单站高频GNSS求解同震位移的新方法及联合强震仪的地震预警应用研究

发布时间：2018-05-10 17:18

本文选题：地震预警 + 同震位移　；参考：《武汉大学》2015年博士论文

【摘要】：地震是地球上最为常见的自然灾害之一,其中极具破坏力的大地震带来的原生灾害、次生灾害和衍生灾害给人类造成了巨大的威胁。虽然无法控制和避免地震的发生,但是人类也在积极地寻求应对之策。目前,在准确的地震短期预测还无法实现的情况下,地震预警便是一种行之有效的防震减灾措施。历经三十年的探索和努力,很多国家和地区建立了以地震仪/强震仪为基础的地震预警系统,并且取得了众多成功预警的实例。但是,这样的地震预警系统在应对大地震时(M7)还存在问题,尤其在地震震级估算上容易“失灵”。这是由于地震震级估算与震时地表位移密切相关,受基线偏差和量程限幅的影响,地/强震仪很难准确把其测量值转化为与地震过程相吻合的地面位移值。因此,如何快速、准确获取震时地表位移成为了地震预警系统可靠运行必须解决的关键问题。近年来,随着高频GNSS技术(高采样率数据,存储技术等)的快速发展,其已能用于监测短期瞬时的地表运动状态。由于GNSS不会受到类似于地/强震仪器记录限幅和倾斜的影响,因而即使在大地震中,也可以较好地获取地震近场同震位移,为地震预警提供了一种新的、可靠的观测手段。虽然高频GNSS在获取震时地表位移上具有优势,但是相对于地/强震仪而言,它在实时地震监测和预警应用中还存在一些限制因素和挑战：(1)GNSS的直接观测量是伪距、载波相位和多普勒观测值,准确的测站位置和速度信息需要通过高精度的数据处理算法才能获得,而且这些算法还需具备实时性；(2)GNSS的测量噪声较大,对微弱地表震动不敏感,不易拾取到初至地震波；(3)GNSS受限于接收机的频带宽度,无法获得与地/强震仪相当的采样率(200 Hz甚至400 Hz)。因此,高频GNSS和地/强震仪具有明显的优势互补性。联合两者将有助于提高地震预警的准确性和时效性,这也是当前地震学的研究热点和前沿课题。本论文旨在将高频GNSS技术引入到现行地震预警系统中,以增强其在面对大地震时的预警能力。围绕这一核心目标,本文主要从数据处理的角度出发,深入研究并提出了单站GNSS单历元求解测站同震位移的新方法,详细推导并建立了GNSS和强震仪组合模型,实现了两者联合进行地震预警应用的目的。论文的主要工作和贡献如下：(1)系统研究了常用的两种利用单站GNSS求解同震位移的方法：精密单点定位(PPP)和单站测速方法。首先介绍了各方法的基本原理,然后建立了相应的求解同震位移模型,最后结合具体地震事件分析了各方法的应用效果。深入分析了接收机钟跳对单站测速法的不利影响,提出了一种顾及接收机钟跳影响的多普勒观测值构造方法,并用动态和静态算例验证了所提方法的可行性和有效性。(2)以单站GNSS实时准确获取同震位移为目标,并结合地震应用的实际需求(获取一段时间的同震位移,并且震前测站坐标已知),从GNSS基本观测模型出发,提出了一种直接求解测站同震位移的新方法——(附有已知坐标的)历元间位置差法。根据数据处理策略的不同,历元间位置差法又可细分成TPP法和R-variometric法。从函数模型的角度出发,证明了PPP法,TPP法和R-variometric法的等价性。详细分析了该方法中所包含的各种主要误差对位移的影响,并给出相应的解决策略。在顾及各项误差改正后,采用历元间位置差法模型能够获得与事后PPP法精度相当的同震位移。实例结果表明：在20分钟的时间里,该法所获得同震位移的精度在水平方向上约为3厘米,在垂直方向上约为5厘米。(3)针对地震监测和预警需要高密度的站网需求以及双频GNSS接收机成本高昂等问题,提出了一种电离层空域建模模型——MSEID模型,将历元间位置差法的适用范围从双频接收机拓展至单频接收机,从而使单频接收机能够高精度的获取同震位移并用于大范围加密现有的双频GNSS观测台网。详细分析和讨论了MSEID模型恢复电离层延迟的有效性,所提模型获得位移的准确性以及适用性。实例结果表明：MSEID模型能够足够精确的恢复出电离层延迟变化情况,并且比SEID模型的计算效率更高；所提模型存在一定的适用范围,对于网内加密,双频测站间距最好小于80公里,对于网外加密,单频测站距离双频测站的平均间距最好在50公里范围内；以双频观测数据计算的同震位移作为参照,所提模型能够获得准确的同震位移,精度在平面方向为2厘米,高程方向上为4厘米。(4)详细对比和分析了强震仪和高频GNSS两种观测手段在获取同震位移时的优缺点。为了克服单一仪器的不足,提出了顾及基线偏差改正的GNSS和强震仪组合模型。实例验证表明：通过Kalman滤波算法能够将两种观测仪器的优势充分结合起来,充分发挥其各自优势,获得了更为准确的宽频带同震位移。在滤波的早期阶段具有更多强震仪结果的特性,可用于探测测站或地面微弱地运动；而在滤波的中后期,则受到更多GNSS位移结果的约束,使得组合位移不会发散,因而能够准确地记录地震波的低频信息。(5)开展了联合高频GNSS和强震仪在地震预警中的应用研究工作,详细探讨并分析了基于组合位移结果进行地震波到时拾取、地震震中位置确定以及地震震级估算的方法及其效果。以东日本大地震为例,联合使用STA/LTA方法和AIC准则进行地震波P波到时拾取,所得结果与理论到时平均时延约0.8秒；采用Geiger法和网格搜索法进行地震震中确定,能在震后18.52秒,分别获得17公里和4公里偏差的震中位置,相应的地震发震时刻偏差在0.5s范围内；根据峰值位移(Pd与PGD)与地震震级的经验关系进行震级估算和预警信息播报,在震后26.35秒可进行第一次预警播报,相应震级为Mw 7.4,在震后102.19秒进行最后一次播报,可获得Mw 8.9的预警最终震级。上述结果表明基于组合的方式,能够获得足够精确的位移结果,可用于P波探测,能够在较短时间内(102.19秒)获得与真实震级十分接近的最终震级而未发生震级饱和现象,这对于增强现有地震预警系统面对大地震时的能力具有极其重要的作用。
[Abstract]:Earthquake is one of the most common natural disasters on the earth, of which the devastating earthquake caused by the great earthquake, secondary and derivative disasters caused great threat to mankind. Although it is impossible to control and avoid the occurrence of earthquakes, human beings are also actively seeking countermeasures. At present, the accurate short-term prediction of earthquake is still available. Earthquake early warning is an effective measure of earthquake prevention and disaster reduction. After thirty years of exploration and efforts, many countries and regions have established seismic early warning system based on seismograph / strong seismograph, and have obtained many examples of successful early warning. M7) still exists problems, especially in earthquake magnitude estimation, which is easy to "fail". This is because the estimation of earthquake magnitude is closely related to the surface displacement of the earthquake. Influenced by the baseline deviation and the range limit, the ground / strong seismograph is difficult to accurately translate its measurement value into the ground displacement value that is consistent with the earthquake process. In recent years, with the rapid development of high frequency GNSS Technology (high sampling rate data, storage technology, etc.), the surface displacement has been the key problem to be solved in the reliable operation of the earthquake early warning system. It has been used to monitor the short-term instantaneous surface movement state. Since GNSS will not be affected by the amplitude and tilt of the records similar to the earth / strong seismic instruments, Therefore, even in a large earthquake, the same seismic displacement can be obtained well, which provides a new and reliable means of observation for earthquake early warning. Although the high frequency GNSS has the advantage of the ground displacement when obtaining the earthquake, there are still some restrictive factors in the real-time earthquake monitoring and early warning application relative to the ground / strong seismograph. And challenges: (1) the direct measurement of GNSS is pseudo range, carrier phase and Doppler observational value. Accurate location and velocity information of stations need to be obtained through high precision data processing algorithms, and these algorithms need to be real-time. (2) GNSS has a large measurement noise and is not sensitive to weak ground motion, and it is difficult to pick up the initial ground to the ground Seismic waves; (3) GNSS is limited to the band width of the receiver and can not obtain the equivalent sampling rate (200 Hz or even 400 Hz) with the earth / strong seismograph. Therefore, the high frequency GNSS and the ground / strong seismograph have obvious complementary advantages. The combination of the two will help to improve the accuracy and time effectiveness of earthquake early warning, which is also a hot and frontier course of seismology. The purpose of this paper is to introduce high frequency GNSS technology into the current earthquake warning system in order to enhance its early warning ability in the face of large earthquakes. In this paper, a new method of single station GNSS single epoch to solve the same earthquake displacement is studied and proposed in detail from the point of view of data processing, and GNS is derived and established in detail. The combined model of S and strong seismograph has been used to achieve the purpose of joint earthquake early warning application. The main work and contribution of this paper are as follows: (1) two methods used to solve the same earthquake displacement using single station (PPP) and single station velocity measurement are studied systematically. First, the basic principles of each method are introduced, and then the method is established. The same earthquake displacement model is solved and the application effect of each method is analyzed in the end. The adverse effect of the receiver clock jump on the single station velocity measurement is analyzed. A method of constructing the Doppler observation value considering the effect of the receiver's bell jump is proposed, and the feasibility of the proposed method is verified by the dynamic and static examples. And effectiveness. (2) taking the single station GNSS to obtain the same earthquake displacement in real time as the target, and combining the actual demand of the seismic application (obtaining a time of the same earthquake displacement and the known coordinates of the station before the earthquake), a new method of directly solving the same earthquake displacement of the station is proposed from the basic GNSS observation model, which is with the known coordinates. Position difference method. According to different data processing strategies, the location difference method between epochs can be subdivided into TPP method and R-variometric method. From the point of view of function model, the equivalence between PPP method, TPP method and R-variometric method is proved. The influence of various main errors on the displacement in this method is analyzed in detail, and the corresponding solutions are given. After taking into account the correction of various errors, the same seismic displacement equivalent to the ex post PPP method can be obtained by using the positional difference method between the epochs. The example shows that the accuracy of the same earthquake displacement obtained in the 20 minute time is about 3 cm in the horizontal direction and about 5 cm in the vertical direction. (3) for earthquake monitoring and early warning. With the need of high density station network and high cost of dual frequency GNSS receiver, an ionospheric spatial domain modeling model, MSEID model, is proposed. The application range of the location difference method between the epochs is extended from the dual frequency receiver to the single frequency receiver, so that the single frequency receiver can obtain the same earthquake displacement with high precision and be used in the large range. The existing dual frequency GNSS observation network is encrypted. The effectiveness of the MSEID model to restore the ionospheric delay is analyzed and discussed in detail. The accuracy and applicability of the proposed model are obtained. The results show that the MSEID model can restore the ionospheric delay change accurately enough, and is more efficient than the SEID model. The model has a certain range of application. For the network encryption, the spacing of the dual frequency station is best less than 80 kilometers. For the outside network encryption, the average distance between the two frequency stations of the single frequency station is best in the range of 50 kilometers, and the same earthquake displacement can be obtained with the same earthquake displacement calculated by the double frequency observation data. The direction of the plane is 2 cm and the elevation direction is 4 cm. (4) the advantages and disadvantages of the strong seismograph and the high frequency GNSS two observation means in obtaining the same earthquake displacement are compared and analyzed in detail. In order to overcome the shortage of the single instrument, the GNSS and the combined model of the strong seismograph are put forward to take into account the correction of the baseline deviation. The example shows that the Kalman filter algorithm can be used. It is sufficient to combine the advantages of the two observational instruments fully, give full play to their respective advantages, and obtain more accurate broadband coisiseismic displacement. In the early stage of filtering, more strong seismograph results can be used to detect the measurement stations or ground motion, but in the middle and later stages of filtering, the results of more GNSS displacement are about. As a result, the combination displacement will not diverge and can accurately record the low frequency information of seismic waves. (5) the application of combined high frequency GNSS and strong seismograph in earthquake early warning is carried out, and a detailed discussion and analysis are made on the basis of the combined displacement results to pick up seismic waves, determine the location of the earthquake epicentre and estimate the magnitude of the earthquake magnitude. The method and its effect. Taking the East Japan earthquake as an example, the STA/LTA method and the AIC criterion are combined to pick up the seismic wave P wave, and the results are about 0.8 seconds in the mean time delay of the theory, and the epicenter of the earthquake epicenter is determined by the Geiger method and the grid search method, and the epicenter position of 17 km and 4 km deviation can be obtained in the post earthquake, and the phase of the epicenter of 17 km and 4 km can be obtained. The seismic time deviation should be in the range of 0.5s, and the magnitude estimation and early warning information are carried out according to the empirical relationship between the peak displacement (Pd and PGD) and the earthquake magnitude. The first early-warning broadcast can be carried out in 26.35 seconds after the earthquake, the corresponding magnitude is Mw 7.4, the last broadcast is carried out in 102.19 seconds after the earthquake, and the final magnitude of the early warning of Mw 8.9 can be obtained. The above results show that the combination method can obtain enough accurate displacement results, which can be used for P wave detection, and can get the final magnitude close to the real magnitude in a short time (102.19 seconds) without magnitude saturation, which is very important to enhance the ability of the existing earthquake warning system to face the large earthquake. The role.

【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：P315;P228.4

【相似文献】