JOURNAL OF NATURAL RESOURCES ›› 2012, Vol. 27 ›› Issue (9): 1590-1600.doi: 10.11849/zrzyxb.2012.09.016

Previous Articles     Next Articles

Estimating Urban Impervious Surface Based on Thermal Infrared Remote Sensing Data and a Spectral Mixture Analysis Model

LI Bo1,2, HUANG Jing-feng2,3, WU Ci-fang1   

  1. 1. Institute of Southeast Land Management, Zhejiang University, Hangzhou 310058, China;
    2. Key Laboratory of Agricultural Remote Sensing and Information System in Zhejiang Province, Hangzhou 310058, China;
    3. Institute of Agricultural Remote Sensing and Information Technology, Zhejiang University, Hangzhou 310058, China
  • Received:2011-12-20 Revised:2012-04-25 Online:2012-09-20 Published:2012-09-20

Abstract: The rapid growth of impervious surface is one of the remarkable characteristics of urbanization. Rapid extraction of urban impervious surface using remote sensing has been a hotspot research at home and abroad for large-scale urban monitoring. This paper explored extraction of impervious surface information of Hangzhou from a Landsat 7 ETM+ image based on the integration of a spectral mixture analysis model and land surface temperature generated by thermal infrared images. The linear combination of high albedo, low albedo, vegetation and soil fraction was used to characterize the different types of urban land. The land surface temperature was considered as a mask to remove the "noise" from low albedo fraction, and soil fraction was used to remove the "noise" from high albedo fraction. The modified high albedo fraction and low albedo fraction were adopted to estimate impervious surface distribution of Hangzhou. The result showed that the average RMSE was 0.0036 in the study area. Impervious surface distribution estimated using the above method and the interpretation from high resolution images in Google earth and SPOT 4 image was comparatively analyzed, and the majority of differences between estimating values and interpreting values of samples were ranged from -0.15 to +0.15. There was a promising accuracy. The result indicated that the method was feasible and reliable to precisely estimate impervious surface based on thermal infrared remote sensing images and a spectral mixture analysis model.

Key words: remote sensing, impervious surface, spectral mixture analysis, thermal infrared data

CLC Number: 

  • TP79