A strategy for analyzing urban forest using Landsat ETM+ imagery

Abstract

Urban forest is of great interest to a variety of scientific and urban planning applications. This paper presents a strategy for calculating tree canopy density in urban areas using Landsat ETM+ imagery and calculating its ecological value. The strategy consists of two key steps: one is to extract urban tree canopy area from remote sensing images; the other is to calculate the ecological value of urban forest. The extraction of tree canopy area from remote sensing imagery is carried out using classification models, which are based on empirical relationships between forest coverage and the spectrum on Landsat imagery, and are generated using regression tree techniques. The calculation about urban forest is carried out introducing CITYgreen model. In the analysis, the zone inside the 4th Ring Road in Beijing is chosen as study area; and a Landsat ETM+ image taken in 2001 is used applying this strategy to analyze the urban forest. In the following part, the results of the analysis are presented. At the end, the accuracy of tree canopy density is reviewed, and some possible reasons for the error are discussed, as well as the advantages and disadvantages of this strategy.