Study On Extraction Methods Of Impervious Surface Information Extraction From Urban Area Using Remote Sensing

Since last century, the rapid development of China’s industrialization process made theurbanization level in China increased continuously. This resulted in the land types changefrom those with good permeability around the farmland and village into urban lands with poorpermeability. The change became the main factor affecting the urban ecological landscape.This paper took the provincial capital Taiyuan city in Shanxi Province as an example, usingthe TM and ASTER images as data sources, to study the methods in extracting impervioussurface information from an urban area in the following three aspects using remote sensingand GIS technologies. The three aspects include:1) impervious surface information extractingmethod,2)urban land surface temperature inversion, and3)the relationship between surfacetemperature and impervious surfaces.This paper analyzes three main types of impervious surface information extractingmethods: Principal Component Analysis (PCA), Normalized Difference Impervious SurfaceIndex (NDISI), and Normalized Difference Built-up Index (NDBI). Combining the ModifiedNormalized Difference Water Index (MNDWI) and Normalized Difference Vegetation Index(NDVI), we revised the NDBI index to a Built Up Area Index (BUAI) using logical operators.Moreover, by studying these methods and Optimum Index Factor (OIF), we proposed animproved methods, Experimental Band Combination method (EBC). Both unsupervised andsupervised classification methods were used to classify the original multi-band image, PCAimage, and EBC image into Impervious surface, Vegetation, and Water three broad categories.Lastly, Accuracy Assessments were conducted on all five IS maps.After inverting land surface temperature, the results of temperature inversion andextracted impervious surfaces werefitted by linear regression line. The results showed that the impervious surface numerically higher areas and higher temperature distribution region arerelatively consistent. Impervious surface is strong positively correlated with the land surfacetemperature. The experimental results show that increased impervious surface coverage inTaiyuan significantly accelerated the urban thermal environment change.