Abstract: Using ETM+ and Landsat8 data, three south districts in Wuhan city were chosen as the research area. The impervious surface changes on the space and time was estimated using the decomposition model of mixed pixels, and the related effect on water environment was analyzed. Through the MNF transform and PPI index calculation, four kinds of spectral endmembers including high and low albedo ground, vegetation and soil were identified. The new band of Landsat8: the Cirrus was used to remove the “noise” such as cloud, soil, sand, and MNDWI index was used to remove water area. Through the addition of high and low albedo components, the distribution of impervious surface could be gotten. The result showed that in the past 14 years, the coverage of impervious surface appeared as radial expansion with development of urbanization due to filling in the lake and reclamation. Area increased 52.6 km2; no matter from the perspective of short-term or long-term, impervious surface deeply affected the hydrological environment of urban areas, including surface runoff mode, water cycle process, and the local microclimate; in addition, impervious surface also had a great influence on non-point source pollution in urban areas.