Geo-scale integration: identifying the urban continuum with people, lights and land cover

• Grant number: 13/22038-0
• Support type: Regular Research Grants
• Duration: March 01, 2014 - February 29, 2016
• Field of knowledge: Humanities - Geography - Human Geography
• Principal researcher: Reinaldo Paul Pérez Machado
• Grantee: Reinaldo Paul Pérez Machado
• Home Institution: Faculdade de Filosofia, Letras e Ciências Humanas (FFLCH). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Assoc. researchers: Ailton Luchiari ; Christopher Sennen Small ; Ligia Vizeu Barrozo

Abstract

Urban areas have long been mapped in terms of both population and land use but changing administrative boundaries and land use definitions make it difficult to quantify change accurately over large areas and time scales. Satellite imagery provides a more consistent basis for mapping land cover, as well as a 30 year archive of imagery, but identification of urban land cover remains challenging. Smaller towns, diffuse development and varying degrees of spatial connectivity combined with the lack of an agreed upon definition of urban complicate the task.The most commonly available imagery over the longest time period is provided by the Landsat missions. However, the 30 m spatial resolution of Landsat imagery combined with the spectral heterogeneity of urban land cover results in most urban areas being imaged as spectrally mixed pixels. Spectral mixture models may provide a physically based solution to the urban spectral heterogeneity because they allow land cover to be represented as areal fractions of specific components, thus making interpretation easier. Some of the spectral ambiguity of daytime imagery can be reduced by the use of ancillary data sources. Imagery provided by the Defense Meteorological Satellite Program Operational Line Scanner (DMSP-OLS) indicate the presence of development by imaging night lights. This sensor has been used to produce annual global composites of temporally stable nighttime lights since 1992 (Elvidge et al., 2001).We conduct change analysis for the Sao Paulo metropolitan area using changes in Landsat-derived land cover fractions in conjunction with changes in night lights observed by DMSP-OLS. OLS imagery contributes to differentiate urban and non-urban substrate according to the intensity of the city lights.To determine the urban continuum, vector census tracts (IBGE, 2011) were overlaid on the Landsat and DMSP-OLS data. The conurbated surface limits were traced following the criteria of 300 m of separation among urbanized spots to be considered within the urbanized surface.Keywords: Urban Area Identification, Spectral Mixture Analysis, Night Lights, Census tracts. (AU)