SHINE: Analysis of Ion Kinetic Instabilities in the Solar Wind Observed Near the Sun with Hybrid Modeling and Machine Learning

SHINE：利用混合建模和机器学习分析太阳附近观测到的太阳风中的离子动力学不稳定性

• 批准号: 2300961
• 负责人: Leon Ofman
• 金额: $ 59.89万
• 依托单位: Catholic University of America
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-05-15 至 2026-04-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2300961&HistoricalAwards=false
• 关键词: SHINE; Analysis; Ion; Kinetic; Instabilities

The solar wind includes charged particles and magnetic fields emanating from the Sun’s outer layers. Space weather results from interactions between the solar wind and the Earth’s geomagnetic field. Therefore, it is important to understand the detailed processes that occur within the solar wind. This project explores the physics of the solar wind through analysis of satellite observations and development of machine learning models. Undergraduate and graduate students will be trained in interdisciplinary research including space plasma physics and machine learning techniques. Also, an early career post-doctoral researcher will be supported. This project is co-funded by the Directorate for Geosciences to support AI/ML advancement in the geosciences.Motivated by new observations with NASA’s Parker Solar Probe (PSP) mission, the science objective of this project is to investigate the heating and acceleration of the solar wind plasma associated with proton and alpha particle temperature anisotropy evolution, their relative drift and beaming velocities, and the associated ion kinetic instabilities. The work will focus on the effects of proton beams, detected by PSP/SPAN-I on the nonlinear evolution of the magnetosonic instability. The magnetic wave spectra and energy partition between the ions and the electromagnetic fields will be determined focusing on ion kinetic scales. The team will analyze the PSP/SPAN-I data of the proton and alpha particle velocity distribution functions (VDFs) with beams during perihelia encounters, as well as plasma moments such as density, anisotropic temperature, and alpha relative abundance data. The FIELDS instrument will provide the corresponding kinetic wave activity magnitude, spectra, and polarizations. Guided by the observations, the team will use 2.5D and 3D hybrid-particle-in-cell (hybrid-PIC) models of kinetic protons and alpha particles with background electron fluid in an expanding box model to study the kinetic instabilities driven by initially unstable non- Maxwellian VDFs such as super-Alfvénic beams and ion relative drifts in the inner solar wind. The models will be used to calculate the physical properties and nonlinear evolution of the proton and alpha particle populations in the expanding solar wind, such as the ion drift speeds, anisotropic temperatures, magnetic energy and spectra, and the associated plasma heating processes. They will develop Artificial Intelligence Machine Learning (AI/ML) methods to automate the detection of unstable VDFs and classification of the kinetic instabilities using semi-supervised (i.e., labeled, and unlabeled data) and supervised (i.e., labeled data) ML methods such as multi-layered (i.e., deep) neural networks (DNNs).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

太阳风包括太阳外层发出的带电粒子和磁场。空间天气是太阳风和地球地磁场相互作用的结果。因此，了解太阳风内部发生的详细过程是很重要的。这个项目通过分析卫星观测和开发机器学习模型来探索太阳风的物理。本科生和研究生将接受跨学科研究的培训，包括空间等离子体物理和机器学习技术。此外，还将支持一名早期职业博士后研究人员。该项目由地球科学局共同资助，以支持AI/ML在地球科学方面的进展。受NASA的帕克太阳探测器(PSP)任务的新观测的推动，该项目的科学目标是研究与质子和阿尔法粒子温度各向异性演化相关的太阳风等离子体的加热和加速，它们的相对漂移和束流速度，以及相关的离子动力学不稳定性。这项工作将集中在PSP/SPAN-I探测到的质子束对磁声不稳定性的非线性演化的影响。磁波谱和离子与电磁场之间的能量分配将集中在离子动力学尺度上确定。该团队将分析近日点相遇期间质子和阿尔法粒子速度分布函数(VDF)的PSP/SPAN-I数据，以及等离子体矩，如密度、各向异性温度和阿尔法相对丰度数据。磁场仪器将提供相应的动态波活动幅度、光谱和偏振。在观测的指导下，该团队将使用2.5D和3D混合粒子单元(HYXED-PIC)模型，在扩展盒模型中使用具有背景电子流体的动能质子和阿尔法粒子，以研究由初始不稳定的非麦克斯韦VDF驱动的动力学不稳定性，如超阿尔夫́NIC光束和内太阳风中离子的相对漂移。这些模型将被用来计算膨胀太阳风中质子和α粒子数量的物理性质和非线性演化，如离子漂移速度、各向异性温度、磁能和光谱以及相关的等离子体加热过程。他们将开发人工智能机器学习(AI/ML)方法，使用半监督(即，标记的和未标记的数据)和监督(即，标记的数据)ML方法，如多层(即，深层)神经网络(DNN)，自动检测不稳定的VDF并对动力学不稳定性进行分类。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。