基于双层聚类信息的极光亚暴自动检测

极光亚暴与太阳风和地球磁场的耦合过程有着紧密的联系,对其发生和发展机制的研究,有助于深入地分析行星际磁场、地球磁层和地球电离层的相互作用,了解太阳风携带的大量能量在地球空间的输运过程,对地球空间环境预警具有重要的意义。Polar卫星搭载的紫外极光图像成像仪能够全天候地获取紫外极光图像,在紫外极光图像中可以完整地展示出极区极光的亮度和尺度变化,尤其是可以清晰地展示出亚暴膨胀相的极光点亮和亮斑膨胀现象。现有的极光亚暴事件检测方法通常需要人工设计特征和相关规则库,耗时耗力。本文利用紫外极光图像提出了基于双层聚类信息的亚暴事件检测方法,实现了紫外极光图像数据中的亚暴事件自动检测。同时,针对极光亚暴事件检测依赖手工设计特征,设计了子空间聚类指导的三维卷积特征自动提取网络;针对极光图像帧间存在成像角度差异,利用地磁纬度和磁地方时信息对极光序列中图像的空间位置进行校正;针对卫星成像位置变化导致的成像噪声,利用极光图像级聚类保留极光亮斑区域和剔除未成像或噪声区域。主观和客观实验结果表明,本算法提升了亚暴事件检测的查全率。

极低频数据中的Pc1地磁脉动自动识别

极低频电磁台网成功观测到大量的Pc1地磁脉动事件,研究极低频Pc1地磁脉动的自动识别方法对于全面分析地球空间电磁物理环境具有重要意义.本文采用了YOLOv8目标检测网络、ResNet残差网络和定向特征增强技术,提出了一种基于计算机视觉的Pc1地磁脉动自动识别模型(Automatic Detection Model for Pc1 Geomagnetic Pulsation,简称ADM-Pc1).以大连台站和丽江台站的极低频观测数据为例,利用2015—2016年的数据作为训练集进行模型的监督学习,并使用2017—2022年的数据作为测试集对模型性能进行评估.实验结果显示,ADM-Pc1模型的F1-Score值达到了95%,错分率仅为0.9%,虚警率仅为5.8%,漏检率仅为9%,处理1天数据平均耗时是2.72 s,显著优于现有的最优识别模型.这表明,ADM-Pc1模型在识别效果和计算速度方面均能更好地满足实际工程需求.

火星感应磁层边界不稳定性研究:K-H波及其产生的磁通量绳

火星感应磁层边界的等离子体不稳定性,如边界层处速度剪切激发的Kelvin-Helmholtz(K-H)不稳定性,对火星感应磁层结构和磁层离子输运过程具有重要的影响.本文利用Mars Atmosphere and Volatile EvolutioN(MAVEN)卫星搭载的Magnetometer(MAG)和Super Thermal and Thermal Ion Composition(STATIC)仪器所提供的磁场和等离子数据,分析了2018年5月21日火星感应磁层边界处观测到的K-H不稳定性事件,发现在磁场、重离子通量、离子成分与密度、和离子速度上存在五个连续的准周期变化,估算K-H波的周期为~100 s,同时在其中两个K-H涡旋内发现了磁通量绳结构.磁通量绳的轴向方向与边界层法向准垂直,且其中的离子成分与涡旋内电离层等离子体成分接近,表明此处的磁通量绳有可能是在K-H波的发展过程中形成.磁通量绳的整体速度也远大于此处的离子逃逸速度,这可能导致重离子O^(+)和O_(2)^(+)从火星大气中快速逃逸.

用非线性磁场梯度方法对空间磁结构模型的重构研究

空间等离子体中的各种磁结构的形成是空间物理过程发展的产物,理解磁结构的拓扑形态对于认识空间物理过程有着重要作用.由于磁结构的三维(Three-dimensional, 3D)图像无法直接被观测到,因此通过卫星的局地观测来重建磁结构就成为一种很重要的研究途径.迄今已发展出多种空间磁结构的重构方法,其中“非线性磁场梯度”(Nonlinear Magnetic Gradients, NMG)方法(Shen et al., 2021)能够完备地利用四颗卫星星座的磁场与电流的局地观测和物理原理来计算星座质心处磁场的一阶梯度和二阶梯度以重构出星座附近磁场的近似形态.本文采用NMG方法重构了四种与磁层磁重联相关的动理学尺度磁结构模型:(i)一维(One-dimensional, 1D)Harris电流片与地球偶极磁场的叠加,星座质心位于赤道面X=-20R_E(地球半径)和Y=0;(ii) X线重联模型,星座质心位于X点(r=0);(iii)离子尺度轴对称的磁通量绳模型,星座质心位于通量绳轴线上;(iv) 3D分形线(separator)重联模型,星座质心位于正、负零点连线的中心.本工作重点对重构结果进行细致的误差分析及检验.结果表明,在距离星座质心显著大于10倍卫星平均间距空间范围内,用NMG方法获得的这四种磁场拓扑形态均与模型磁场相符.进而,对于前三个不存在磁零点的磁结构,NMG方法可以近似重现星座周边~10倍卫星平均间距空间范围内磁场模型的数值和方向,重构结果与模型磁场基本一致.对于存在两个磁零点的分形线重联模型,在接近与两个磁零点扇形(fan)面平行的方向上,NMG方法可以近似重现星座周边~5倍卫星平均间距空间范围内的磁场,在其他方向上NMG方法仍可以近似重现星座周边~10倍卫星平均间距范围内的磁场.由此可见,NMG方法为重构空间磁结构提供了一种有效的途径.但是必须注意,对于有磁零点的磁结构,扇形面的存在,会导致重构精度降低.

基于BiLSTM-Attention的F_(10.7)指数预测模型与中国自主数据集的应用

F_(10.7)指数是太阳活动的重要指标,准确预测F_(10.7)指数有助于预防和缓解太阳活动对无线电通信、导航和卫星通信等领域的影响.基于F_(10.7)射电流量的特性,在双向长短时记忆网络(Bidirectional Long Short-Term Memory Network,BiLSTM)基础上融入注意力机制(Attention),提出了一种基于BiLSTM-Attention的F_(10.7)预报模型.在加拿大DRAO数据集上其平均绝对误差(MAE)为5.38,平均绝对百分比误差(MAPE)控制在5%以内,相关系数(R)高达0.987,与其他RNN模型相比拥有优越的预测性能.针对中国廊坊L&S望远镜观测的F_(10.7)数据集,提出了一种转换平均校准(Conversion Average Calibration,CAC)方法进行数据预处理,处理后的数据与DRAO数据集具有较高的相关性.基于该数据集对比分析了RNN系列模型的预报效果,实验结果表明,BiLSTM-Attention和BiLSTM两种模型在预测F_(10.7)指数方面具有较好的优势,表现出较好的预测性能和稳定性.

地磁平静期赤道电集流的建模研究

本文利用Swarm-A卫星从2013年11月到2023年2月近十年观测的赤道电集流(Equatorial Electrojet,EEJ)数据,挑选地磁活动平静时(Kp≤3)EEJ的峰值电流密度为研究对象,构建了一个以经度、地方时、季节、太阳辐射通量和月相为参数的EEJ经验模型.将模型计算的EEJ与上述建模所用的Swarm-A卫星实测的EEJ进行比较发现,两者的相关系数约为0.76,平均绝对误差(均方根误差)约为20.91(27.32)mA·m^(-1).该模型可较好地反映Swarm-A卫星从2013年11月到2023年2月观测到的EEJ典型特征,比如,EEJ的幅度随着太阳辐射通量的增加而增加;EEJ有明显的日变化特征,在正午出现峰值,日出附近常出现反向EEJ;EEJ随经度变化呈现显著的波状结构;EEJ的季节变化特征明显,二分季节附近出现峰值,二至季节附近出现谷值;月相对EEJ产生明显影响,正午的EEJ在新月和满月附近出现极大值.为了进一步检验上述建立的EEJ模型的性能,我们利用Swarm-B卫星2013年11月到2023年2月期间地磁平静条件下观测的EEJ数据,对上述模型进行了验证,结果表明本文所建立的EEJ模型能较好地重现Swarm-B卫星观测的EEJ变化.

火星弓激波上游高能拾起氧离子实例分析

火星外逸层中的氧原子(O)经光致电离、电荷交换或电子碰撞产生氧离子(O^(+)),新生O^(+)随即被太阳风拾起,进而被对流电场加速.这些O^(+)被称为“拾起”O^(+).2018年11月17日08∶54∶00—10∶24∶00 UT期间,火星大气和挥发物演化任务(Mars Atmosphere and Volatile EvolutioN,MAVEN)卫星从上游太阳风进入磁鞘中,持续观测到背离太阳运动的高能拾起O^(+)(约30 keV),在相空间呈现为初速度不为零的部分环分布.分析表明:这部分高能O^(+)的初始能量较高,回旋半径较大,位于摆线运动的高能区域,穿越火星弓激波进入磁鞘,离子在该范围内能量变化不显著,只是使得相应环分布在穿越弓激波后发生偏转.这些结果有助于进一步认识火星空间高能拾起O^(+)的产生与运动轨迹.

F_(10.7)指数与Ap指数短期预报检验

国家空间天气监测预警中心(NCSW)从2004年7月1日开始对用户提供空间天气预报服务,其中包括未来24 h,48 h和72 h的F_(10.7)指数和Ap指数预报.本文对2005-2022年NCSW的F_(10.7)指数和Ap指数的预报结果进行了检验.通过检验发现,NCSW预报的未来24 h,48 h和72 h的F_(10.7)指数平均比实测值偏小;未来24 h的Ap指数平均比实测值偏大,未来48 h和72 h的Ap指数平均比实测值偏小.太阳活动水平越高,F_(10.7)指数的预报误差越大;但是Ap指数预报误差最大值反而出现在太阳活动下降段.此外,通过与跟报模型、14天回归模型、14天回归修正模型、27天回归模型和27天回归修正模型等简单的统计预报模型对比发现,NCSW的预报性能整体优于简单的统计预报模型.对于F_(10.7)指数,NCSW的预报性能略好于跟报模型,明显好于4个回归模型,但是当太阳活动水平较高时,跟报模型未来72 h的F_(10.7)指数预报性能优于NCSW;对于Ap指数,多数情况下,NCSW的预报性能明显优于统计预报模型,但是地磁扰动剧烈时,27天回归模型预报的Ap指数反而比NCSW的预报更准确.

Using restored two-dimensional X-ray images to reconstruct the three-dimensional magnetopause

Astronomical imaging technologies are basic tools for the exploration of the universe,providing basic data for the research of astronomy and space physics.The Soft X-ray Imager(SXI)carried by the Solar wind Magnetosphere Ionosphere Link Explorer(SMILE)aims to capture two-dimensional(2-D)images of the Earth’s magnetosheath by using soft X-ray imaging.However,the observed 2-D images are affected by many noise factors,destroying the contained information,which is not conducive to the subsequent reconstruction of the three-dimensional(3-D)structure of the magnetopause.The analysis of SXI-simulated observation images shows that such damage cannot be evaluated with traditional restoration models.This makes it difficult to establish the mapping relationship between SXIsimulated observation images and target images by using mathematical models.We propose an image restoration algorithm for SXIsimulated observation images that can recover large-scale structure information on the magnetosphere.The idea is to train a patch estimator by selecting noise–clean patch pairs with the same distribution through the Classification–Expectation Maximization algorithm to achieve the restoration estimation of the SXI-simulated observation image,whose mapping relationship with the target image is established by the patch estimator.The Classification–Expectation Maximization algorithm is used to select multiple patch clusters with the same distribution and then train different patch estimators so as to improve the accuracy of the estimator.Experimental results showed that our image restoration algorithm is superior to other classical image restoration algorithms in the SXI-simulated observation image restoration task,according to the peak signal-to-noise ratio and structural similarity.The restoration results of SXI-simulated observation images are used in the tangent fitting approach and the computed tomography approach toward magnetospheric reconstruction techniques,significantly improving the reconstruction results.Hence,the proposed technology may be feasible for processing SXI-simulated observation images.

Tomographic reconstruction of the Earth’s magnetosheath from multiple spacecraft:a theoretical study

Following our earlier work on tomographic reconstruction of the magnetosheath soft X-ray emissions with superposed epoch analysis of many images recorded from a single spacecraft we now explore the instantaneous reconstruction of the magnetosheath and magnetopause using a few images recorded simultaneously from a few spacecraft.This work is motivated by the prospect of possibly having two or three soft X-ray imagers in space in the coming years,and that many phenomena which occur at the magnetopause boundary,such as reconnection events and pressure pulse responses,do not lend themselves as well to superposed epoch analysis.If the reconstruction is successful-which we demonstrate in this paper that it can be-this collection of imagers can be used to reconstruct the magnetosheath and magnetopause from a single image from each spacecraft,allowing for high time resolution reconstructions.In this paper we explore the reconstruction using,two,three,and four spacecraft.We show that the location of the subsolar point of the magnetopause can be determined with just two satellites,and that volume emissions of soft X-rays,and the shape of the boundary,can be reconstructed using three or more satellites.

基于范艾伦卫星观测数据的地球辐射带电子能谱分布的统计分析

利用范艾伦卫星2014年至2018年观测的高精度电子能谱数据,统计分析了不同辐射带高能电子能谱类型的时空分布特征.结果表明,辐射带电子能谱主要被分为三类:指数分布能谱、幂律分布能谱和反转能谱.指数分布能谱通常在等离子体层以外占主导地位.幂律分布能谱通常在磁暴的主相出现在高L处,并逐渐转移到低L处.在地磁活动平静时期,幂律分布能谱在高L处存在时间更长,占比更多.反转能谱在等离子体层内L>2.5处占主导地位,随着等离子体层层顶升高,反转能谱增加,指数分布能谱减少.在长时间的地磁活动平静年份,辐射带电子反转能谱局地峰值(约2 MeV)附近的电子通量明显较低,反转能谱占比较少.辐射带电子能谱类型对等离子体层层顶和地磁活动依赖性的统计分析表明,反转能谱发生率的峰值位置约位于L=(L_(pp)–2)(L_(pp)为等离子体层层顶位置),指数分布能谱发生率的峰值约位于L=(L_(pp)+1.5),幂律分布能谱的出现与电子的注入密切相关.磁暴越剧烈,反转能谱被指数分布能谱取代的L范围越广泛.在平静时期,由于等离子体嘶声的散射作用,等离子体层层顶内的指数分布能谱被反转能谱取代.本文给出了辐射带电子能谱类型长期的时空分布特征,可以为辐射带动态建模和高能电子通量预报提供理论参考.

Preface to the Special Issue on Modeling and Data Analysis Methods for the SMILE mission

The SMILE(Solar wind Magnetosphere Ionosphere Link Explorer)project(http://www.nssc.cas.cn/smile/,https://www.cosmos.esa.int/web/smile/mission)is a joint spacecraft mission of the European Space Agency(ESA)and the Chinese Academy of Sciences(CAS)with an expected launch in 2025.SMILE aims to study the global interactions of solar wind–magnetosphere–ionosphere innovatively by imaging the Earth’s magnetosheath and cusps in soft X-rays and the northern auroral region in ultraviolet(UV)while simultaneously measuring plasma and magnetic field parameters in the solar wind and magnetosheath along a highly-elliptical and highly-inclined orbit.This special issue is composed of 22 articles,presenting recent progress in modeling and data analysis techniques developed for the SMILE mission.In this preface,we categorize the articles into the following seven topics and provide brief summaries:(1)instrument descriptions of the Soft X-ray Imager(SXI),(2)numerical modeling of the X-ray signals,(3)data processing of the X-ray images,(4)boundary tracing methods from the simulated images,(5)physical phenomena and a mission concept related to the scientific goals of SMILE-SXI,(6)studies of the aurora,and(7)ground-based support for SMILE.

Background removal from global auroral images:Data-driven dayglow modeling

Global images of auroras obtained by cameras on spacecraft are a key tool for studying the near-Earth environment.However,the cameras are sensitive not only to auroral emissions produced by precipitating particles,but also to dayglow emissions produced by photoelectrons induced by sunlight.Nightglow emissions and scattered sunlight can contribute to the background signal.To fully utilize such images in space science,background contamination must be removed to isolate the auroral signal.Here we outline a data-driven approach to modeling the background intensity in multiple images by formulating linear inverse problems based on B-splines and spherical harmonics.The approach is robust,flexible,and iteratively deselects outliers,such as auroral emissions.The final model is smooth across the terminator and accounts for slow temporal variations and large-scale asymmetries in the dayglow.We demonstrate the model by using the three far ultraviolet cameras on the Imager for Magnetopause-to-Aurora Global Exploration(IMAGE)mission.The method can be applied to historical missions and is relevant for upcoming missions,such as the Solar wind Magnetosphere Ionosphere Link Explorer(SMILE)mission.

Magnetopause properties at the dusk magnetospheric flank from global magnetohydrodynamic simulations,the kinetic Vlasov equilibrium,and in situ observations--Potential implications for SMILE

We derived the properties of the terrestrial magnetopause(MP)from two modeling approaches,one global–fluid,the other local–kinetic,and compared the results with data collected in situ by the Magnetospheric Multiscale 2(MMS2)spacecraft.We used global magnetohydrodynamic(MHD)simulations of the Earth’s magnetosphere(publicly available from the NASA-CCMC[National Aeronautics and Space Administration–Community Coordinated Modeling Center])and local Vlasov equilibrium models(based on kinetic models for tangential discontinuities)to extract spatial profiles of the plasma and field variables at the Earth’s MP.The global MHD simulations used initial solar wind conditions extracted from the OMNI database at the time epoch when the MMS2 observes the MP.The kinetic Vlasov model used asymptotic boundary conditions derived from the same in situ MMS measurements upstream or downstream of the MP.The global MHD simulations provide a three-dimensional image of the magnetosphere at the time when the MMS2 crosses the MP.The Vlasov model provides a one-dimensional local view of the MP derived from first principles of kinetic theory.The MMS2 experimental data also serve as a reference for comparing and validating the numerical simulations and modeling.We found that the MP transition layer formed in global MHD simulations was generally localized closer to the Earth(roughly by one Earth radius)from the position of the real MP observed by the MMS.We also found that the global MHD simulations overestimated the thickness of the MP transition by one order of magnitude for three analyzed variables:magnetic field,density,and tangential speed.The MP thickness derived from the local Vlasov equilibrium was consistent with observations for all three of these variables.The overestimation of density in the Vlasov equilibrium was reduced compared with the global MHD solutions.We discuss our results in the context of future SMILE(Solar wind Magnetosphere Ionosphere Link Explorer)campaigns for observing the Earth’s MP.

喉区极光过程中的能量输入估算

喉区极光是一种只发生在磁正午附近的特殊分立极光,具有明显的从分立极光卵赤道侧边界向低纬延伸的特征.已有研究表明,喉区极光对应于一种由磁重联导致的磁层顶局地凹陷.磁重联是太阳风能量注入到地球磁层空间的重要方式之一,但伴随喉区极光过程有多少能量传输进入电离层尚不明确.利用2015年12月的地基全天空极光观测与经过电离层上方的卫星粒子共轭观测,建立了极光强度与沉降电子能通量之间的相关关系.在此基础上,针对发生在2015年12月8日的一个典型喉区极光事件,在综合考虑持续时间与空间尺度基础上,估算出伴随该事件由电子携带的输入到电离层的总能量约为1.07×10^(11)J,单位面积输入能量的平均功率约为1.31×10^(-4)W·m^(-2),对比以往研究中MHD模拟给出的通常情况下极隙区能量输入结果,本文给出的喉区极光过程中能量输入的效率约为MHD模拟结果的两倍,表明伴随喉区极光过程的能量输入非常可观,可能引起局地空间天气效应.

The magnetosheath at high spectral resolution

While we eagerly anticipate SMILE’s(Solar wind Magnetosphere Ionosphere Link Explorer)unprecedented X-ray observations of the Earth’s magnetosheath and the initiation of a new era of magnetospheric research,it seems appropriate to look ahead to the abilities of the next generation of astrophysics missions.Of these,the Line Emission Mapper(LEM),a large aperture micro-calorimeter based mission,is currently planned to be able to observe the magnetosheath at high spectral resolution(~2 eV).With a field of view of~30′,LEM will allow higher spatial resolution and higher cadence measurement of the motion of a very small portion of the magnetopause over relatively short periods of time(multiple hours),complementing SMILE’s global mapping.LEM’s strength is its spectral resolution.It will be able to measure the abundance of a broad range of elements and ionization states,many of which are inaccessible to current in situ instruments,and will be able to separate the emission from the magnetosheath from the emission from the cosmic X-ray background using the difference in their relative velocities.

Finding the magnetopause location using soft X-ray observations and a statistical inverse method

Variability in the location and shape of the dayside magnetopause is attributed to magnetic reconnection,a fundamental process that enables the transfer of mass,energy,and momentum from the solar wind into the magnetosphere.The spatial and temporal properties of the magnetopause,under varying solar and magnetospheric conditions,remain largely unknown because empirical studies using in-situ observations are challenging to interpret.Global wide field-of-view(FOV)imaging is the only means to simultaneously observe the spatial distribution of the plasma properties over the vast dayside magnetospheric region and,subsequently,quantify the energy transport from the interplanetary medium into the terrestrial magnetosphere.Two upcoming missions,ESA/CAS SMILE and NASA’s LEXI will provide wide-field imagery of the dayside magnetosheath in soft X-rays,an emission generated by charge exchange interactions between high charge-state heavy ions of solar wind origin and exospheric neutral atoms.High-cadence two-dimensional observations of the magnetosheath will allow the estimation of dynamic properties of its inner boundary,the magnetopause,and enable studies of its response to changes in the solar wind dynamic pressure and interplanetary magnetic field orientation.This work introduces a statistically-based estimation approach based on inverse theory to estimate the spatial distribution of magnetosheath soft X-ray emissivities and,with this,identify the location of the magnetopause over the Sun−Earth line.To do so,we simulate the magnetosheath structure using the MHD-based OpenGGCM model and generate synthetic soft X-ray images using LEXI’s orbit and attitude information.Our results show that 3-D estimations using the described statistically-based technique are robust against Poisson-distributed shot noise inherent to soft X-ray images.Also,our proposed methodology shows that the accuracy of both three-dimensional(3-D)estimation and the magnetopause standoff distance calculation highly depends on the observational point.

Mshpy23:a user-friendly,parameterized model of magnetosheath conditions

Lunar Environment heliospheric X-ray Imager(LEXI)and Solar wind−Magnetosphere−Ionosphere Link Explorer(SMILE)will observe magnetosheath and its boundary motion in soft X-rays for understanding magnetopause reconnection modes under various solar wind conditions after their respective launches in 2024 and 2025.Magnetosheath conditions,namely,plasma density,velocity,and temperature,are key parameters for predicting and analyzing soft X-ray images from the LEXI and SMILE missions.We developed a userfriendly model of magnetosheath that parameterizes number density,velocity,temperature,and magnetic field by utilizing the global Magnetohydrodynamics(MHD)model as well as the pre-existing gas-dynamic and analytic models.Using this parameterized magnetosheath model,scientists can easily reconstruct expected soft X-ray images and utilize them for analysis of observed images of LEXI and SMILE without simulating the complicated global magnetosphere models.First,we created an MHD-based magnetosheath model by running a total of 14 OpenGGCM global MHD simulations under 7 solar wind densities(1,5,10,15,20,25,and 30 cm)and 2 interplanetary magnetic field Bz components(±4 nT),and then parameterizing the results in new magnetosheath conditions.We compared the magnetosheath model result with THEMIS statistical data and it showed good agreement with a weighted Pearson correlation coefficient greater than 0.77,especially for plasma density and plasma velocity.Second,we compiled a suite of magnetosheath models incorporating previous magnetosheath models(gas-dynamic,analytic),and did two case studies to test the performance.The MHD-based model was comparable to or better than the previous models while providing self-consistency among the magnetosheath parameters.Third,we constructed a tool to calculate a soft X-ray image from any given vantage point,which can support the planning and data analysis of the aforementioned LEXI and SMILE missions.A release of the code has been uploaded to a Github repository.

子午工程二期大气电场仪及其初步观测

大气电场是大气科学和空间物理学科中共同的一个重要电学参量,大气电场的变化代表着全球大气电路和近地面大气电荷分布状态,其变化同时受到气象活动、地质活动和太阳活动影响.大气电场的平均日变化特点代表着平静时期该地区近地面大气电场的平均值,这对于研究不同纬度地区近地面大气电场由于地质活动或太阳活动引起的扰动具有重要意义.为实现对日地空间环境全圈层、多要素综合的立体式探测,子午工程在两极、中国多个地区建成了多个大气电场观测台站,本文主要介绍子午工程中的大气电场观测,包括场磨式大气电场仪及其数据格式,展示了子午工程建设的西藏噶尔站2021年11月至2022年10月一年期间的晴天大气电场平均日变化,并对其进行了分析以及与“卡耐基曲线”的比较.

基于神经网络的未来3天Kp指数预报建模与可解释AI应用

当前业务中对未来3天Kp指数预报需求强烈.但地磁暴中多参数耦合导致难以量化各预报因子对Kp值的贡献,制约了预报精度提升.本文构建了神经网络3天Kp指数预报模型,并使用人工智能(AI)可解释性算法定量化各因子贡献.结果显示,行星际磁场南向分量在提前3 h对Kp指数的贡献为37.15%,为主要因子,说明模型能捕捉符合物理特征的主要预报因子.Kp指数历史特征贡献随提前量逐渐增加,提前3天总体贡献占68.06%,验证了对冕洞高速流引起的地磁暴事件的预报能力.对2015和2017年特大地磁暴进行贡献分析,模型准确捕捉了地磁暴多参数耦合的复杂特性.研究表明,可解释AI算法在一定程度上能定量化各预报因子对Kp指数的预报贡献,有助于改进未来3天Kp指数AI预报模型.