under study are calibrated preliminary results interviews with relevant experts and local stakeholders from the public and private sectors to assess in the light of these views the robustness of the preliminary results. The second objective of the visit is the observation in situ of the major trends and patterns obtained. As an empirical study of cross contact between researchers studied the realities – in a reasonable depth beyond the cabinet, is fundamental.
- Adjusted preliminary results are developed temporal patterns of urban expansion (change detection) and establishes the vectors of expansion. Stages of the methodology:
Acquisition of satellite images of the Latin American case studies, orthorectification, georeferencing and calibration.
Literature review and secondary data collection, land registers, census, plans, financial information, statistics, photos, etc.. Revision of theses and papers. Preparation of databases with secondary information. Collection of European information.
Classification of images. Determination of classification levels expansion. Obtaining maps with temporary cuts urban expansion processes categorized for the different vectors of expansion. Calculation of densities and other ad-hoc parameters.
Construction of models of urban sprawl. Change detection (ERDAS) and construction of urban expansion firms. Calculation of rates of land consumption and per capita annually.
Identification and quantification of the parameters of urban fragmentation. This characterization was made using morphologic definitions (compactness, etc). In general parameters for estimating the behavior of forms: Nump, MPS, MedPS, PSSD, AWMPFD, etc. (Patch Analyst, ArcGIS).
Comparative analysis with the study of European cities, regression models and determination of weights.
Comparison of expansive patterns of fragmentation and their respective regulatory frameworks. Discussion and results.
1) multispectral satellite imagery (or bundle type) high spatial resolution.
2) Census and official data sources for each case of Latin America. In the European case can access the database and they are free
3) European Environment Agency,
4) SCATTER project, European Commission,
5) CORINE data base
6) URBAN ATLAS
7) Database GIS and Remote Sensing Laboratory, TU Dresden,
8) Information obtained during visits to cities and by local coordinators and the network of collaborators.