【作者】 李汇军；

【导师】 魏奉思； 冯学尚；

【作者基本信息】 中国科学院研究生院（空间科学与应用研究中心） ， 空间物理学， 2009， 博士

【摘要】 本文首先对日地物理过程、空间天气预报的相关概念进行了综述,然后重点介绍了行星际磁通量绳结构的观测研究结果。主要研究成果包括发展了一个磁通量绳轴向推断的新方法;证认了两个具有典型等离子体场向剩余流结构的磁云事件,提出了磁云结构中等离子体惯性效应的概念。此外,利用太阳瞬变事件的观测资料,对行星际激波传播模型(原SPM )在行星际激波波达时间(SAT)预报中的应用进行了研究,通过观测资料和预报结果的分类统计分析掌握了原SPM的SAT预报误差特性,提出了该模型在预报应用中的改进方向。(1)发展了一个磁通量绳轴向推断的新方法。利用行星际磁通量绳模型存在不同场线不变量的几何性质,我们提出了新的残差定义,用来测量通量绳轴向推断结果与真实轴向之间的偏差,并在此基础上发展了一个残差最小化(MR)方法,用于通过单飞船观测资料推断行星际通量绳结构的轴向。利用通量绳解析模型输出的数据,加上“趋势噪音”得到行星际磁通量绳的单飞船观测模拟数据,对残差最小化方法进行基准测式。结果显示:新方法可以应用于实际观测数据的分析;相应的轴向推断结果在数据采样路径离磁通量绳轴线最短距离不太大,噪音水平不太高的情况下可以接受。与传统磁通量绳轴向推断方法相比,新方法的轴向推断结果在精度上有所提高。新方法还成功地应用于由WIND飞船观测到的一个典型磁云事件的轴向分析,通过不同磁通量绳模型得到的结果对比发现:模型越精细,结果越好。表明在磁云轴向推断中,适当地采用复杂度较高的模型有利于改善轴向推断结果的精度。(2)证认了两个具有典型场向剩余流结构的磁云事件。利用WIND飞船的观测资料,我们通过MR分析得到磁云结构的轴向,通过deHo?man-Teller(HT)分析得到两个磁云事件HT速度,最后建立了磁云的自然坐标系。在此坐标系中详细地分析了磁云共行参照系下,等离子体剩余流与磁场之间的关系,结果发现两个磁云结构中的等离子体剩余流与磁场之间均保持反向平行的关系。基于磁云磁场的螺旋结构特征,我们提出:这两个磁云事件中发现的场向剩余流结构可能是磁云结构内的一种大尺度磁流体拟序涡结构。随后对两个磁云事件的动力学、热力学特性质进行分析,发现:(a)在低速背景太阳风中运动的磁云结构具有膨胀磁云的特征,而在高速背景太阳风中运动的磁云结构则不具有典型膨胀磁云的特征;(b)两个磁云结构中,质子的热力学行为近乎为等温过程;(c)两个磁云结构在1 AU附近表现出随背景太阳风对流运动的特征;(d)在磁云共行参照系中,两个磁云事件中等离子体场向剩余流的动压与磁压之比均超过热压与磁压之比,表明等离子体场向剩余流的动力学效应在磁场结构演化过程中的作用要比等离子体热压更重要。这一研究工作也显示出等离子体剩余流的惯性效应在磁通量绳结构的生成、演化与传播等物理问题的研究中有潜在价值。(3)采用分类统计分析方法研究了原激波传播模型的行星际激波波达时间预报的性能。基于点源爆炸波理论的激波传播模型(SPM)已经在行星际激波波达时间(SAT) (也称作渡越时间:Shock Transit Time)预报中得到应用。和其它半经验半物理预报模型一样,激波传播模型在应用中存在哪些问题,需要作什么样的改进?围绕着这些问题,我们利用原SPM模型的内蕴渡越时间上极限性质对其SAT预报性能进行了分类统计分析。统计结果显示:(I)激波能量的估计误差并不是SAT预报误差的唯一来源,因此我们不能指望仅通过提高激波能量估算精度就能大幅度提高SAT预报的质量;(II) SPM模型在SAT预报应用中存在系统性误差,特别是对强激波在低速背景太阳风中传播以及能量相对较低的激波在高速背景太阳风中传播时,系统性误差明显。随后针对原SPM模型的这些误差特性,我们从模型的数学基础、行星际激波的磁场位形、激波的非对称性传播、激波驱动机制等不同的角度对预报误差特性进行了解释,同时指出了原SPM模型在SAT预报应用中改进方向。基于统计分析的结,果我们对原SPM模型进行简单地修订,剔除了分类结果中SAT预报误差的趋势性,并应用修订后的SPM模型进行预报试验,结果显预报精度有明显的改善。这说明SPM模型仍有改进空间,值得进一步发展。更多还原

【Abstract】 In this dissertation, attentions are paid to the observational study on In-terplanetary Flux Ropes (IFRs) after review of the solar-terrestrial physics andthe associated concepts for space weather forecasting. A new Minimum Residue(MR) method is proposed for the axial orientation inferring of the IFRs. Twomagnetic cloud (MC) events with the field-aligned residual ?ows are identifiedand the inertial e?ects due to the plasma inside the magnetic clouds are alsohighlighted in this dissertation. Additionally, theories about the blast-wave-propagation from a point source in a moving, steady-state, medium with variabledensity are applied to the Shock Arrival Time (SAT) forecasting with the ShockPropagation Model (SPM). A statistical analysis on the forecasting capability ispresented in this dissertation. Ways of further development of the original SPMare discussed and we highlight the potential value of the non-similarity-theory-based SPM in application to the SAT forecasting.There are many 2 D models that can be used to describe the structure ofthe IFRs, such as the force-free model, the non-force-free model, and the inertialmodel. For each model, one or multiple field line invariants exist. We introducea new definition of the quantity, residue, based on all field line invariants of aspecified ?ux rope model to measure the de?ection between the assumed axisand the true ?ux rope axis. Then, a new MR method is proposed to infer theaxial orientation of IFR with the observational data from a single spacecraft.For an arbitrarily assumed ?ux rope axis, the natural coordinate system can beconstructed, then a magnetic ?ux function, A, and each invariant of the specified?ux rope model can also be concurrently calculated under this coordinate system.The direction corresponding to the minimum residue is expected to be the realaxial orientation. In previous study, the residue was first defined with A and asingle invariant Pt of a static equilibrium ?ux rope model. Here, the new MRmethod is tested with simulated magnetic cloud (MC) data sets constructed fromthe analytical model outputs of two di?erent ?ux rope models with“trend noise” added. It shows that the new MR method is applicable in real case analysis andthe inferring results are acceptable for cases with small closest approach distanceand proper noise level. Compared with results from traditional methods, accu-racy of the inferred axial orientation is improved by the new method. The newMR method is also applied to a typical in-situ event observed by WIND space-craft. The comparison of the inferring results from di?erent models indicate thatapplication of a more accurate ?ux rope model is useful for inferring techniques.Two MC events observed by WIND spacecraft are identified that there aregood examples of field-aligned residual ?ow inside those MC structures. Forboth events, the co-moving frames are determined through the deHo?man-Teller(HT) analysis and the axial orientations are inferred by the newly developedminimal residue (MR) method. The nature coordinate system for both eventsare constructed with velocity of the HT frame and the inferred MC axis, the fieldand ?ow remaining in the HT frame are analyzed at this coordinate system. As aresult, we find that the residual ?ows in the co-moving HT frame of the two MCevents are almost anti-parallel to the helical magnetic field. We speculate thatthe field-aligned residual ?ows are large scale coherent hydrodynamic vortices co-moving with the MCs at the supersonic speed near 1 AU. Data analyses show thatthe event in slow ambient solar wind is expanding at 1 AU and another one in fastsolar wind does not show apparent expansion. Proton behaviors for both eventsare quasi-isothermal. Accelerated HT analysis shows that both events have nosuitable HT frame with constant accelerations, which suggests that both eventsmay be moving at the constant speed near 1 AU under the assumptions of theHT analysis. For both events, the ratio of the dynamic pressure to the magneticpressure is larger than that of the thermal pressure to magnetic pressure, whichsuggests that the dynamic e?ects due to the plasma ?ows remaining in the co-moving HT frame are more important than the thermal e?ects in the study ofMC evolution and propagation.Methods to improve the predictions of SAT of the original SPM are inves-tigated in the last portion of this dissertation. According to the classical blastwave theory adopted in the SPM, the shock propagating speed is determinedby the total energy of the original explosion together with the background solar wind speed. Noting that there exists an intrinsic limit to the transit times com-puted by the SPM predictions for a specified ambient solar wind, we present astatistical analysis on the forecasting capability of the SPM using this intrinsicproperty. Two facts about SPM are found: (1) the error in shock energy estima-tion is not the only cause of the prediction errors and we should not expect thatthe accuracy of SPM to be improved drastically by an exact shock energy in-put; (2) there are systematic di?erences in prediction results both for the strongshocks propagating into a slow ambient solar wind and for the weak shocks intoa fast medium. Statistical analyses indicate the physical details of shock prop-agation and thus clearly point out directions of the future improvement of theSPM. A simple modification is presented here, which shows that there is roomfor improvement of SPM and thus that the original SPM is worthy of furtherdevelopment.更多还原