RII Track-4: Investigating Solar Wind-Magnetosphere Coupling

点击翻译按钮获取中文摘要

The Sun drives almost all of the (neutral and charged particle) fluctuations in near Earth space or geospace. The energy from the Sun carried by solar wind is transferred to the geospace at the edge of the Earth's magnetic field or magnetosphere through an interaction process called magnetic reconnection. The geospace fluctuations caused by this energy transfer can adversely impact critical national infrastructure (e.g. satellites, power grids, etc.) and endanger the astronauts. However, we do not yet completely understand how the different solar wind conditions affect the amount of energy transferred into the magnetosphere. The fellowship will combine the magnetospheric expertise at the Los Alamos National Laboratory (LANL) with the data science/machine learning expertise of the PI at West Virginia University (WVU) to develop a new model for the geospace/magnetosphere that will help to improve our understanding of this energy transfer process. The project will provide a unique training opportunity for the PI and a graduate student at WVU through visits to LANL and foster long-term collaboration partnerships that will improve research and education in the state of West Virginia.The geospace environment is a highly complex system made up of regions that are closely coupled together. The regions undergo significant variations caused by energy input from the solar wind. The energy transfer occurs primarily through a process known as magnetic reconnection which fuels almost all of the physical and dynamical geospace processes. Understanding the solar wind-magnetosphere coupling is crucial to modeling and predicting the geospace variations that can threaten safety, infrastructure, and property on Earth, in the air, and in space. The goal of the proposed project is to gain new insights into the physics of solar wind-magnetosphere coupling. Simulation studies can provide invaluable insights into the process; however, the impact of existing models is limited by their considerable computational cost. The project will combine the magnetospheric expertise at LANL and PI's expertise in Physics-Informed Machine Learning to develop a reduced order model (ROM) that embeds the physics of the full physical model but significantly reduces the computational cost. The ROM will permit, for the first time, a comprehensive sensitivity and uncertainty analyses to identify the dominant solar wind drivers. Combined with systematic and comprehensive model-data comparisons, the ROM will reveal the paramount processes and may even discover overlooked or missing elements in the solar wind-magnetosphere coupling. The project will provide one graduate student valuable training and help develop a long-term partnership between WVU and LANL that will enhance research and education in the state of West Virginia and make it more competitive.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.