Announcements of Opportunity
SURF: Announcements of Opportunity
Below are Announcements of Opportunity posted by Caltech faculty and JPL technical staff for the SURF program. Each AO indicates whether or not it is open to non-Caltech students. If an AO is NOT open to non-Caltech students, please DO NOT contact the mentor. Announcements of Opportunity are posted as they are received. Please check back regularly for new AO submissions!
Remember: This is just one way that you can go about identifying a suitable project and/or mentor. Click here for more tips on finding a mentor. Announcements for external summer programs are listed here.
*Students applying for JPL projects should complete a SURF@JPL application instead of a "regular" SURF application.
*Students pursuing opportunities at JPL must be U.S. citizens or U.S. permanent residents.
Showing
Record
1 of
53
Next >>
Back To List
| Project: | A Machine Learning Approach to Surveying Solar Energetic Particle Time Profiles | ||||||||
| Disciplines: | Physics, Computer Science, Data Science, Mathematics | ||||||||
| Mentor: |
Allan Labrador,
Staff Scientist, (PMA),
labrador@srl.caltech.edu, |
||||||||
| Background: |
During the 11 year solar cycle, the Sun occasionally erupts in flares or coronal mass ejections which emit large amounts of Solar Energetic Particles (SEPs) composed of energetic electrons and ions. These particles can reflect a variety of physical processes, including ionization at the source region, acceleration processes, and transport through the heliosphere. Determination of ionic charge states are a vital clue in understanding these physical processes. The Advanced Composition Explorer (ACE) has been in space since 1997, observing the Sun and heliosphere. The Solar Isotope Spectrometer (SIS) aboard ACE precisely measures the abundances of elements from He through Ni, at energies ranging from 5 to 150 MeV/nucleon. When an SEP event is detected by SIS, the data include time profiles for all the elements (from He through Ni) and a variety of energies. The time profiles for SEP events includes an intense rise in particle counts, followed by a decay phase. The shapes of the time profiles vary greatly between SEP events, and the shapes contain a wealth of physical information. For example, although the SIS instrument does not measure ionic charge states directly, mean ionic charge states may be inferred for a given SEP event if there are extended time periods of well-behaved, exponential decay in particle intensity. We have been developing a machine learning (ML) approach to classify SEP decay phases as either exponential, power law, or irregular, and we are in the early stages of using regression analysis to extract time decay constants for these profiles. We have a machine learning ready data set of SEP events using 2017-2024 ACE/SIS data spanning elements from He to Fe and spanning a variety of energy ranges. This data set can be used to train Classifiers and Regressors, and the results will be compared to results of more traditional analysis. |
||||||||
| Description: |
The student will build on the existing codebase to use the data set (the 2017-2024 ACE/SIS data) to train machine learning algorithms to identify time profiles with useful exponential and power-law decays, using the time profile data directly, and from those time profiles, to identify and extract the exponential and power-law decay periods and time constants. The student will then generalize the approach in multiple ways: (1) Incorporate more data into the study (e.g. more data from ACE/STEREO; also include STEREO/HET and Parker Solar Probe EPI-Hi data), and (2) classify additional time profile types (e.g. multiple injections, passage of transient events, and other anomalies). This project will be co-mentored by Ashish Mahabal aam@astro.caltech.edu. |
||||||||
| References: |
ACE Science Center: https://izw1.caltech.edu/ACE/ASC/ Scikit-learn: https://scikit-learn.org/ Mean Ionic Charge States for the September 2017 SEP events: https://pos.sissa.it/358/1102/pdf |
||||||||
| Student Requirements: | Knowledge of Python or C/C++, and familiarity with machine learning concepts and/or common machine learning libraries. Familiarity with the physics of solar energetic particles or heliospheric physics are a plus but not required. | ||||||||
| Programs: |
This AO can be done under the following programs:
|
Showing Record 1 of 53 Next >> Back To List