基于探地雷达的土壤水遥感监测地面验证及尺度问题研究

发布时间：2018-03-22 00:26

本文选题：探地雷达　切入点：土壤含水量　出处：《中国水利水电科学研究院》2017年硕士论文　论文类型：学位论文

【摘要】：准确定量地监测土壤含水量,对研究区域水循环运动、监测干旱的发生及发展过程、指导农业生产实践和合理进行水资源调控具有重要意义。随着遥感技术的快速发展,快速获取大尺度土壤含水量的遥感方法得到广泛关注。但由于地面仪器测量数据与卫星接收数据的尺度存在很大差异,通过传统的点土壤含水量测量数据验证遥感数据产品的准确性和精确性存在尺度不匹配问题。土壤含水量的定点测量和遥感观测之间缺少对像元尺度土壤含水量验证的科学方法研究,如何有效地通过方便快速、准确可靠的土壤含水量测量方法验证像元尺度的土壤含水量遥感反演结果,是本文的主要科学问题。为研究基于探地雷达的土壤含水量遥感监测地面验证方法,以探地雷达测量土壤含水量的精度验证和影响因素分析为基础,通过基于探地雷达的像元土壤含水量地面验证实验,利用地统计法分析像元尺度土壤含水量的空间异质性及其尺度效应。利用随机组合法建立科学合理的像元尺度土壤含水量验证方法,并应用探地雷达的土壤含水量遥感监测地面验证方法评估30m和1km分辨率土壤含水量遥感反演结果的精度。论文的主要成果如下:(1)利用烘干法验证了探地雷达测量土壤含水量的可靠性:探地雷达的土壤含水量测量结果与烘干法一致性高,适用于田间尺度的土壤含水量测量,其中Topp公式比Roth公式更适合计算荒漠草原区域的土壤含水量。(2)利用地统计法分析了像元尺度内土壤含水量的空间变异性及其尺度效应:样方大小对土壤含水量空间变异性的影响显著,土壤含水量的最大相关距离随样方尺度增大而增大。对于相同尺度样方,由于样方内受地形、微地貌、土壤和植被等复杂因素影响,样方间的空间变异性存在差异,变异系数在12%-19%范围内变化。(3)提出了基于探地雷达的土壤含水量遥感监测结果的地面验证方法:通过随机组合法和泛克里金法分析发现,4条井字形测线即可基本得到稳定可靠的像元土壤含水量观测值,比均匀分布的25个传统测点观测值更准确,而且能够准确地反映高分辨率的土壤含水量空间变化特征。(4)应用探地雷达法验证了不同尺度的土壤含水量遥感监测结果:根据探地雷达的土壤含水量遥感监测地面验证方法,由4条井字形测线实际评估了 30m分辨率的土壤含水量遥感反演结果的精度,得到土壤含水量反演结果的相对误差约为5.9%。对于1km分辨率的土壤含水量反演结果的验证,通过对现有小样方(60m×60m以内)上的观测数据的比较分析,60m×60m样方内的测量数据难以反映1km像元的土壤含水量情况。为准确验证1km像元土壤含水量遥感监测结果,建议在1km像元内布设4条井字形的测线,将其平均值作为该像元的土壤含水量。
[Abstract]:Accurate and quantitative monitoring of soil water content is of great significance for studying regional water circulation movement, monitoring the occurrence and development of drought, guiding agricultural production practice and rationally regulating water resources, with the rapid development of remote sensing technology. The remote sensing method to obtain large scale soil moisture content quickly has been paid much attention to, but because of the difference between the scale of ground instrument measurement data and satellite data, The accuracy and accuracy of remote sensing data are verified by the traditional point soil moisture measurement data. There is a problem of scale mismatch between the fixed point measurement of soil moisture content and remote sensing observation. The lack of pixel scale soil moisture content between soil moisture content measurement and remote sensing observation. Validation of scientific methods, How to effectively verify the results of remote sensing retrieval of soil water content on pixel scale by convenient, rapid, accurate and reliable soil moisture measurement method, In order to study the ground verification method of remote sensing monitoring soil moisture content based on ground penetrating radar, based on precision verification of soil moisture content measured by ground penetrating radar and analysis of influencing factors, Based on ground penetrating radar (GPR), the ground verification experiment of pixel soil moisture content is carried out. The spatial heterogeneity of soil water content in pixel scale and its scale effect are analyzed by means of geostatistics, and a scientific and reasonable method for verifying soil water content in pixel scale is established by using random combination method. The accuracy of 30 m and 1km resolution soil moisture remote sensing inversion results is evaluated by using the ground verification method of remote sensing monitoring of soil moisture content of ground penetrating radar. The main results of this paper are as follows: 1) the ground-penetrating radar measurement is verified by means of drying method. Reliability of soil moisture content: the soil moisture content measured by GPR is in good agreement with the drying method. Suitable for field scale soil moisture measurement, Topp formula is more suitable than Roth formula to calculate soil water content in desert steppe area.) the spatial variability of soil water content in pixel scale and its scale effect are analyzed by using the method of geostatistics. The effect of intervariability was significant. The maximum correlation distance of soil water content increases with the scale of the sample. For the same scale, the spatial variability of the sample is different because of the complex factors such as topography, microtopography, soil and vegetation, etc. The variation coefficient varies from 12% to 19%.) the ground verification method of soil moisture content monitoring results based on ground penetrating radar is put forward: through random combination method and pankrikriging method, we can find four well zigzag lines. To a stable and reliable pixel soil water content observation, It is more accurate than the observed values at the 25 traditional points of uniform distribution. Moreover, it can accurately reflect the spatial variation characteristics of soil water content with high resolution. The GPR method is used to verify the results of remote sensing monitoring of soil moisture content at different scales: according to the ground verification method of GPR remote sensing monitoring of soil moisture content, In this paper, the accuracy of 30 m resolution remote sensing inversion of soil water content is evaluated by 4 well zigzag lines, and the relative error of soil water content inversion result is about 5.9. The validation of soil water content inversion result based on 1km resolution is given. Through the comparison and analysis of the observed data on the existing sample within 60m 脳 60m), it is difficult to reflect the soil moisture content of the 1km pixel by the measured data in the 60m 脳 60m sample square. In order to verify the results of the remote sensing monitoring of the soil moisture content of the 1km pixel accurately. It is suggested that four well-shaped lines should be arranged in the 1km pixel, and the average value of the line should be taken as the soil moisture content of the pixel.
【学位授予单位】：中国水利水电科学研究院
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S152.7;S127

【参考文献】