| 题名 | 基于GPS的电离层层析算法及其应用研究 |
| 作者 | 闻德保 |
| 学位类别 | 博士 |
| 答辩日期 | 2003-12 |
| 授予单位 | 中国科学院测量与地球物理研究所 |
| 授予地点 | 武汉 |
| 导师 | 欧吉坤 |
| 关键词 | 全球卫星定位系统 掩星 电离层 总电子含量 电子密度 电离层层析成像 不适定问题 秩亏 病态 代数重构算法 改进的代数重构算法 截断奇异值分解法 选权拟合法 组合层析算法 磁暴 赤道异常 |
| 学位专业 | 大地测量学与测量工程 |
| 中文摘要 | 电离层总电子含量（Total Electron Content，简称为TEC）是监测和了解电离层时空变化规律的一个重要物理量。然而，实际研究中，用来计算电离层TEC的模型通常是一个薄层假设模型。由于常用的薄层假设模型假定电离层是一个固定在某个高度上的薄层，因而，基于薄层假设模型获得的电离层TEC只能反映电离层的水平结构，难以反映电离层的垂直结构。电离层层析成像技术是计算机层析成像技术在电离层监测中的一种新的应用。该技术通过对电离层进行分层研究，不仅克服了薄层假设电离层模型的局限性，而且特别适合于监测大尺度电离层电子密度的垂直分布及其扰动状态。基于GPS的电离层层析成像技术是近十年来发展起来的一种新的电离层探测手段，在电离层形态与扰动的监测研究方面具有重大的科学意义和应用价值。尽管近年来一些研究者已经进行了这方面的研究工作，但由于电离层层析成像问题的复杂性，还有许多问题尚需进一步深入研究。如：电离层层析中不适定问题的存在引起反演结果不唯一或不稳定性问题及其相应算法的研究，空间天气异常条件下大尺度电离层的层析响应以及电离层层析成像技术的计算机程序尚需进一步发展和完善等。本文研究工作的主要目标是：系统地分析了电离层电子密度层析反演过程中不适定问题产生的原因以及常用的电离层层析反演算法存在的问题，进而提出了相应的解决办法。在前述工作的基础上，联合利用地基和空基GPS观测数据，分别反演了不同电离层活动状态下电离层电子密度的时空变化规律，初步获得了一些有意义的研究结果。概括起来，具体研究工作主要集中在以下几个方面： 1. 提出了一种改进的代数重构算法通过系统地分析经典的代数重构算法，发现对于大尺度电离层电子密度分布图像重构，经典的代数重构算法存在迭代收敛速度慢、计算精度低的缺陷。针对该问题，提出了一种改进的代数重构算法，其核心思想是利用新算法中上一轮迭代结果自适应地调整松弛参数向量中各元素，使其达到加快迭代收敛速度和提高重构精度的目的。数值模拟实验和GPS实测数据的反演结果证实了该算法的有效性，与经典的代数重构算法相比，新方法有效地提高了迭代收敛速度和电离层电子密度反演的精度。 2. 创新性地将选权拟合法应用到电离层电子密度层析重构中对于基于GPS的电离层层析问题，在观测数据充分且误差较小的理想情况下，利用最小二乘法能够得到比较好的反演结果。然而，由于GPS观测站分布几何条件的限制，实际用来反演电离层电子密度分布的地基和空基GPS数据通常情况下是不充足的，导致电离层电子密度反演是一个不适定问题。本文通过对重构区域内各个像素电离层电子密度之间存在的关系进行系统的分析，将欧吉坤研究员提出的选权拟合法，创新性地应用到电离层电子密度反演中，特别设计了参数权矩阵或参数约束矩阵的构造方案，增加了信息量，使得法矩阵的条件达到正常水平，有效地解决了电离层层析中的不适定问题。数值模拟实验和GPS实测数据的反演结果，验证了选权拟合法在电离层电子密度反演中的有效性和稳定性及其相对于常用的电离层层析算法的优越性。 3． 提出了一种组合电离层层析算法系统地分析了单一电离层层析算法在电离层电子密度反演过程中存在的问题，基于此，提出了一种组合电离层层析算法。利用该方法，不仅解决了不适定问题引起的解的不唯一和不稳定的问题，而且有效地解决了迭代重构算法在初值的选择上对经验电离层模型的依赖。数值模拟实验的结果证实了组合电离层层析算法的可行性及其相对于单一重构算法的优越性。在此基础上，利用组合电离层层析算法和实测的GPS数据，反演了两种不同电离层活动状态（磁静和磁暴）下区域性电离层电子密度的时空变化，初步获得了一些有意义的结果。 4. 首次利用我国实测的GPS资料，反演了磁暴发生期间中国区域上空电子密度的时空变化利用中国地壳运动观测网络提供的高精度双频GPS资料，研究了2003年8月18日强磁暴发生期间中国区域电离层电子密度的时空变化以及电离层扰动结构的演化过程。层析结果显示，此次磁暴发生期间，电离层结构存在强烈的扰动现象，且赤道异常结构向北扩展，另外，在不同纬度和不同高度上的电离层暴相不同。 5. 首次利用我国实测的GPS资料，反演了磁静条件下中国区域上空电子密度的时空变化利用2003年8月15日磁静条件中国区域的GPS资料，反演了中国区域电离层电子密度在磁静条件下的时空变化规律。初步分析了赤道异常结构的周日演化、赤道异常结构倾斜变化、赤道异常核的形成、发展和消亡的过程以及电离层电子密度垂直结构的变化 |
| 英文摘要 | Ionospheric total electron content (TEC) is an important physical parameter for monitoring and knowing the temporal-spatial variations of the ionosphere. However, all of TEC values are derived from the ionospheric models based on the single-layer assumption that the free electrons in the ionosphere are concentrated on a single-thin ionospheric shell with a fixed altitude. This assumption is an approximation to the reality and physically untrue. Hence, the single-layer ionospheric models can only be used to investigate the horizontal structure of the ionosphere, but they are inherently insufficient to describe the ionosphere in the vertical dimension. Computerized ionospheric tomography (CIT) is the new application of computerized tomography (CT) to ionospheric monitoring. This technique aims to investigate the ionosphere by delaminating mode. So it not only overcomes the limitation of those single-layer ionospheric models, but also suits to monitor large-scale spatial distribution and the perturbation of the ionosphere. GPS-based CIT, which has been developed in recent decade, is a new technique of ionospheric sounding, and it has importantly scientific significance and the application value for monitoring and investigating ionospheric morphology and perturbations. Although some researchers have carried out the investigation of GPS-based CIT in recent years, due to the complicated nature of GPS-based CIT, there still remains many problems needed to be further investigated, such as the effect of ill-posed problem of GPS-based CIT on the reconstruction results, the investigation of reasonable CIT algorithms，and the ionospheric response to the space weather such as geomagnetic storm and the high-efficiency computer program of CIT etc. The main aim of this work is to carefully analyze the arisen reason of ill-posed problem in ionospheric tomography system and the limitations of the conventional CIT algorithms, and then some new algorithms are proposed to cope with the above problems. Based on the above work, we investigate the temporal-spatial variations of ionospheric electron density under the conditions of different ionospheric activities (eg. geoagnetic quite and geomagnetic storm) by using ground-based and space-based GPS observations, respectively. Some valuable results are preliminarily obtained from the reconstructed ionospheric images. The major constribution of this dissertation is summarized as follows: 1. Putting forward an improved algebraic reconstruction technique. By analyzing the classical algebraic reconstruction technique (ART), we found that the ART have some limitations to the reconstruction of large-scale ionospheric structure, such as the low convergent speed and inversion accuracy of ionospheric electron density. To resolve these problems, an improved algebraic reconstruction technique is put forward, which improves the convergent speed and the reconstruction accuracy of ionospheric electron density by adaptively adjusting each parameter value of the relax parameter vector. Numerical simulation experiment and the inverted results from actual GPS observations demonstrated its feasibility to the reconstruction of ionospheric electron density distribution. Comparing with the classical ART, the new algorithm accelerates the convergence and improves the inversion accuracy of ionospheric electron density. 2. Creatively applying the Fitting Method by Selection of the Parameter Weights into the tomographic reconstruction of ionospheric electron density. For the GPS-based CIT, the least quare method can obtain high-quality ionospheric imaging under the ideal conditions of having sufficient GPS data and small observation errors. However, due to the limitation of the geometrical distribution of GPS receivers, ground-based and space-based GPS observations are often insufficient to the tomographic inversion of ionospheric electron density. The insufficiency of GPS observations makes the CIT an ill-posed problem. In this paper, Fitting Method by Selection of the Parameter Weights, which is proposed by Professor Ou Jikun, is creatively applied into the tomographic inversion of ionospheric electron density by analyzing the correlation of the ionospheric electron densities between those neighbouring voxels and the arisen reason of ill-posed problem. An approache to construct the constraint weight matrix is given, and the ill-posed problem of CIT is efficiently resolved. The reliability and feasibility of the Fitting Method by Selection of the Parameter Weights and its superiority to the conventional CIT algorithms are demonstrated by numerical simulation experiment and the reconstruction results based on actual GPS observations. 3. Putting forward a combined ionospheric tomography algorithm. A combined ionospheric tomography algorithm is proposed by analyzing the limitations of the conventionally single ionospheric tomography algorithms. The new algorithm not only resolves the nonunique and instablity of the solutions, but also overcomes the dependences of iterative reconstruction algorithms on the initial values, which are often obtained from experiential ionospheric models such as International Reference Ionospheric (IRI) model. Numerical simulation experiment demonstrates its superior to the single tomographic algorithm and its feasibility to the imaging reconstruction of ionospheric electron density. And then the new algorithm is used to reconstruction the temporal-spatial distribution of ionospheric electron density under the conditions of magnetic quiet and magnetic storm. Some significant results are preliminarily obtained from the reconstruction ionospheric images. 4. The temporal-spatial variations of ionospheric electron density over China during geomagnetic storm are first inverted by using the actual GPS observations. By using the GPS data from the Crustal Movement Observation Network of China, the temporal-spatial variations of ionospheric electron density over China during the storm on 18 August 2003 and the evolution of ionospheric disturbed structure are investigated. The tomographic results show that intense disturbance structure can be found, and the equatorial anomaly structure spread to northern China. In addition, the storm phase of the ionosphere varies with altitude and latitude. 5. The temporal-spatial variations of ionospheric electron density over China during geomagnetic quiet day are inverted by using the actual GPS observations for the first time. For the first time, the temporal-spatial variations of the ionosphere over China are investigated by using the CIT algorithm and GPS data in China. From the reconstructed images on 15 August 2003, we can find the diurnal evolution of equatorial anomaly, the gradient variation of equatorial anomaly (i.e. the form, development and disappearance of the equatorial anomaly core) and the variations of the vertical structure of ionosphere. |
| 公开日期 | 2013-01-17 |
| 内容类型 | 学位论文 |
| 源URL | [http://ir.whigg.ac.cn//handle/342008/3667]  |
| 专题 | 测量与地球物理研究所_学生论文_学位论文 |
| 推荐引用方式 GB/T 7714 | 闻德保. 基于GPS的电离层层析算法及其应用研究[D]. 武汉. 中国科学院测量与地球物理研究所. 2003. |
| 个性服务 |
| 查看访问统计 |
| 相关权益政策 |
| 暂无数据 |
| 收藏/分享 |
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
修改评论