【作者】 吴云龙；

【导师】 罗志才； C.C.Tscherning；

【作者基本信息】 武汉大学 ， 固体地球物理学， 2010， 博士

【摘要】 地球重力场研究历来是现代大地测量学领域的主要研究任务之一。随着卫星重力探测技术的发展,卫星跟踪卫星技术(SST)和卫星重力梯度测量技术(SGG)被国际上公认为当今获取全球高精度高分辨率地球重力场及其时变信息的最有效技术手段。GOCE卫星已经于2009年3月17日成功发射,利用卫星重力梯度数据确定地球重力场的理论与方法已成为国际大地测量学界研究的热点和趋势。高质量的重力梯度测量数据预处理将为恢复高精度静态地球重力场模型提供数据质量保障,既是卫星重力梯度测量数据处理及应用中的重要环节,也是实现GOCE预期科学目标的关键任务之一。在此研究背景下,本文旨在紧跟国际卫星重力梯度技术的发展步伐,系统研究GOCE卫星重力梯度数据的预处理理论与方法,包括重力梯度观测数据的时变重力场效应改正、粗差探测方法和外部校准方法,研制自主知识产权的卫星重力梯度数据预处理软件,对发展我国卫星重力梯度测量技术具有重要的科学意义和应用价值。本文的主要研究工作及成果如下：(1)详细阐述了卫星重力探测技术的发展和现状,归纳总结了利用卫星重力梯度测量技术确定地球重力场的理论与方法、卫星重力梯度测量数据的预处理方法及相关研究进展。(2)系统研究了卫星重力梯度测量的基础理论,详细推导了重力梯度测量观测方程,给出了GOCE卫星涉及的时间系统、坐标系统和相应转换公式,以及重力梯度张量在不同坐标系间的转换关系；详细介绍了GOCE卫星任务的科学目标、技术特点、卫星系统的组成、数据产品及分类,分析了重力梯度观测值和卫星轨道精密定轨的误差特性。(3)系统研究了GOCE重力梯度测量数据时变效应中的潮汐影响,包括天文潮汐、固体潮、海潮和极潮等潮汐效应对卫星重力梯度测量数据的影响,给出了卫星重力梯度观测数据的潮汐改正公式。计算结果表明：潮汐对重力梯度数据的总体影响量级为10-3E量级,接近于GOCE卫星设计的测量精度指标。考虑到潮汐效应具有系统性,在GOCE重力梯度数据预处理中可看作是有色噪声源,需要在数据预处理环节将其剔除。(4)详细分析了时变重力场效应非潮汐影响中的大气与海洋质量变化、陆地水质量变化等影响源对卫星重力梯度测量数据的影响,给出了相应的重力梯度观测数据的改正公式。计算结果显示：大气与海洋质量变化、陆地水质量变化对重力梯度观测数据的影响总体上比GOCE重力梯度测量精度指标低1-2个量级。(5)深入研究了重力梯度观测数据的粗差探测理论与方法,包括阈值法、Grubbs检验法和Dixon检验法等统计检验方法和小波分析方法,给出各种方法的实用计算模型。利用含有噪声和不同分布类型粗差的模拟卫星重力梯度测量数据计算结果表明：统计检验方法具有计算简单、快速的特点,能够较好的识别离散态粗差,总体探测成功率较高,适用于粗差探测的快速解算；小波分析方法粗差探测成功率很高,失败率非常低,且能适应多种粗差分布类型,适合于粗差探测的精密解算。(6)提出了基于小波收缩阈值降噪方法和Dixon检验法的组合粗差探测方法,计算结果表明：组合粗差探测方法的成功率和失败率都优于上述几种单独的粗差探测方法,能够正确识别基本上所有的离散态粗差,同时对于区域态粗差也有很好的识别率。建议在GOCE实测数据预处理中采用该组合粗差探测方法。(7)深入研究了重力梯度观测值的外部校准方法,包括基于地球重力场模型的校准方法、基于高-低卫星跟踪卫星数据的校准方法和基于高精度地面重力数据的校准方法,给出了各种方法的计算模型。计算结果表明：利用高精度地球重力场模型能够以104的相对精度校准卫星重力梯度数据的尺度因子和偏差；利用局部高精度地面重力数据能够以10-2的相对精度校准卫星重力梯度数据的尺度因子。(8)研究制订了GOCE卫星重力梯度测量数据处理方案,给出了L1b数据预处理流程,独立研制了GOCE卫星重力梯度数据预处理软件,该软件具备重力梯度观测数据的时变重力场效应改正、粗差探测和外部校准等功能。更多还原

【Abstract】 THE EARTH’S gravity field is always one of the main tasks of modern geodesy research field. With the breakthrough of satellite gravimetry technique, the Satellite-to-Satellite Tracing (SST) and Satellite Gravity Gradiometry (SGG) are regarded as the most effective techniques for the determination of the earth’s gravity field and its temporal variation with high accuracy and high resolution. As GOCE satellite successfully launched on 17th March 2009, theory and methodology of gravity field recovery using Satellite Gravity Gradiometry data become a hot issue in international geodesy research field. The satellite gradiometry data pre-processing in high quality can provide pure data for recovering static earth’s gracity field model with high accuracy, which is a key link in satellite gradiometry data process and application. Study on pre-processing of GOCE satellite gradiometry data is one of the key tasks to achieve the scientific goals of GOCE mission. Under the background of the scientific research, keeping the pace with the development of the satellite gravity gradiometry technique, the theory and methodology of pre-processing of GOCE satellite gradiometry data are studied in this dissertation, including temporal gravity field correction, outliers detection and external calibration on GOCE data. The corresponding GOCE data pre-processing software with autonomous copyright are developed, which has scientific significance and can be applicated to develop our own satellite gravity gradiometry technique.The main research work and contributions of this dissertation are as follow:THE milestone and current status of satellite gravimetry technique is reviewed in detail. The theory and methodology for the determination of the earth’s gravity field base on satellite gravity gradiometry technique, pre-processing of GOCE satellite gradiometry data and related research progress home and abroad are summarizedFUNDAMENTAL theories of the satellite gradiometry are researched. The satellite gradient observation equations are derived in detail. Then time reference system and coordinate reference frame related to GOCE satellite and the corresponding transformations are introduced. And the gravity gradient tensor transformations between different coordinate system are derived. The GOCE satellite mission is comprehensively reviewed, including scientific objectives, technique characteristics, satellite system components, data product and its classification etc. The error properties of satellite gradiometry data and precise orbit determination are analysed.TIDAL effects among temporal gravity field correction in GOCE satellite gradiometry data are deeply studied, including the effect of astronomical tides, solid earth tides, ocean tides and pole tides on satellite gradiometry data. And tidal effect corrections for satellite gradiometry data are derived. The results indicate that total tidal effect is 10-3E in order, close to GOCE measurement accuracy. Considering the systemanic of tidal effect, it should be removed as colored noise in GOCE data pre-processing.(?)ON-TIDAL effects on GOCE gravity gradiometry observation are discussed, including atmospheric and oceanic high-frequency mass variations, land hydrology. The corresponding correction models are derived. It can be indicated that the effect of atmospheric and oceanic high-frequency mass variations, land hydrology are 2 degrees lower in order than GOCE measurement accuracy.THEORY and methodology of outlier detection for satellite gradiometry data are deeply stuied, including Threshold method, Grubbs’s test, Dixon test, wavelet analysis method. Practical numerical models of ech method are derived. The computation results using satellite gradiometry data with noise and different types of outliers indicate that statistics method, which is easy and fast, can detec innovative outlier with a rather high outlier ratio of success. The statistics method can be applied to quick-look outlier detection. Wavelet method is valid for all three outlier distribution types, with a high outlier ratio of success and low outlier ratio of failure.BASE on wavelet shrinkage denoising method and Dixon test, combined algorithm foroutlier detection is presented. The computation result indicates that combined method shows best performance than the other outlier detection methods, with successful detection of nearly all the innovative outliers and most bulk outliers. It is suggested to apply the combined method for pre-processing of satellite gradiometry dataMETHODOLOGY of external calibration of gradiometry observations using global-gravity models, regional terrestrial gravity data and high-low SST data are deeply stuied. The computation results indicate that it can achieve relative accuracy at 10-4 order to calibrate scale factor and bias base on high accuracy earth’s gravity model. The relative accuracy of calibration on scale factor base on regional terrestrial gravity data can achieve 10"2 order.GOCE satellite gradiometry data processing scheme are studied, L1b data pre-processing flow is presented. GOCE satellite gradiometry data pre-processing software is developed independently, with the functions as temporal gravity field correction, outlier detection and external calibration.更多还原