This study focuses on monitoring landscape change of Taipei from 1993 to 2007. The objective is to investigate land use effect and suggest future development. The research processes include applying the technique of remote sensing and GIS, landscape index, and Shannon diversity t-test to acquire the landscape change information of Taipei and its administrative districts; investigating if Taipei’s landscape is obviously affected by land use change within 15 years; deriving common factors from the selected landscape indices and analyzing the similarity among Taipei’s administrative districts via multivariate statistical analysis such as principal component analysis and factor analysis and cluster analysis. The results indicated that Taipei’s landscape had a obvious change after 15 years. The area of build-up increases from 33.65% of 1993 to 41.06% of 2007 and the bare land decreases from 16.59% to 10.54%. Both Shannon diversity index and Shannon Evenness index are decreasing. The result of Shannon diversity t-test also showed that the land use change in Taipei did have an obvious effect for the entire landscape. The application of principal component analysis and factor analysis selected four principal component and four factor of the cluster analysis results indicate, that principal component analysis of cluster analysis in 1993 and 2007, that the administrative districts was clustered into four and two groups; as for the application of factor analysis and cluster analysis, the result pointed out that the administrative districts was clustered into six and two groups no matter in 1993 or 2007 information, but if we investigate the administrative districts within the same group, clearly there was different between 1993 and 2007 information. The above results can be concluded as follows: the application of integrating the technique of remote sensing and GIS, landscape index, and Shannon diversity t-test on Taipei’s landscape change is an effective and feasible approach. Meanwhile, the landscape similarity among different administrative districts and the selection of suitable landscape indices on landscape analysis can be achieved by the use of multivariate statistical analysis. The result obtained from this study can be extended as a reference of future development model for Taipei’s land use.
Keywords：Geographical Information System, Landscape Change, Landscape Index, Principal Component Analysis, Factor Analysis, Cluster Analysis
