非渗透表面丰度提取方法运用——以广州市海珠区为例  

An Approach of Extracting the Percent of Impervious Surface and Its Application in Haizhu District of Guangzhou City

作　　者：李明杰[1,2,3] 钱乐祥[1] 陈健飞[1] 

机构地区：[1]广州大学,广州510006 [2]中科院烟台海岸带研究所,烟台264003 [3]中国科学院研究生院,北京10049

出　　处：《遥感信息》2011年第2期36-40,119,共6页Remote Sensing Information

基　　金：建设部科技资助项目(2007-K9-26);广州市属高校科研资助项目(10A004)

摘　　要：非渗透表面丰度即非渗透表面含量百分比,是水文学中不透水层与遥感领域混合像元分解概念相结合的产物。在城市区域,非渗透表面多指房屋、道路、停车场等建设用地区。本文基于广州市海珠区2000年ETM+数据,首次采用单窗算法反演的地表温度数据与植被-非渗透表面-土壤(V-I-S)模型、归一化线性混合光谱模型(NSMA)相结合的方法提取研究区非渗透表面丰度值,并对提取结果进行49个300m×300m有效样区的抽样检验。结果表明抽样值与真实值之间存在10.03%的均方根误差以及2.65%的系统误差,从而进一步验证了三模型结合使用提取非渗透表面丰度的可行性。The concept of impervious surface fraction has been defined as the percentage of impervious surface to the total and is widely used to indicate the surface structure in urban region for environmental studies to such important problems as the land surface runoff,urban heat islands effect,environment and water pollution.Using an ETM＋ image of the Haizhu district,Guangzhou city in 2000 as the data source,we developed an approach to compute the fraction from the ETM image through combination of the Vegetation-Impervious-Soil Model（V-I-S）,the Normalized Mixing Spectral Analysis Model（NSMA） and the Land Surface Temperature（LST） data,in detail.The LST retrieved by mono-window algorithm was used to modify the extracted percent of impervious surface data by the V-I-S model and NSMA model.And then,validating the extracted result by 49 samples with a size of 300m×300m,we found that the Root Mean Square Error（RMSE） and Systematic Error（SE） of the approach were 10.03% and 2.65% respectively,which indicated that the approach could be alternative to improve the accuracy of extracting the impervious surface percent from ETM＋ images.

关 键 词：非渗透表面丰度 单窗算法 V-I-S模型 NSMA模型 广州市海珠区 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]