How to enter theoretical physics with astronomy background?

[Brian Leist]

Hi all,

I am interested in advancing my education in theoretical physics, specifically topics concerning quantum field theories and quantum gravity. However, I have a somewhat rocky background as an early undergrad and while my Master's degree was achieved with great success in the classroom and in research, I am not an appealing candidate on paper.

I have a bunch of the background posted below, and I'm sorry if this amount is not appreciated, but I include it to better paint a picture of my situation. To cut to the advice that I am looking for is in how to find opportunities (from most preferable to least) to

1. Find a PhD program in theoretical physics which includes a stipend (anywhere on the globe).
2. Find a PhD program in theoretical physics in which I could find a job in the university or something.
3. Find a Masters program, where although I already have a Msc Phyiscs, I could prove my chops in
theoretical physics and start networking in that community.
4. Find a job at a university which had a theoretical physics group that would be open to me hanging
around. I would be satisfied running the labs and learning what I could from others.
5. Is anyone else, now or have been, in a similar situation?

It should also be known that I'm:
a. willing as well as eager to study abroad
b. willing to work for very little money if I can advance my education
c. I am planning to use this summer to study for and take the physics GRE

I spent the first years of college a total mess. When I was later into the physics program, my professors couldn't believe my early transcript. I was unsure of what I wanted to do, unprepared for college, and spending most of my time outside of the classroom working to pay for tuition. I was also better at working than I was a student at the time. Eventually I decided to take a few physics courses, as I'd always had an interest and read any popular book on the subject I could but just assumed that was a topic for better students than I. But if I was going to go out of University, I might as well take a few physics courses.

I wasn't immediately a hit student from that point on. I still had to figure out how to drop the number of hours I had to work out of the classroom and sand off much of the mathematical rust. However, I improved each semester, got in with a research group in astronomy, earned a few grants and scholarships to supplement working, and by the end of my undergraduate degree I was among the top performers in the classroom. I proved my capabilities to myself and to my own university, at least, and earned a full ride to the graduate program as a GTA. Those years in late undergraduate and graduate were the best of my life, so far, but I'm still hungry for more.

Now onto issue number two. My research has been in astrophysics and has been moderately successful and was a hit as far as a Master's Thesis goes. However, it has not produced any publishable results, at least not as of yet. It was a very time consuming project, other team members were in and out with personal issues, my advisor set me to work in python and couldn't keep up with what was happening with the code, and so on. I've since taken a job as an high school AP Physics, AP Calculus and Astronomy teacher. I had hopes of continuing the research to get it to a publishable state, but that is a laughable thought now. I've made improvements here and there but without the time it won't get there until next year. So I won't really have that to show off for a while.

Not only that, but most of my education has been in physics and not astronomy, my interests are more in physics and not astronomy, and I would like to advance in theoretical physics rather than the sort of work my research has thus far been in. Frankly, my research has involved more programming challenges than physics and math.

On the positive side, I would have great recommendations coming from my professors (especially in EM, Theoretical Mech and QFT) as well as my current advisor.

If you've read this long post, thank you for your time. Your advice, whether it directly applies to the numbered list above or is a perspective I haven't thought about, is greatly appreciated.

 

[Vanadium 50]

If you can 't develop publishable work as a grad student in your field, why do you think you can do this in a different field? That might sound harsh, but that's the question that admissions committees will ask. You need a good answer to that, and you need to be able to convey this clearly. You also need to understand "I really really want to do this" is something that does not distinguish between applicants.

 

[Brian Leist]

If you can 't develop publishable work as a grad student in your field, why do you think you can do this in a different field? That might sound harsh, but that's the question that admissions committees will ask. You need a good answer to that, and you need to be able to convey this clearly.

Of course, this is a very reasonable question and certainly requires more of an explanation than this forum, but I'll try to convey the gist of it. Apologies in advance and for the length this post is taking.

I had been given a project for my Master's Thesis and I completed that project to the satisfaction of my advisor, others in the university and in the field, and beyond. However, this does not make the project worthy of publication in a scientific journal, though some were pushing for that after "completion". We may put it in a journal of a more computational nature.

In the field of galaxy formation and evolution, there exists large catalogs of galaxies in various regions, or fields, of the sky. Though I've worked on other fields, my field of focus was the COSMOS field (Scoville et al. 2007). The typical routine for generating a catalog is to use the software SExtractor (Bertin, E. & Arnouts, S. 1996)...for a fun aside, I defended my thesis in front of the one that wrote the more popular reference manual for this software and he actually pushed to at least try to publish it...This software creates a catalog from a single image. Then one would take a series of these SExtracter searches from a few dozen images across the spectrum to get a full multi- wavelength data set. The hop-and-a-skip way of generating a catalog is to use SExtractor in "dual image mode" which uses only a single image as detection. This, however, leaves out many other objects that may not be bright enough for detection at that wavelength. There are other methods involving cross-correlation of positions. Our method focused on the latter. There are issues one runs into with each, many of which involve the challenges with dealing with an inhomogenous data set.

My project was to created a Python package which automated the generation of such "master" catalogs. This process was typically done via a by hand sort of way, and adding to catalogs later could also be a process. To this end, I have a package which does so. It generates the catalog in a much shorter time (the method used by the group at the time took a weekend to add a single catalog and involved editing scripts by hand in IDL). My code generates an entire catalog in a few hours using a single configuration file. It's even faster to add and access the data, which is now stored in a FITS file. I also worked on testing various software packages available for generating photometric redshift estimations. I also explored with incorporating our own software for this and for performing fixed aperture photetry. The latter two things were not part of the thesis.

The reason that this work is not publishable is because it does not yet push the frontier of science. Our catalog does add a hundred thousand or so extra objects to the existing catalogs generated in the COSMOS field, but there are other issues here. Other similar catalogs in the literature was created by large teams with far more resources and experience than we had. Not only that, but I was the only one working on the project. My advisor could only do research on his own time, as he was with the University with another position (he and his wife were a part of a two body problem). These other groups applied more techniques that were more precise than ours in certain aspects. My advisor wanted to pursue some of these at a later time or did not feel they were necessary, to my objection as well as a referee down the road.

This project was a fine one and I learned much from it. However, to build a catalog generation scheme from scratch that is competitive with current methods as well as producing the benefits of our method was a bite a little larger than a Master's thesis timeline, especially with our resources. In retrospect, I would have rather worked with existing catalogs, as many many do, and pushed on the scientific frontier from there. I would have certainly spent a lot more time learning astrophysics. This is still only part of the picture and there is much more that I could get into that this isn't the correct forum for.

In any case, for what it is worth, I have a Msc Physics degree in excellent standing, research and programming experience which led to a successful Master's Thesis, and professors and advisors that support advancement. It isn't simply that I was incapable of "develop[ing] publishable work as a grad student in [my] field", it is slightly more complicated than that. I know it is difficult to leave the field I am in for another, but my interests are in another direction.

You also need to understand "I really really want to do this" is something that does not distinguish between applicants.

Of course. In no way do I mean to imply that this post qualifies me as anything but a whiny little brat. I'm simply casting out lines for perspective and advice. I understand all too well the position I am in. I also hope that the tone in which I write does not seem optimistic or entitled. I hope it rather has a taste of desperation to it as that is the state in which I write it.

I am simply tired of being the only one in my city interested in studying quantum field theory. The best part of my day is when I've finished my chores and can snag an hour or two working through Zee's "In a Nutshell" series on QFT, group theory, and gravitation. How great are those books, by the way?! I am simply posting hoping that I can find an opportunity to place myself in an environment, or adjacent to one, where others are studying the same thing. Perhaps, even, give myself an open door to proving to someone else that I could earn a PhD in the subject.

Perhaps I should not have included (1) or (2) on the list. At least, I should have made it clear the laughable optimism required to include them. I'm just in search for possibilities that are currently in my view.

Thank you.