基于宇宙射线中子法的荒漠草原像元尺度土壤水分监测与验证

发布时间：2018-02-02 13:03

本文关键词： 土壤水分宇宙射线中子法遥感像元尺度耦合验证　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】：土壤水分是水循环、能量循环和生物地球化学循环过程中的基本组成部分,在水文、气候、农业和生态等研究领域都具有十分重要的作用。大范围的土壤水分遥感监测过多地依赖点测量结果,这两种方法空间尺度上的较大差异导致遥感反演结果的精度难以得到保证。因此,本研究基于区域尺度的宇宙射线中子法,开展烘干称重法和介电特性法FDR点测量、宇宙射线中子法CRS区域监测和像元遥感反演的多尺度土壤水分协同观测试验和耦合验证研究。该研究不仅能够丰富土壤水分测量技术,而且能够从像元尺度上为遥感反演结果提供有效的验证手段,对水文、生态和气候等科学研究都具有重要作用。本论文的研究内容和结论主要有:荒漠草原宇宙射线中子法土壤水分测量。利用土壤水分的烘干称重法测量结果以及FDR和CRS的连续测量结果,开展了宇宙射线中子法在荒漠草原的土壤水分测量应用研究,证明该方法对荒漠草原土壤水分的测量精度能够满足科学研究和实际应用的要求,具有较好的发展潜力。通过对植被和降雨进行影响因素分析,证明研究区植被对测量结果的影响很小,同时证明CRS对降雨的响应比FDR灵敏。基于中子数与氢的密切相关性,本研究利用不同位置的中子强度在中子总量中的权重作为对应位置土壤水分的权重,构建土壤水分尺度转换模型,并提出了基于中子强度的加权法;通过四种加权方法的对比分析,发现基于中子强度加权法的平均土壤水分与CRS 土壤水分的拟合效果最好,相关系数高达0.92;择优选取该加权法作为模型参数(权重系数)校准的最佳方案,由此建立了 CRS测量足迹内的尺度转换模型;基于权重系数的最优解,分析出CRS测量足迹的实际空间特征:①CRS的足迹范围不是标准的圆柱体,而是扁平的陀螺型;②关于土壤水分点测量结果的权重,垂直权重随深度的增大而减小,水平权重随径向距离的增大而减小。热惯量法土壤水分遥感反演。采用热惯量法,分别开展Landsat 8 30m卫星影像数据和MODIS 1km数据产品的土壤水分遥感反演。其中,对于Landsat8数据的热惯量法反演参数温度日较差△T,通过结合Landsat8日地表温度反演结果和MODIS夜地表温度产品计算得到。多尺度土壤水分测量结果耦合验证。通过FDR点测量、CRS区域测量和遥感像元反演的多尺度土壤水分协同观测试验,基于CRS足迹范围内的尺度转换模型,统一不同测量方法的空间尺度,开展不同尺度土壤水分观测结果的耦合验证。CRS与FDR 土壤含水量拟合的R2高达0.84,RMSE仅为0.0128kg/kg,相对误差多数都在20%以内,说明宇宙射线中子法的测量结果精度较高,能够满足科学研究和生产应用的要求。FDR 土壤含水量与30m像元和1km像元遥感反演结果拟合的R2均为0.74,CRS与两种像元反演结果拟合的R2分别为0.82和0.81,说明CRS区域土壤含水量对遥感反演结果的验证效果更佳。
[Abstract]:Soil moisture is a basic component of the water cycle, energy cycle and biogeochemical cycle processes in hydrology and climate. Research fields such as agriculture and ecology play a very important role. A wide range of remote sensing monitoring of soil moisture depends too much on the results of point measurement. The difference between the two methods on spatial scale makes the precision of remote sensing inversion difficult to be guaranteed. Therefore, the cosmic ray neutron method based on regional scale is studied in this paper. The measurement of FDR points by drying weighing method and dielectric characteristic method was carried out. The multi-scale soil moisture cooperative observation experiment and coupling verification study of cosmic ray neutron method CRS area monitoring and pixel remote sensing inversion can not only enrich the soil moisture measurement technology. And it can provide an effective verification method for remote sensing inversion results from the pixel scale to hydrology. Ecological and climatic studies play an important role. The main contents and conclusions of this paper are as follows:. Soil moisture was measured by cosmic ray neutron method in desert steppe. The results were measured by drying and weighing method of soil moisture and the continuous measurement results of FDR and CRS. The application of cosmic ray neutron method to soil moisture measurement in desert steppe has been studied. It is proved that the precision of this method can meet the requirements of scientific research and practical application. Through the analysis of the influence factors of vegetation and rainfall, it is proved that the vegetation in the study area has little effect on the measured results. It is also proved that the response of CRS to rainfall is more sensitive than that of FDR, based on the close correlation between neutron number and hydrogen. In this study, the weight of neutron intensity in the total neutron amount was used as the weight of soil moisture in the corresponding position, and the scale conversion model of soil moisture was constructed, and the weighting method based on neutron intensity was proposed. Through the comparison and analysis of four weighting methods, it was found that the fitting effect of average soil moisture and CRS soil moisture based on neutron intensity weighting method was the best, and the correlation coefficient was as high as 0.92; The weighting method is chosen as the best method for calibration of model parameters (weight coefficient), and the scale conversion model within the footprint of CRS is established. Based on the optimal solution of weight coefficient, the actual spatial feature of CRS measurement footprint: 1: 1CRS is analyzed. The footprint range of CRS is not a standard cylinder, but a flat gyroscope. 2 for the weight of soil moisture measurement results, the vertical weight decreases with the increase of depth, and the horizontal weight decreases with the increase of radial distance. The thermal inertia method is used to retrieve soil moisture by remote sensing. Soil moisture remote sensing inversion of Landsat 830m satellite image data and MODIS 1km data products were carried out respectively. For the Landsat8 data, the thermal inertia method is used to retrieve the daily temperature difference T of the parameters. By combining the Landsat8 daily surface temperature inversion results with the MODIS night surface temperature products calculation. The results of multi-scale soil moisture measurement coupled verification. Through the FDR point measurement. Based on the scale conversion model in the range of CRS footprint, the multi-scale soil moisture cooperative observation experiment based on CRS regional measurement and remote sensing pixel inversion unifies the spatial scale of different measurement methods. The results of soil moisture observation at different scales were coupled to verify that the R2 fitting between CRS and FDR soil moisture content was as high as 0.84kg / kg, only 0.0128kg / kg. Most of the relative errors are less than 20%, which indicates that the accuracy of the cosmic ray neutron method is high. FDR soil moisture content fitting with 30m pixel and 1km pixel fitting R2 is 0.74. The R2 fitting between CRS and two kinds of pixel inversion results is 0.82 and 0.81respectively, which indicates that the soil moisture content in CRS area is more effective for the validation of remote sensing inversion results.
【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】：S812.2

【参考文献】