Authors: Xue WANG*, , Tung FUNG, The Chinese University of Hong Kong
Topics: Remote Sensing, Urban Geography
Keywords: Urban impervious surface, Spectral mixture analysis (SMA), V-I-S model, Coarse resolution, Multi-sensor
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Lafayette, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded
Although urban remote sensing has focused on mapping, monitoring, and understanding urban phenomena for many years, mapping urban areas at regional and global scales remains one of the most challenging tasks. The reasons are not only that there are few effective methods for urban mapping, but also that the definition of urban is always debatable, which influences the accuracy assessment of urban extraction. Therefore, instead of extracting urban areas directly, this study estimated urban impervious surface fraction at subpixel level for analyzing urban environments.
Spectral mixture analysis (SMA) has been widely used in estimating impervious surface distribution based on high and medium resolution images, but rarely been applied to coarse resolution images due to the difficulty of deriving endmember spectra using traditional endmember selection methods. However, in order to monitor urban phenomena at regional and global scales, different coarse resolution datasets (i.e. MODIS NDVI, MODIS LST and DMSP/OLS NTL data) were used in this study because of their coverage and complementary information they offered. Therefore, to address the problem of conducting SMA on coarse resolution images, the method of this study can be described as follows. Firstly, the vegetation-impervious-soil model (V-I-S model) was employed in the SMA. Then, spectra of the three endmembers (i.e. vegetation, impervious surface and soil) were extrapolated through linear spectral mixing model and least square straight (LSS) line method with known abundance knowledge of sample pixels. Finally, with the derived endmembers, each endmember fraction was calculated, including the impervious surface fraction.