TM影像VI提取植被信息技术研究

应用规一化植被指数（ＮＤＶＩ）、土壤调节植被指数（ＳＡＶＩ）和修正后的土壤调节植被指数（ＭＳＡＶＩ）,研究了ＴＭ影像植被信息的提取技术。结果表明,ＳＡＶＩ较之ＮＤＶＩ和ＭＳＡＶＩ,消除土壤背景影响和适应植被密度变化的能力更强,是植被信息提取效果较好的植被指数。使用ＳＡＶＩ,关键是要确定一个合适的土壤调节参数Ｌ,以最大限度地消除土壤背景的影响,ＳＡＶＩ提出时推荐的最佳Ｌ值（０．５）应用于卫片处理有很大局限性;基于研究结果,本文给出了适于不同植被条件卫片处理的最佳Ｌ取值。     

基于遥感分层技术的平南县森林蓄积抽样调查方法研究

采用遥感数据辅助分层可解决分层抽样在大范围森林资源调查中分层面积不准确的缺点.以ALOS数据为基础,将平南县的森林资源分为A层（有林地、疏林地层）和B层（其它地类层）.在各层内机械预布设样地,比较预布样地缓冲区（角规控制检尺所能绕测到的最大范围）的SAVI值、DNN IR值及对明显地物的目视解译,确定各样地缓冲区的地类,A层样地数有578个,B层有978个.根据分层抽样各层所需样本数,在确定好地类的样地中,随机抽取各层所需样本数并调查其蓄积量.结果表明,抽样的估计精度为91.5%,全县森林蓄积量为5 900 186.7 m3.     

Land cover change detection based on satellite data for an arid area to the south of Aksu in Taklimakan desert

An experiment is made to detect the land-cover change in the area located to the south of Aksu in the northern Taklimakan desert through analyses of satellite data pixel by pixel basis. The analyzed data are those observed in the late summer and early autumn of 1973, 1977, 1993 and 1995. As a parameter of land-cover, SAVI (Soil Adjusted Vegetation Index) derived from the data of Landsat MSS and JERS-1 OPS (Optical Sensor) is used. The result indicates the increase of vegetation in the oasis areas, confluent area of the Yarkant and Kashgar Rivers and around reservoirs while little change occurs in the desert area. The 1973 satellite image shows the abundant flow in the Yarkant River while the river is almost dried up in the satellite images of later years. The trend of the decrease in the Hotan River flow is recognized although not so dramatic as that of the Yarkant River.     

Estimation of dwarf green bean water use under semi-arid climate conditions through ground-based remote sensing techniques

Determination of temporal and spatial distribution of water use (WU) within agricultural land is critical for irrigation management and could be achieved by remotely sensed data. The aim of this study was to estimate WU of dwarf green beans under excessive and limited irrigation water application conditions through indicators based on remotely sensed data. For this purpose, field experiments were conducted comprising of six different irrigation water levels. Soil water content, climatic parameters, canopy temperature and spectral reflectance were all monitored. Reference evapotranspiration (ET₀), crop coefficient K_c and potential crop evapotraspiration (ET_c) were calculated by means of methods described in FAO-56. In addition, WU values were determined by using soil water balance residual and various indexes were calculated. Water use fraction (WUF), which represents both excessive and limited irrigation applications, was defined through WU, ET₀ and K_c. Based on the relationships between WUF and remotely sensed indexes, WU of each irrigation treatments were then estimated. According to comparisons between estimated and measured WU, in general crop water stress index (CWSI) can be offered for monitoring of irrigated land. At the same time, under water stress, correlation between measured WU and estimated WU based on CWSI was the highest too. However, canopy-air temperature difference (T_c − T_a) is more reliable than others for excessive water use conditions. Where there is no data related to canopy temperature, some of spectral vegetation indexes could be preferable in the estimation of WU.     

基于RS、GIS 的呈贡县总体活立木蓄积量分层抽样控制方法

在传统的森林资源调查中,总体活立木蓄积量抽样控制法所需的样地数多,需耗费大量的人力物力.利用SPOT 5卫星数据,基于RS和GIS将呈贡县森林资源先分为A层(有疏林地层)和B层(其它地类),在各层内机械抽样初步确定样本数为1 816个,再根据样地SAVI值和DNNIR值确定各层最终样本数为424个.可极大地减少外业工作量.     

科尔沁沙地植被覆盖变化及其与气候因子的关系研究

干旱半干旱地区生态系统脆弱,植被覆盖变异性大,明晰荒漠化地区植被覆盖变化,进一步确定其时空变异性和气候因子之间的关系,可为荒漠化防治提供理论依据。以我国北部荒漠化典型地区科尔沁沙地为研究区,利用中分辨率成像光谱辐射计(Moderate Resolution Imaging Spectroradiometer,MODIS)地表反射率产品数据,提取土壤调节植被指数(SAVI),结合信息熵理论,分析了研究区SAVI时空变异性,进一步通过相关分析明确研究区植被覆盖变化和气候因子的关系。整体上,研究区植被生长状况东部好于西部,南部好于北部。植被覆盖时空变异性在时间上以2000年最大,2005年最小;空间上以研究区西部南缘最大,东部南缘最小。降雨量、气温均与SAVI呈正相关,其中降雨对于SAVI的影响较大。     

基于EM38和WorldView-2影像的土壤盐渍化建模研究

在干旱半干旱地区,土壤盐渍化是常见的土地退化问题之一。本研究选取于田县克里雅河上游边缘典型盐渍化区域作为研究靶区,通过EM38大地电导率仪实测土壤表观电导率,提取不同系数下的土壤调节植被指数(SAVI),分析了SAVI指数与土壤电导率间的相关性,并利用同时期WorldView-2影像的敏感波段建立了基于高分辨率影像数据的土壤盐渍化偏最小二乘回归(PLSR)模型并进行了精度验证。结果表明:①从遥感影像提取SAVI指数时,在系数(L)调节范围内选取固定系数值,系数值(间隔为0.1)从0.1变化到1.0的过程中,相应提取的SAVI指数与土壤电导率的相关性明显提升,相关性系数(r)从0.30提高到0.50,并通过显著性检验(P0.01)。②选取的SAVI1.0、B6、B7、B8四种变量中,以SAVI1.0+B6+B8为变量组合所建立的PLSR模型为最优,该模型较其他变量组合建模的决定系数(R2p)提高了0.11,因此,在研究区该模型具有更好的预测能力,模型精度为RMSEC=0.77dS/m、RC2=0.68、RMSEP=0.79 dS/m、RP2=0.66、RPD=2.2。     

基于温度植被旱情指数的徐州市郊干旱遥感监测

利用Landsat TM/ETM+数据,以徐州市郊为研究区,获取归一化植被指数(NDVI)、土壤调整植被指数(SAVI)和地表温度(Ts)信息,分别构建NDVI-Ts和SAVI-Ts特征空间,依据这两个特征空间计算出研究区2001年4月3日和2007年5月14日的温度植被旱情指数TVDI(NDVI)和TVDI(SAVI),并分别与地表温度(Ts)和降水量进行了相关评价.结果表明,TVDI可用于实现大范围的干旱监测,SAVI能够修正NDVI对土壤背景的敏感,基于SAVI的反演结果明显优于基于NDVI的反演结果,能够有效地运用于干旱监测.     

Relationships Between Landscape Greenness and Condition of Elk,Mule Deer,and Pronghorn in New Mexico

Nutritional condition drives large herbivore population performance and is related to precipitation and forage quality in the arid Southwest. Because precipitation is difficult to measure at home-range scales, we tested whether satellite-derived vegetation indices of landscape greenness (i.e., indices of vegetation phenology or photosynthetic activity including normalized difference vegetation index [NDVI], soil-adjusted vegetation index [SAVI], and enhanced vegetation index [EVI]) were correlated to the condition of three species of large herbivores (elk, mule deer, pronghorn). We used canonical correlation analysis to relate seasonal landscape greenness with several measures of large herbivore condition. We also used linear mixed models to relate each measure of condition to seasonal landscape greenness separately for each herbivore population-year to see if any patterns were masked by multivariate analysis. Landscape greenness indices were only weakly related to condition of large herbivores, and the effect of landscape greenness on condition was always weaker than lactation status with the exception of pronghorn, an income breeder. Different indices also frequently gave highly variable and conflicting relationships between seasonal landscape greenness and condition of large herbivores. Overall, expected positive relationships between herbivore condition and landscape greenness indices were seen in only 8% of 2 988 possible outcomes. Because indices of landscape greenness are increasingly being used to relate wildlife population demographics to precipitation through a presumed effect on forage quality and resultant nutritional condition, we caution this use in arid environments unless a direct landscape greenness-forage quality or greenness-condition link is demonstrated.