Analysis of Vegetation Indices in Urban Areas from TM-Landsat and HRV-SPOT Orbital Data

摘要

The spatial organization of the brazilian cities in several areas of the country reflects the impact of the accelerated and disorganized urbanization that has occurred during the last years. This urbanization without an adequate planning has caused a process' of degradation of the natural urban environment. An imbalance between built up areas and green areas has occurred where the presence of the vegetation element has grown less. A quantitative and qualitative survey of urban green areas establishes basic information to elaborate adequate planning in order to improve the quality of the urban environment. Through the transformation of orbital data into numerical models, called Vegetation Indices (VI), it is possible to obtain a qualitative and quantitative indicator of the vegetation cover relative to built up areas. The main objective of this study is to analyse the Normalized Difference Vegetation Index (NDVI) in the urban environment obtained from orbital data. The study area is located in the city of Taubate in Sao Paulo state. The NDVI is calculated from orbital data from TM-Landsat (TM3, TM4, TM5) and HRV-SPOT (XS2, XS3) corresponding to spectral bands in the red, near-infrared and middle-infrared ranges. The data was taken on 8 August, 1988 and 19 July, 1988, respectively. It is calculated by the formulae: VI velence gain x (NIR - R)/(NIR + R) + offset and VI velence gain x (MIR - R)/(MIR + R) + offset. The influence of spectral and spatial characteristics from TM-Landsat and HRV-SPOT data was taken into account in analyzing the performance of classification for the VI calculated from those sensor systems. Ground information and the percent vegetation cover were determined from panchromatic aerial photographs (in scale 1:10 000) and planimetric maps (in scales of 1:25 000 and 1:50 000). The different classes of urban land use were discriminated and classified on the basis of VI. In the results, it was found that the NDVI calculated by TM-Landsat and HRV-SPOT data allowed a distinct classification associating urban land use and vegetal cover to be obtained. It is concluded that the VI is a good estimator to compare green areas with built up areas and it permits a global view of the spatial distribution and density of vegetal cover.