基于LST-EVI特征空间的江汉平原腹地渍害风险研究

发布时间：2018-06-22 20:05

  本文选题：渍害 + 温度-植被指数特征空间　； 参考：《华中师范大学》2016年硕士论文

【摘要】：渍害低产农田在中国南方湖区、沿江平原广泛分布。由于渍害对粮食生产影响严重,发生存在历史悠久,一直是渍害低产田治理中的重要内容。传统对渍害的鉴定、识别是在实地调研、实验和测量的基础上进行的,费时费力。因此,迫切需要寻求一种快速有效的监测识别的方法,而遥感技术就为此提供了技术支持。本研究选取江汉平原腹地四个县市作为研究区,利用遥感数据和DEM数据对研究区的渍害农田进行尝试判定提取并进行空间分析。研究结果可以为渍害农田的快速识别、监测提供一种方法,丰富了湿润湖区平原的土壤水分含量的研究,具有重要的理论和现实意义。本研究选取2005年4月20日的Land sat 5 TM遥感影像数据,在融合、裁剪等预处理的基础上,进行辐射定标和大气校正,并对校正前后地物波谱曲线变化进行了分析；对研究区遥感影像进行地物分类并进行分类精度评价；在对地表真实温度反演中,地表比辐射率的估算采用像元分解法。其次,通过LST-NDVI和LST-EVI构建的特征空间对“干边”和“湿边”进行拟合,可以发现构建的LST-EVI特征空间具有明显的三角形特征。因此,采用LST-EVI特征空间的拟合系数计算出研究区的TVDI。最后,在ArcGIS 10.0中对地物分类中的农田像元、DEM数据和TVDI进行空间分析,获取渍害农田像元；分析了渍害的三个主要影响因子：水系、地势起伏度和高程,并对渍害农田进行了空间分布统计。本研究验证了渍害遥感识别原理的科学性、合理性、可行性和实践性,为渍害农田的遥感识别判定提供了一种快速有效的方法。本文的主要成果有：(1)利用LST-Ⅵ特征空间对渍害农田进行判定识别。LST-Ⅵ特征空间表征了土壤含水量、温度和植被之间的关系,因此可以用于渍害研究。本研究提出和验证了该方法的可行性。(2)基于GIS空间分析,提出遥感识别判定渍害农田的阈值标准：同时满足0.12EVI0.31和23.1 mDEM34 m的农田耕地像元；农田耕地像元的TVDI满足0TVDI0.45。(3)研究结果显示研究区2005年渍害农田的面积约为339.69km2,集中分布的高程范围为24 m～29m。
[Abstract]:Waterlogging and low yield farmland are widely distributed in the lake region of southern China and the plain along the Yangtze River. Because of the serious impact of waterlogging on grain production and the existence of a long history, it has been an important content in the treatment of waterlogging and low yield fields. Traditional identification of stains is based on field investigation, experiment and measurement, which is time-consuming and laborious. Therefore, there is an urgent need to find a rapid and effective method of monitoring and identification, and remote sensing technology provides technical support for this. In this study, four counties and cities in the hinterland of Jianghan Plain were selected as the study areas. Remote sensing data and Dem data were used to determine and extract the waterlogged farmland in the study area and spatial analysis was carried out. The results can provide a method for rapid identification and monitoring of waterlogged farmland, and enrich the study of soil moisture content in humid lake plain, which has important theoretical and practical significance. In this study, the data of Land sat 5 TM remote sensing image from April 20, 2005 were selected to carry out radiometric calibration and atmospheric correction on the basis of pre-processing such as fusion and clipping, and the changes of spectral curve of ground object before and after correction were analyzed. The classification accuracy of the remote sensing image is evaluated and the pixel decomposition method is used to estimate the surface specific emissivity in the inversion of the real surface temperature. Secondly, by fitting "dry edge" and "wet edge" with the feature space constructed by LST-NDVI and LST-EVI, we can find that the constructed LST-EVI feature space has obvious triangle characteristics. Therefore, the fitting coefficients of LST-EVI feature space are used to calculate TVDI in the study area. Finally, in ArcGIS 10.0, the field pixel Dem data and TVDI were analyzed in ArcGIS 10.0, and the three main influencing factors of waterlogging were analyzed: water system, topographic fluctuation and elevation. The spatial distribution of waterlogged farmland was analyzed. This study verifies the scientific rationality feasibility and practicality of the principle of remote sensing identification of waterlogging damage and provides a rapid and effective method for the identification of waterlogging damage farmland. The main results of this paper are as follows: (1) the relationship among soil moisture content, temperature and vegetation is characterized by LST- 鈪，

本文编号：2054081