MONITORING DIACHRONIC CHANGES OF THE CITY OF SAO PAULO WITH MULTI-TEMPORAL SPECTRAL MIXTURE ANALYSIS. Reinaldo Paul Pérez Machado Departamento de Geografia, Universidade de São Paulo, Av. Prof. Lineu Prestes, 338 CEP 0558-000. São Paulo – SP. Brasil rpmgis@usp.br Christopher Small 304b Oceanography Lamont Doherty Earth Observatory of Columbia University, Palisades NY 10964. USA small@ldeo.columbia.edu ABSTRACT Spatio-temporal change mapping with remotely sensed imagery has the potential to inform a wide variety of urban research and management questions. Despite its great potential, moderate resolution (10-30 m) remotely sensed imagery is underutilized for urban applications. Traditionally, urban applications of remote sensing use high spatial resolution imagery, while moderate spatial resolution is widely use on non-urban applications. But Landsat’s 30 m resolution has proved sufficient for measurement of some environmental urban parameters that would be both very difficult and expensive to measure directly. The spatiotemporal distribution of the increase of urban built-up areas coincident with the apparent diminishing of the vegetation is an important relationship of the urban fabric that can be quantified using multispectral imagery. However, conventional classification methods used at scales compatible with the sensor spatial resolution have proven to be of limited utility in urban areas, due to the spectral heterogeneity of this complex environment, which commonly combines several different materials. Spectral mixture models, on the other hand, may provide a physically based solution to the urban spectral heterogeneity issue, especially because it is possible to reduce the dimensionality of the multispectral reflectance by converting it to areal fractions of land cover components, thus facilitating the interpretation. The purpose of this study is to analyze the most conspicuous changes occurred on the city of São Paulo using multi-temporal spectral mixture analysis of Landsat TM and ETM+ imagery. The analysis was based on a three component linear mixture model incorporating substrate, vegetation and dark targets, directly used for visualization on false color composites of red, green & blue respectively. Winter and summer image pairs, were selected for quality and consistency of solar illumination for two time intervals: 1986 – 2005 and 2000 – 2010. Fraction difference images reveal the location and extent of urban development – both at the periphery and within São Paulo. Changes in fraction mixtures and texture also highlight high rise construction as well as vegetation abundance. A quantitative and visual analysis of these changes at different spatial scales is presented. Keywords: São Paulo’s diachronic changes; Multi-temporal Spectral Mixture Analysis, Fraction difference images.