基于ETM+的植被覆盖信息提取及其动态变化分析——以广州市为例

以2000年和2005年ETM+影像为数据源,在水体掩膜和端元光谱优选对比基础上,利用线性光谱模型提取了广州市植被覆盖信息,并对植被覆盖信息的时空动态变化进行了定量分析,结果表明:应用线性光谱模型之前对所占面积不大的水体进行处理是非常必要的,选取植被、硬化地面、土壤三端元法比包含阴影的四端元法更有利于线性光谱分离,而且能达到很高的精度;此外,通过对植被覆盖信息空间结构和景观格局的研究,显示:2000-2005虽然广州市绿化取得了初步成效,但总体植被覆盖状况呈下滑趋势,植被景观破碎程度增加,各植被景观形状趋于复杂化,城市化的发展加之人类活动的干扰,低植被覆盖区不断扩大,而中、高植被覆盖区正处于不断退化时期。

Estimation and spatio-temporal characteristics analyses of vegetation abundances based on ETM+ images-A case study in Guangzhou

The ETM+ images acquired on November 1,2000 and November 23,2005 were chosen and the vegetation abundances in Guangzhou were estimated using linear spectral mixing model–LSMM by applying series processes including masking water,choosing better endmembers and so on.Based on this,the dynamic characteristics of vegetation abundances from 2000 to 2005 in study area were also analyzed.Research indicates that,on the one hand,the process of masking water is necessary before applying LSMM and the three-endmember is better than four-endmember which including shadow.On the other hand,the spatio-temporal analyses of vegetation abundances shows that although the improvement of urban environment has made progress in the ecological environment,as a whole,it is being deteriorated in Guangzhou.