关键词: 线性光谱混合模型, 不透水地表, 景观格局, 武汉市

Abstract:

Impervious surface is considered as an indicator of urban ecological environment and impervious surface area data, which is important to urban planning and environmental and resources management. The reconciliation between the V-I-S model and LSMA provided a continuum field model, which offered an alternative, effective approach for characterizing and quantifying the spatial and temporal changes of impervious surface. In this article, we extracted the impervious surface information from Landsat images of 2002, 2009 and 2015 within the metropolitan area of Wuhan by a fully constrained linear spectral mixture model based on the vegetation-impervious surface-soil (V-I-S) model. Here, gradient analysis was adopted to analyze spatial distribution and four different landscape indicators were chosen to analysis landscape patterns dynamics of impervious surface from 2002 to 2015. Rusults of this study were as follows: The average impervious surface coverage of Wuhan was respectively 27.53% in 2002, 34.65% in 2009 and 40.51% in 2015, which showed a trend of rising. The areas of high impervious surface coverage value of Wuhan are mainly distributed along the Yangtze River and Han River as well as in some secondary centers such as Jiangxia and Hannan and that of low value are mainly distributed in suburban counties. For the period 2002 to 2009, areas of new impervious surfaces mainly formed around existing urban areas and mostly concentrated in circumjacent areas of Wuhan Economy and Technology Development Zone and Donghu New Technology Development Zone as well. After 2009, it was observed that the distribution of new impervious surfaces was scattered. The impervious surface coverage of main urban area is obviously higher than new urban district, but the impervious surface coverage of new urban district increase rapidly. With the increase of the distance to downtown, impervious surface coverage tended to be stable after progressive decrease. The range of 10km outside the third-ring road and 4 km within it was the area of largest increment of average impervious surface coverage. Landscape pattern analysis results showed that natural surface and area of very high density impervious surface had a low degree of fragmentation, strong spatial continuity and a very simple shape. Conversely, low density and medium density impervious surface area had a high degree of fragmentation, weakest spatial continuity and a very complex shape. The patches shape shows that natural surface had a high dominant position in 2002, but after that changed to high density impervious surface after 2015.

Key words: linear spectral mixture model, impervious surface, landscape pattern, Wuhan City