以成都市为研究区，定量分析了各地表特征参数与地表温度之间的线性关系。通过对地表温度与归一化植被指数（ NDVI）、归一化建筑指数（ NDBI）、归一化水汽指数（ NDMI）进行局部区域逐像元分析和总体区域统计分析，结果表明NDVI，NDMI，NDBI与地表温度间都存在明显的线性关系，可用于说明地表温度的动态变化，在3月份，NDMI与地温的相关性更优于NDVI。对传统城市热现象研究中，NDMI与ND BI能够用来以NDVI作为分析地表温度随季节而变化的互补的度量标准。

This paper quantitative analysis the linear relationship between the surface characteristic parameters and the land surface temperature(LST) in Chengdu Area .Through the statistical analysis module in ArcGIS , we can make a statistical analysis between land surface temperature (LST) and the normalized difference built -up index(NDBI), the normalized difference vegetation index ( NDVI) , the Normalized Difference Moisture Index ( NDMI) .The analysis indicates that there are strong linear relationships among LST, NDVI, NDMI and NDBI, whereas the relationship between LST and NDMI is stronger than NDVI .This result suggests NDMI and NDBI can be used as complementary metrics to the traditionally applied NDVI for analyzing LST quantitatively over the seasons for surface urban heat island(SUHI) studies.

本文以武汉三镇为研究对象，利用2009年夏季Lalldsat TM主要遥感数据源，进行地表温度反演和土地覆盖解译，计算了武汉市城市建筑指数NDBI、归一化植被指数NDVI；并利用NDⅥ计算地表比辐射率，并使用单窗算法精确反演得到了武汉市2009年夏季地表温度LST。研究结果表明：武汉城市热岛由武昌、汉口、汉阳三大城区热岛构成，且三镇交界区存在多数热岛效应更为严重的“岛中岛”，以汉口区中心商业区最为集中；进一步对植被覆盖区和建筑区温度及NDVI、ＮＤBI相关性研究表明，武汉市不同土地覆盖类型有较大差异，LST与NDBI指数成正相关性，且相关系数建筑用地〉植被〉水体；LST与陆地表面NDⅥ指数成负相关性，且相关系数城市用地〉植被，与水体NDⅥ指数成正相关性。

In this paper, three towns of Wuhan are used as the research object, and Landsat TM in 09/2009 is use d as the main remote sensing data sources. Based on the land surface temperature interpretation after the processin g of image correction, the urban construction index NDBI and normalized difference vegetation index NDVI were calculated in the study area, meanwhile, the surface emissivity was counted by NDVI index and the single window algorithm was used to calculate the surface temperature in the summer of 2009 LST in the study region. The resea rch results show that the urban heat island in the summer of Wuhan consisted of three heat islands in Wuchang,Ha nkou and Hanyang city area, and there existed some more serious "island to island" of heat effect in the border ar ea of three towns, while it was more focus on the commercial center of Hankou. Further research on the relations hip of LST, NDVI and NDBI in the vegetation area and construction area showed that, there was a bigger dif

以3S技术为支持,采用TM数据、应用单窗算法对1989-2010年间北京市五环内地表温度( land surface temperature,LST)进行反演,并采用同步的MODIS 温度产品对反演结果进行对比验证。在此基础上,利用距平值分析和剖面线分析等方法揭示了20多a来北京市地表温度的时空分异总体特征；基于归一化水汽指数( normalized difference moisture index,NDMI)和归一化建筑指数( normalized difference building index,NDBI)分析了研究区内地表温度的时空变化特征,定量揭示并阐述了地表温度和NDMI及NDBI的相关关系。研究结果表明：反演温度可较为真实地反映研究区地表热量的空间差异；北京地表温度时空变化特征明显,同一时段、不同土地覆盖类型的地表温度差异较为显著,热岛效应明显；地表温度和NDBI呈显著的正相关,与NDMI则呈显著的负相关,NDMI和NDBI在表征地表热特征方面均是较好的指标。该结果为监测北京地表温度变化及其时空分布特征提供了依据。

Land surface temperature ( LST) is an important parameter in such fields as meteorology, hydrology and ecology, and the inversion of land surface temperature by use of remote sensing data is a simple and effective method. In this paper, with the support of the 3S technology, the authors took Beijing City as the study area to study the land surface temperature inversion and analyze spatial and temporal characteristics. Using TM data obtained from 1989 to 2010 within the 5th Ring Rroad of Beijing, the authors adapted the single-window algorithm in 20a to the inversion of land surface temperature, with a comparative study of the verification of synchronous MODIS temperature products. On such a basis, the anomaly analysis and hatches analysis were conducted to reveal the overall characteristics of temporal and spatial variation of Beijing surface temperature in more than 20 a. The characteristics of temporal and spatial variation of NDMI and NDBI index were analyzed based on 20 y

首先,阐述了决策树分类器的结构与理论基础;然后,以安徽省滁州市为例,利用Landsat ETM+遥感影像数据和DEM数据,在ENVI软件的支持下,结合影像的光谱特征及NDVI,NDBI特征值,参考C4.5算法(决策树生成算法),建立了土地利用分类的决策树模型并进行了分类;最后,对分类结果进行了精度评价。研究区结果表明决策树分类法效果较好,精度较高。

This paper first describes the structure and theoretical basis of the decision tree classifier .Taking Chuzhou city , Anhui province as a study case and with the support of ENVI software , Landsat ETM+image and DEM are used to analyze the spectral char-acteristics of the image and its value of NDVI , NDBI.Then by referencing C4.5 algorithm (the generation algorithm decision tree ), a decision tree model of land -use classification was built and the accuracy of results are evaluated .The classification results of the study area show that the method is more effective and accurate .

随着社会的快速发展,人类活动的日益频繁,地球表面发生了翻天覆地的变化。遥感技术以其独特的优势成为土地利用调查和变化研究的重要技术手段。本文以淮安市1995年和2006年的TM遥感影像为主要数据,依据土地利用类型的光谱特征,采用NDVI,NDWI,NDBI和非监督分类方法分别提取耕地、林地、草地、水体和建筑用地类型,并对提取结果从结构和变化角度进行分析。结果证明:1)依据地物光谱特征,采用不同模型分别提取土地利用类型的方法切实可行;2)淮安市是一个以农用地为主、水资源丰富、城镇化水平较低的地区;在1995~2006年间,耕地、水域和草地分布呈减少的趋势,林地和建筑用地呈不断增加的趋势。

Society rapidly developing and human constantly acting , the surface of earth has been taking place great changes .Remote sensing on unique advantage becomes the important technical means of land -use survey and change .TM remote sensed images of Huai′an in 1995 and 2006 as mainly data, spectral characteristic based on , cultivated, forest, meadow, water and build -up land were respectively extracted by the methods of NDVI , NDWI, NDBI and unsupervised classification .Moreover, the structure and change of land-use was analyzed based on the extracted consequence .The results show that the method is feasible which the types of land-use are extracted with different models on spectral characteristic .And Huai’an is a region with mainly agriculture constitution , abundant water resources and lower urbanization .In the study period of 1995 to 2006 , cultivated land , water and meadow appear the decrease trend and forest land and build -up land appear the increase trend .