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.
<< Prev Record 41 of 59 Next >> Back To List
|Project:||Multiple Projects in Galaxy Evolution and Data Science|
|Disciplines:||Astronomy, Computer Science, Physics|
|Mentor URL:||https://dawn.nbi.ku.dk/events/surfdawn (opens in new window)|
NOTE: This project is being offered by a Caltech postdoc alum and is open only to Caltech students. The project will take place at the University of Copenhagen in Copenhagen, Denmark.
This is one of several projects available at the Niels Bohr Institute this summer, and we expect that in total 6-10 students (some from Caltech and some from elsewhere) will come to Copenhagen during our seventh year running a summer program. Since travel within Europe is inexpensive, this will be an 11-week program, so that students can take a 1 week vacation the week of August 7 to see other parts of Europe. Other projects in astronomy with different mentors will also be available. We hope to finalize who will be coming by mid-January, so that there will be plenty of time to both write a SURF proposal and take care of any necessary visa/housing.
A range of projects related to early-Universe galaxy formation and evolution are available, ranging from observational to computational depending upon your background and interest.
Much of the recent focus of the group has been on improved techniques for inferring the properties of high-redshift galaxies. Analysis of high-redshift galaxies can often only be done from very limited data, in many cases using only the colors of galaxies in very broad filters (called photometry). Everything we learn about these galaxies therefore comes from fitting models derived from local galaxies to photometric data. However, there are several good reasons to think that the first galaxies to form in the Universe actually aren’t exactly like local galaxies, and therefore we would need to use different models.
Over the past few years, we have been developing models for the ways in which the first galaxies might differ from local ones, particularly in terms of the temperatures in star-forming regions. The inferred properties are highly sensitive to the temperature assumed, and getting this wrong might mean incorrectly estimating, e.g., the stellar mass by as much as a factor of 100! Indeed, nearly everything we thought we have learned about distant galaxies is potentially altered with improved assumptions.
Much of this work has involved summer students over the past few years, including seven publications and several others in preparation.
Possible projects related to this work might include:
1) We are now looking to apply this technique to galaxies where there is not enough information to actually determine which temperatures to use. Perhaps the most promising approach is to look for a relationship between properties of gas, which are very difficult to measure, and properties dust, which are easier because of strong infrared emission. The goal would then be to use dust as a proxy for gas in order to figure out which temperature to use in order to infer the correct properties. This could then be applied to a wide range of galaxies of various ages and types for which reliable estimates were previously not possible.
2) Another approach for galaxies with insufficient information to constrain gas temperatures would be to find similar galaxies for which there is enough information to determine gas temperature, then assume that temperature. Such an approach would potentially allow us to get accurate galaxy properties for the first time from deep surveys with Hubble and the James Webb Space Telescope (JWST) which are wonderful for detecting faint objects but which do not provide enough information to measure gas temperatures.
3) A similar approach could be applied to finding the very highest-redshift galaxies detected by JWST. In the first few months of use, a large number of ultra-high-redshift galaxy candidates have been found, but all rely on very limited information and in many cases, followup work has shown that they are much less distant than previously claimed. In addition to the limited information, this is because current models are making physically-implausible assumptions about what high-redshift galaxies should actually look like. This becomes a sharp problem because it is easy to confuse the similar spectral energy distributions of an ultra-high-redshift galaxy with a Lyman break and a far-more common galaxy at a more moderate redshift with a Balmer break when only one of the two breaks is seen by JWST.
With the improved relationship between gas, dust, and stellar populations that our group has been developing, we have more accurate model spectral energy distributions, and thus should be able to separate these two cases more effectively than in previously work. We would go through the first-year JWST archive in order to systematically search for more robust ultra-high-redshift galaxy candidates available.
4) These claimed ultra-high-redshift detections are also remarkably massive given how young the Universe is at the time their light was emitted. Indeed, if correct, they would likely disprove our current consensus cosmological model (see Labbe et al. 2022; Boylan-Kolchin et al. 2022). However, by correcting the gas temperatures and resulting stellar populations, these masses come down by nearly a factor of 100. The goal of this project would be to take the most robust high-redshift detections available by this summer, re-fit them with improved models, and then evaluate whether there is still tension between their masses and the standard cosmological paradigm.
5) We recently found that novel machine learning techniques allowed us to categorize gamma-ray bursts far more efficiently than was previously possible. We are now looking at other places where these methods might be applicable, not just in astronomy but also with groups at the Niels Bohr Institute in other fields. This project would be well suited for a computer science student, as the creative aspects are almost entirely computer science rather than astronomy (or biology). It also may be a good fit for a student with a strong mathematical background.
One additional possible aspect of this problem would center around trying to find a formalism in which dimensionality reduction methods can be proven to be optimal. Establishing optimality for noisy or missing data is often not part of research in machine learning, which tends to focus on more idealized benchmark datasets.
6) It is difficult to probe the assembly mechanisms and history of galaxy clusters because we can only see the luminous parts of a galaxy, but most of the mass is instead dark matter, which as the name implies does not give off light. Filamentary structures are thought to be particularly important to the assembly of galaxy clusters and groups as part of the “cosmic web”. A recently-developed learning technique based on the behavior of a species of slime mold might allow us to infer the presence of these filamentary structures. We would use this to search for early cluster formation in a new, 20 sq. deg. survey called the Cosmic Dawn Survey.
|Student Requirements:||Variable, depending upon the project, but some computational background is strongly recommended. Some projects will be suitable for frosh, and others will require a more formal astronomy, physics, computer science, or even math background.|
This AO can be done under the following programs:
<< Prev Record 41 of 59 Next >> Back To List