REU - Research Projects for Summer 2025 | Institute for Space Weather Sciences

Focus Area 1: Solar Astrophysics

Project #1.1: Feature Identification of Solar Prominences

Primary Mentor: Dr. Vasyl Yurchyshyn
Co-Mentor: Dr. Xu Yang
Type of Project: Image processing and Data Analysis
Project Description: We propose to use an automatic routine to detect solar prominences seen off the limb in BBSO Halpha and SDO/AIA images. Solar prominences are protrusions above the solar limb best visible in hydrogen alpha line (Ha, 6563A) and ultraviolet images at 304A. The importance of prominences lies in the fact that due to loss of stability they erupt, carrying a significant amount of plasma and magnetic field out into interplanetary space. These eruptions impact the heliosphere in general and the Earth's magnetosphere in particular. Over the past years the solar community has acquired large amounts of data on solar prominences. The current NASA SDO mission generates large quantities of excellent multi-temperature prominence data. We propose to perform a statistical study of solar prominences and a comparative study of the prominence statistics detected during the past and the current solar cycles. In this study the student will apply the prominence identification software to the full disk H-alpha data stored in BBSO archives as well as to full disk AIA 304A images. The software will need to be modified in order to be applicable to newer data sets. After that the student will perform statistical analysis of the generated data in order to reveal temporal and spatial distributions of the prominence parameters over solar cycles.
Expected Outcomes: The students will learn the basics of data processing and feature identification. The data processing will be done either using IDL and/or Python programming language. Also, the student will learn the basics of statistical analysis and data visualization. Moreover they will be introduced to solar observations and learn how solar data are acquired with the 1.6 meter Goode Solar Telescope and the full disk H-alpha telescope.

Project #1.2: Exploring pre-erupting configuration of magnetic fields in solar active regions

Primary Mentor: Dr. Vasyl Yurchyshyn
Co-Mentor: Dr. Xu Yang
Type of Project: Image processing and Data Analysis
Project Description: Coronal Mass Ejections (CMEs) have long been identified as a prime cause of large, non-recurrent geomagnetic storms. They often have a magnetic flux rope (MFR) embedded in them, meaning that they harbor a large-scale, coherent magnetic structure with a large amount of twist. One of the goals of solar physics research is to comprehend the evolution of an AR toward an eruption. From a practical standpoint, it is desirable to know the location and timing of the next eruption. We propose a novel approach to solving this long standing problem. The main objective is to detect and parameterize a MFR in extrapolated coronal fields prior to a flare or an eruption. Coronal field extrapolations will be performed using Fleishman et al. (2017) tool that exploits the optimization and weight function methods. The numerical realization of this approach is part of the GX Simulator package, which is freely available from the SolarSoft IDL library. The boundary conditions for the extrapolations will be vector magnetograms provided by the Helioseismic and Magnetic Imager on board Solar Dynamics Observatory. Using extrapolated data cubes we will fine tune our methodology for detecting signatures of strong current systems in the AR corona. We will use existing Q factor package and develop new tools and approaches to quantifying evolution of magnetic configurations and development of MFR that lead to eruptions.
Expected Outcomes: The students will learn the basics of data processing and feature identification. The data processing will be done either using IDL and/or Python programming language. Also, the student will learn the basics of statistical analysis and data visualization. Moreover they will be introduced to solar observations and learn how solar data are acquired with the 1.6 meter Goode Solar Telescope and the full disk H-alpha telescope.

Project #1.3: Investigation of Mini-filament Eruptions and Their Relationship with Small Scale Magnetic Flux Ropes in Solar Wind

Primary Mentor: Prof. Haimin Wang
Co-Mentor: Dr. Nengyi Huang
Type of Project: Data Analysis
Project Description: In recent years, small-scale MFRs (SMFRs) in both solar surface and solar wind are receiving significant attention. Initial evidence shows that they are numerous and ubiquitous. Studies of SMFRs are advanced significantly with high resolution observations from the 1.6m Goode Solar Telescope (GST) of Big Bear Solar Observatory (BBSO) on the solar surface, and solar wind observations from Parker Solar Probe (PSP). Students will carry out comprehensive case and statistical studies of SMFRs in solar surface and solar wind.
Expected Outcomes:
Students will investigate the kinematic, thermal and magnetic properties of them, as well as possible photospheric magnetic field evolution associated with eruption of mini-filaments. In addition, they will anticipate to find possible connection between solar mini-filament eruptions and detected SMFRs in solar wind.
Preferred Skill: Took courses in college physics, and some programming skill

Project #1.5: Investigating Subsurface Flow Patterns Preceding the Emergence of Active Regions

Primary Mentor: Dr. John Stefan
Co-Mentor: Prof. Alexander Kosovichev
Type of Project: Data Analysis (possibly with Machine Learning methods)
Project Description: It is well known that active regions---areas of strong magnetic field that appear to the naked eye as sunspots---produce the majority of solar flares and eruptions. However, the specifics of how these active regions form remains unclear. Additionally, prediction of where and when these regions will develop has become increasingly important given the growing human presence in Earth orbit and beyond. In a similar way that monitoring tropical depressions on Earth gives us advance warning about where hurricanes are likely to develop, monitoring the subsurface flows of plasma within the Sun may allow us to provide advance warning about where active regions will form. To this end, the student will analyze maps of subsurface flows generated from the data provided by the Helioseismic and Magnetic Imager (HMI) aboard the Solar Dynamics Observatory (SDO) spacecraft. These flow maps, as well as their physical characteristics such as vorticity and divergence, will be compared and correlated with the surface magnetic field to search for signatures that precede active region emergence. Depending on the student's comfortability with programming and statistics, there is the potential to apply Machine Learning (ML) methods to train a Convolutional Neural Network (CNN) to predict the likelihood of active region emergence based on the input flow maps.
Expected Outcomes: At the completion of this project, the student will have analyzed the evolution of many active regions to identify relevant flow characteristics that precede the regions' emergence. The student will gain experience in Python programming as well as working with science-grade observational data. Throughout the course of the project, the student will learn about how the Sun's magnetic field is generated and evolves (basic Dynamo theory) as well as a casual understanding of how electrically-conducting fluids behave (magnetohydrodynamics, or MHD, theory). There is potential for the student to also begin working with ML methods, which have broad applications beyond heliophysics.

Project #1.6: Small-scale Ejections and Eruptions from the Solar Chromosphere

Primary Mentor: Dr. Jeongwoo Lee
Co-Mentor: Dr. Aabha Monga
Type of Project: Data Analysis
Project Description: Small-scale jet-like ejections are ubiquitous in the solar atmosphere, and are believed to play an important role in the energy balance and structuring of the corona as well as the mass transport in the solar wind. We generally surmise that these events may be associated with small-scale magnetic reconnection, but alternative mechanisms are also compelling. Excellent quality data of imaging and spectroscopy are gathered with on-going NASA missions, Parker Solar Prove (PSP), Interface Region Imaging Spectrograph (IRIS), Solar Dynamics Observatory (SDO), and Solar Orbiter in international collaboration with ESA. Observational analysis combined with modeling of such small-scale jet-like ejections is essential in order to advance our understanding of their physical nature and role in coronal heating and mass loading of the solar wind, which are aligned with recent NASA mission goals.
Expected Outcomes: Students will mostly work on finding suitable events using a web-based tool called the Helioviewer to look at EUV images of the Sun taken by several space telescopes in various wavelengths over selected observing times of interest. They will receive training on retrieving, visualizing, and processing high-resolution EUV images and spectra using IDL-based solar software (SSW) tools as well as a Python-based software package for solar physics (SunPy). Students will apply basic physics knowledge to understand and interpret these solar images made at multiple wavelengths.

Project #1.7: Investigation of Quasi-Periodic Pulsations in Long-Duration Solar Eruptive Flares

Primary Mentor: Dr. Sijie Yu
Co-Mentor: Dr. Xingyao Chen
Project Type: Data Analysis
Project Description: Quasi-periodic pulsations (QPPs) are commonly observed in microwave and X-ray emissions from solar eruptive flares associated with coronal mass ejections (CMEs). This project aims to investigate the relationship between QPPs and the morphology of CMEs in long-duration solar flares. The student will analyze observational data from the Expanded Owens Valley Solar Array (EOVSA) and Geostationary Operational Environmental Satellite (GOES) (and Solar Orbiter/The Spectrometer Telescope for Imaging X-rays (STIX), if applicable) to identify QPPs in microwave and X-ray emissions. The morphology of the associated CMEs will be examined using multi-wavelength imaging data from a suite of NASA spacecraft—such as the Solar Dynamics Observatory (SDO) and GOES-R—and white-light data from the Solar and Heliospheric Observatory (SOHO) coronagraph. The student will perform comprehensive case studies and statistical analyses of QPPs and CMEs in long-duration solar flares.
Expected Outcomes: The student will gain an introductory understanding of solar physics—particularly solar flares and coronal mass ejections (CMEs)—while developing practical data processing and analysis skills using Python-based open-source software for modern solar observations. Working closely with the mentors, the student will deepen their knowledge of solar eruptive flares and quasi-periodic pulsations (QPPs). Ultimately, this project will contribute to understanding the relationship between QPPs and CME characteristics in long-duration solar flares.
Preferred Skills: Some background knowledge in college-level physics and some Python programming experience is preferred.

Focus Area 2: Terrestrial Physics

Project #2.1: Investigating ionospheric activities during solar eclipse using radio observations

Primary Mentor: Dr. Surajit Mondal
Co-Mentor: Prof. Lindsay Goodwin
Type of project: Data analysis
Project description: During astronomical observations, radio waves pass through the Earth’s ionosphere (a layer of the atmosphere that contains a high concentration of plasma), before they are incident on a radio telescope. During this process, the ionosphere leaves its imprint on the radio waves, which can manifest in astronomical images. One of the prominent ionospheric effects seen in these images is the apparent shift of astronomical sources from their fiducial positions. This is related to plasma density gradients in the ionosphere altering radio wave propagation. These effects need to be corrected before these images can be used for astronomical studies, however, these corrections also serve as a rich dataset that can be used for granular studies of the ionosphere. In this project, we will use low-frequency radio observations from the Owens Valley Radio Observatory’s Long Wavelength Array (OVRO-LWA), taken during the October 2023 eclipse,to generate a new ionospheric dataset. The student will initially test and improve already existing shift-correction strategies using these data. After that, the student will work on deriving some ionospheric parameters using the developed method and compare them using coincident ground-based GPS receiver observations.
Expected outcomes: 1) Gain a basic understanding of radio wave propagation through the ionosphere and its effects on radio images; 2) Gain a basic understanding of ionospheric dynamics; 3) Gain image processing and data analysis skills.

Project #2.2: Short wavelength infrared observations of the nightglow emissions

Primary Mentor: Dr. John Meriwether
Co-Mentor: Prof. Andy Gerrard
Type of project: data analysis and screening of observations
Project description: Examine SWIR imager observations obtained at the Jiffer Observatory at Jenny Jump and generate data products (keograms) for several wavelengths. Also help with the installation and operations of the MaxFPI instrument that is expected to be installed this spring.
Expected outcomes: One or two papers should be forthcoming.

Project #2.3: Electrostatic Waves in the Earth's Magnetosheath

Primary Mentor: Dr. Ilya Kuzichev
Type of project: Data Analysis
Project description: Recent electric field measurements in the Earth's magnetosheath have demonstrated that the electron scale waves are dominated by electrostatic solitary waves, such as electron holes. These waves can efficiently interact with electrons and ions, affecting local plasma dynamics. In this project, we propose to collect a large dataset of electrostatic waves observed aboard the Magnetospheric Multiscale (MMS) Mission and perform a statistical investigation of the wave properties.
Expected outcomes: A large statistical survey of electrostatic solitary waves in the Earth's magnetosheath. A paper presenting the corresponding results.

Focus Area 3: Data Science in Space Weather

Project #3.1: An Interpretable Machine Learning Approach to Understanding the Relationships between Solar Flares and Source Active Regions

Primary Mentor: Prof. Jason Wang
Co-mentor: Prof. Haimin Wang
Type of project: data analysis
Project description: Solar flares are explosive events on the Sun’s surface, posing significant risks to Earth-based technologies. Accurate prediction of these flares is crucial for mitigating potential disruptions. This research focuses on the binary classification task of predicting whether an active region (AR) will produce a ≥C-class flare or a ≥M-class flare. We apply various machine learning models, including Random Forest (RF), Support Vector Machines (SVMs), XGBoost, and neural networks, to analyze observational data from SolarMonitor.org and the XRT flare database (2011-2021). The models will be trained and evaluated on a dataset of ten solar parameters associated with each AR. We will employ SHAP (SHapley Additive exPlanations) in conjunction with LIME, ALE, and PDP to enhance model interpretability and assess feature importance for this classification task. Furthermore, we aim to explore time series characteristics by integrating data from the NJIT Space Weather Benchmark Dataset. Finally, the potential of Large Language Models (LLMs), particularly TimeLLM, for time series-based forecasting of flare occurrence will be investigated.
Expected outcomes: The primary outcome will be a comparative performance analysis of the studied models for the binary classification of ARs. We seek to identify the physical features of a source AR that significantly influence the potential of the AR to trigger ≥C-class or ≥ M-class flares.

Project #3.2: H-alpha Image Super Resolution at BBSO with Advanced Deep Learning

Primary Mentors: Prof. Bo Shen
Co-Mentor: Dr. Qin Li, Chenyang Li
Type of Project: Data Analysis, Machine Learning, Software Development
Project Description: The Halpha Image Super-Resolution project aims to enhance the spatial resolution of chromospheric observations using advanced deep learning techniques. By training neural networks on pairs of full disk and high-resolution Halpha data from the Goode Solar Telescope (GST), this initiative will reconstruct fine-scale features such as fibrils and flare dynamics with unprecedented clarity. This transformative approach leverages the power of convolutional neural networks to bridge the gap between lower-resolution full-disk observations and the fine details captured in GST's high-resolution targeted observations. The resulting high-resolution dataset will enable solar physicists to study chromospheric dynamics across multiple spatial scales simultaneously, providing crucial insights into energy transport mechanisms in the solar atmosphere and potentially improving space weather forecasting capabilities.
Expected Outcomes: We will provide students with hands-on experience applying cutting-edge deep learning techniques to real data. Students will develop valuable skills in image processing, neural network architecture design, and scientific programming while contributing to meaningful solar physics research. By the project's completion, participants will have created a working super-resolution pipeline that enhances the clarity of chromospheric features, authored scientific documentation, presented their findings to the research community, and potentially contributed to a peer-reviewed publication.

Project #3.3: Detection of Solar Radio Bursts with Machine Learning

Primary Mentor: Dr. Peijin Zhang,
Co-mentor: Anastasia Kuske, Prof. Bin Chen and Mengjia Xu
Type of project: AI-ML, solar physics, radio astronomy
Project description: This project focuses on designing an event trigger and detection method for solar radio bursts targeting the observation data from the Owens Valley Radio Observatory Long Wavelength Array (OVRO-LWA). Solar radio bursts serve as key indicators of solar activity and space weather, which can impact satellite operations, communication systems, and power grids. With OVRO-LWA having collected a big dataset, it is compelling to have an efficient, trigger-based system for identifying and processing significant events. This will enhance both scientific research and future space weather monitoring efforts. Students will gain knowledge of solar physics and solar radio bursts and will leverage cutting-edge machine learning techniques, including state-of-the-art object detection algorithms, integrated with physics-based models to develop a robust tool for detecting radio burst events. The project involves training the tool on historical OVRO-LWA data, performing statistical analyses of detected events, and adapting it for real-time data processing to enable automated event triggering. This work will contribute to optimizing the OVRO-LWA’s data pipeline and advancing our understanding of solar phenomena.
Expected outcomes: 1) Gain knowledge of solar physics and solar radio bursts; 2) A deeper understanding of ML methods and their implementation; 3) Contribute to the development of a machine learning-based radio burst event detector
Preferred qualifications: Basic knowledge of Python and programming.

Focus Area 4: Astronomical Instrumentation and Observational Techniques

Project #4.1: Installation and First-light Observations of an He I 10830 Synoptic Telescope

Primary Mentor: Prof. Wenda Cao (BBSO/NJIT)
Co-Mentor: Prof. Deqing Ren (CSUN)
Type of Project: Instrument Development and Data Processing
Project Description: A synoptic solar telescope has been developed to provide high-speed and high-sensitivity polarimetry measurement with a spectral line at Helium 10830 Å, under a joint effort between CSUN (California State University, Northridge) and NJIT. The telescope is a standard Celestron 11-inch EDGEHD that offers a clear aperture of 355 mm and a field of view of 320" by 256". Its polarimeter consists of a tunable Fabry-Perot Interferometer, a high-speed Liquid Crystal Variable Retarder (LCVR)-based modulator, and a NIR camera. As of now, all the hardware development and system assembly has been completed in a campus observatory at CSUN. After initial system characterization and trial observation tests on CSUN campus, this telescope will be relocated in Big Bear Solar Observatory (BBSO) of NJIT this summer for first-light observations and scientific data verification. The student will participate in the He I 10830 synoptic telescope installation and operation, first-light observations, and scientific data processing and calibration.
Expected Outcomes: The student will gain hands-on experience in astronomical instrument development, observations, and data analysis.
Preferred Skills: Some background knowledge in college-level physics/astronomy and some LabVIEW, MATLAB and/or Python programming experience are preferred.