科尔沁沙地固沙植被高光谱特性及叶片含水率反演研究

发布时间:2018-06-05 03:40

  本文选题:沙丘植被 + 光谱特性 ; 参考:《内蒙古农业大学》2017年硕士论文


【摘要】:地物光谱特征的测量和研究不仅是各种遥感应用分析的基础,同时也为遥感理论研究的重要内容。由于实测地面光谱数据的光谱分辨率较高,因此,它能够准确反映目标地物的光谱信息。本文采用Field Spec4(ASD)便携式光谱仪对科尔沁5种沙丘植被(麻黄、冷蒿、黄柳、差巴嘎蒿和小叶锦鸡儿)进行了高光谱测定,并对其特征进行提取分析,研究了它们不同生长时期和不同覆盖度下差巴嘎蒿的光谱特性,为今后该区域植被特性的反演和光谱数据库的建立提供数据支持,同时对小叶锦鸡儿叶片含水率建立了高光谱反演模型,为进一步探索高分辨率卫星遥感数据在固沙植被叶片含水率反演方面的方法和应用提供依据。主要成果如下:1.5种典型沙丘植被的光谱形态大致相同,但又略有差异。在600-690nm与750-900nm波段处差异较大,根据光谱差异波段和导数光谱参数可以区分冷蒿、麻黄和差巴嘎蒿,但不能区分黄柳和小叶锦鸡儿。2.随植被生长期的推移,5种沙丘植被的"绿峰"位置经历了先向蓝光方向偏移而后又向红光方向偏移的过程,黄柳、冷蒿、麻黄和小叶锦鸡儿4种沙丘植被光谱对应的"红边"均具有明显的"双峰",而差巴嘎蒿光谱"红边"区的"双峰"现象并不十分的明显;麻黄、黄柳和小叶锦鸡儿光谱的"红边"区(同其他植被相同)出现较为明显的"红移"现象,黄柳和冷蒿的"红边"区在9月末出现"蓝移"现象,而冷蒿和差巴嘎蒿的"红边位置"呈微弱减少态势,"红移"现象并未出现。5种典型沙丘植被在8、9月份均具有"红边平台",其中差巴嘎蒿和小叶锦鸡儿在9月27日均持续处在"红边平台"位置。3.5种典型沙丘植被在红光波段处的吸收明显比蓝光波段处强烈,吸收深度也比蓝光波段处深。随植被生长推移,5种沙丘植被的吸收深度、吸收面积和NDVI变化趋势相同。不同生长时期的"红边"、"吸收"参数和时序NDVI数据识别植被的效果明显好于其他参数。研究区进行植被分类时,应选择春季和秋季中晚期的多时相遥感影像,进行联合解译会取得更好的效果。4.分析不同覆盖度下差巴嘎蒿的光谱特性,当覆盖度大于40%时,"红边位置"与覆盖度并无直接联系,但"红边斜率"与覆盖度呈正比。其吸收深度随覆盖度的增加而增大,吸收峰面积随覆盖度的增加而减小,对覆盖度敏感性参数比较可知,研究该地区差巴嘎蒿覆盖度变化优先考虑吸收峰面积。5.对采集的113组小叶锦鸡儿叶片高光谱反射率进行小波变换,并构建基于"小波植被指数"反演小叶锦鸡儿叶片含水率的线性模型,将所得结论与构建的光谱植被指数反演模型进行比较,进而得出结果:基于6个小波植被指数构建反演叶片RMC的多元线性模型RMCw和采用3个光谱植被指数建立的多元线性模型的R2分别为0.72、0.46,检验精度R2分别为 0.88、0.57。
[Abstract]:The measurement and study of the spectral characteristics of ground objects is not only the basis for the analysis of various remote sensing applications, but also an important part of the theoretical study of remote sensing. Because of the high spectral resolution of the measured ground spectral data, it can accurately reflect the spectral information of the target ground object. In this paper, five kinds of sand dune vegetation (Ephedra ephedra, Artemisia chrysanthemi, Salix willow, Artemisia japonica and Caragana microphylla) were determined by Field Spec4ASD portable spectrometer, and their characteristics were extracted and analyzed. The spectral characteristics of Artemisia variegata in different growth periods and different coverage were studied, which provided data support for the inversion of vegetation characteristics and the establishment of spectral database in the region in the future. At the same time, the hyperspectral inversion model of leaf moisture content of Caragana microphylla is established, which provides the basis for further exploring the method and application of high resolution satellite remote sensing data in the retrieval of leaf moisture content of sand fixation vegetation. The main results are as follows: 1. 5 typical dune vegetation has the same spectral form, but slightly different. There is a great difference between 600-690nm and 750-900nm bands. According to the spectral difference band and derivative spectral parameters, Artemisia chrysanthemi, Ephedra ephedra and Artemisia charantia can be distinguished, but they can not be distinguished from Caragana microphylla. With the development of vegetation, the "green peak" position of 5 kinds of dune vegetation experienced the process of first shifting to blue light and then moving to red light, yellow willow, Artemisia frigida, The "red edges" of Ephedra equisetifolia and Caragana microphylla have obvious "double peaks", but the "double peaks" in the "red edge" region of the spectrum of Artemisia ephedra are not very obvious. The "red edge" region of the spectrum of willow and Caragana microphylla (similar to other vegetation) showed a more obvious "red shift" phenomenon, and the "red edge" region of Salix willow and Artemisia frigida showed "blue shift" at the end of September. However, the "red edge position" of Artemisia chrysanthemi and Artemisia japonica decreased slightly, and the "red shift" phenomenon did not show that there were 5 typical dune vegetation with "red edge platform" in August and September, in which Artemisia chrysanthemi and Caragana microphylla had a "red edge platform" on September 27. The absorption of 3.5 typical dune vegetation in red light band was significantly stronger than that in blue light band. The absorption depth is also deeper than the blue band. With the growth of vegetation, the absorption depth, absorption area and NDVI change trend of 5 kinds of dune vegetation are the same. The effect of "absorption" parameter and time series NDVI data in identifying vegetation is better than that of other parameters. In the classification of vegetation in the study area, the multitemporal remote sensing images in the middle and late spring and autumn should be selected, and the combined interpretation will achieve better results. The spectral characteristics of Artemisia variabilis under different coverage were analyzed. When the coverage is greater than 40, the "red edge position" is not directly related to the coverage, but the "red edge slope" is proportional to the coverage. The absorption depth increases with the increase of coverage, and the absorption peak area decreases with the increase of coverage. The hyperspectral reflectance of 113 groups of Caragana microphylla leaves was analyzed by wavelet transform, and a linear model based on "wavelet vegetation index" for retrieving water content of Caragana microphylla leaves was constructed. The conclusions are compared with the constructed spectral vegetation index inversion model. The results are as follows: the multiple linear model (RMCw) based on six wavelet vegetation indices and the multivariate linear model (R2) based on three spectral vegetation indices are 0.72 ~ 0.46 and 0.880.57, respectively.
【学位授予单位】:内蒙古农业大学
【学位级别】:硕士
【学位授予年份】:2017
【分类号】:Q948

【参考文献】

相关期刊论文 前10条

1 郝芳芳;陈艳梅;高吉喜;吕国旭;田美荣;;河北坝上地区草地退化指示种的高光谱特征波段识别[J];生态与农村环境学报;2016年06期

2 童庆禧;张兵;张立福;;中国高光谱遥感的前沿进展[J];遥感学报;2016年05期

3 李真真;郑翔;牛德奎;郭晓敏;谢碧裕;张学玲;;武功山山地草甸主要群落类型高光谱特征[J];草业科学;2016年08期

4 毛新安;夏小伟;安沙舟;;基于高光谱遥感的荒漠草地围栏封育监测[J];草食家畜;2016年03期

5 徐旭;陈国庆;王良;叶桂香;王振林;李勇;;基于敏感光谱波段图像特征的冬小麦LAI和地上部生物量监测[J];农业工程学报;2015年22期

6 刘美;吴世新;潘伯荣;;沙冬青(Ammopiptanthus Cheng f.)高光谱特征提取和分析[J];干旱区研究;2015年06期

7 孙少波;杜华强;李平衡;周国模;徐小军;高国龙;李雪建;;基于小波变换的毛竹叶片净光合速率高光谱遥感反演[J];应用生态学报;2016年01期

8 马维维;巩彩兰;胡勇;魏永林;李龙;刘丰轶;孟鹏;;牧草品质的高光谱遥感监测模型研究[J];光谱学与光谱分析;2015年10期

9 蒋金豹;何汝艳;STEVEND Michael;;基于连续小波变换的地下储存CO_2泄漏高光谱遥感监测[J];煤炭学报;2015年09期

10 刘吉凯;钟仕全;梁文海;;基于多时相Landsat8 OLI影像的作物种植结构提取[J];遥感技术与应用;2015年04期

相关博士学位论文 前2条

1 李向阳;烟草高光谱特性与农艺、生理、品质指标关系和估测模型研究[D];河南农业大学;2007年

2 刘顺喜;高光谱遥感土地利用信息提取技术研究[D];北京林业大学;2005年

相关硕士学位论文 前7条

1 林海军;塔里木河下游主要荒漠植物高光谱特征分析[D];新疆农业大学;2013年

2 苗茜;基于小波变换的芦苇叶绿素含量的地物高光谱反演研究[D];首都师范大学;2013年

3 王志辉;高光谱遥感在森林树种识别中的应用[D];浙江农林大学;2011年

4 赵钊;新疆荒漠植物含水率高光谱特征分析[D];新疆农业大学;2011年

5 郑长春;水稻种植面积遥感信息提取研究[D];新疆农业大学;2008年

6 郝建亭;光谱数据处理及其在植被信息提取中的应用[D];成都理工大学;2008年

7 谢秋昌;基于高光谱数据的分类技术研究[D];长安大学;2008年



本文编号:1980330

资料下载
论文发表

本文链接:https://www.wllwen.com/shoufeilunwen/benkebiyelunwen/1980330.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户ce76a***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com