For what aspects of cosmology research can a software engineer contribute?

Question

I'm a software engineer and I would love to learn more about cosmology by participating in this field. I'd like to change my career path closer to it and further from web programming. But I don't know how software engineers can contribute to the field of cosmology.

I'm now interested in finding a way to create a bridge between cosmology and software engineering in order to work in this field.

So far I have only thought of one position that can be helpful in this field, that of the data scientist.

Are there others?

Answer 1

You can regularly find Software Engineeer positions advertised on the AAS Job Register (check the bottom several sections). There are listings in various countries. With that kind of job, you would make lots of contacts and be able to branch out more into science if you like.

Another option (not a job) would be making substantial contributions to open-source science software projects, like Astropy as one example. It is collaborative work, so again you would make contacts to help you branch out from there.

Answer 2

I'm a software engineer and I would love to learn more about cosmology by participating in this field.

This seems like putting the cart before the horse to me.

I'd suggest you need to learn some basic physics and basic cosmology before trying to participate. It's not impossible to contribute without these things, but the more you know the more attractive you are to potential employers and the less confused you'll be by the material you'll have to deal with.

The problem with the way you phrased it in your question is that it sounds like you want to learn cosmology on the job, and that's probably not a practical option in a competitive scenario.

I'd like to change my career path closer to it and further from web programming

At the risk of stating the obvious, even science organizations need good websites. These help them communicate and educate, so maybe look at this is a potential way to contribute or at least enter such an organization.

You say "futher from programming" but you don't really say towards what goal. There's a world of difference between developing new IT skills that may be useful in science organizations and research compared to actually becoming good at the theory of cosmology and it's not clear what your end point is. You'll need to firm up your idea of "participation" to identify what you really want to be doing e.g. ten years from now and decide if it's realistically achievable.

I'm now interested in finding a way to create a bridge between cosmology and software engineering in order to work in this field.

Computers are used everywhere so again it's up to you to work out what you find most interesting in IT to serve your own needs in relation to being involved in cosmology. Is e.g. managing a data network going to fulfil your needs as long as it's a nework involved in cosmology in some way ? You need to identify what will fulfil your need here.

Answer 3

There is very much a scope of software engineering in cosmology and astronomy in general. I myself have transitioned from computer science to astronomy. My goals were different - I wanted to understand the science, so studying astrophysics and cosmology were more suited to me. This was interesting but also a long route and took a lot of effort. If you want to go that route, you'd need a degree in astronomy (e.g. a PhD) as others mentioned. But you maybe simply be interested in improving your software engineering skills, with applications in cosmology and other scientific domains. In that case, you would be in the niche of a science software engineer, which although not as well paid as other software engineering positions in the industry, is a very exciting job and is in high demand these days at research institutes.

You mentioned about data science. That's certainly a natural way to transition in astronomy (in fact its also a way for astronomers who can't find a job in the field to transition into a different field in the industry). But beware that when it comes to data science, astronomers (and scientists in general) are more interested in statistics/science side of things than just programming. So you may not get very far along just taking cool deep learning models and getting results if you cannot explain what's going on. This is based on my experience though, so take it with a grain of salt.

The easiest way to transition into the industry for you maybe to get in touch with researchers running computer simulations, like what @Justin T mentioned. Cosmological simulations are the backbone of research in the field, they are used extensively to propose new physics and explain new observations. From a software engineering point of view, designing a simulation to run efficiently is a huge challenge. You are running the code on hundreds if not thousands of nodes in parallel, and they must communicate effectively to produce accurate output. The field of high performance computing is very hot right now. See GADGET-4 for instance.

Lastly, as you start looking, you may find other ways to be involved in research. For e.g. I know that certain institutes use automated telescopes or other systems for research in cosmology or astronomy in general. Building models to efficiently run these systems - e.g. scheduling observations to optimize signal-to-noise ratio on the most important science goals, detecting clouds in images and avoiding observations during these periods, etc - are well respected contributions. As a software engineer, you are much better poised to solve these challenges than a physicist who self-teaches to do this from scratch.

Answer 4

tl;dr You’re a lot closer to being able to do this than you might think, but you should consider a PhD if you want to do research heavy work.

I think at the risk of repeating what has already been said, it has been my personal experience that computer scientists are essential collaborators in some of the biggest undertakings, with often times having very little physics background.

For context, I’ve been doing computational physics research for some years now with relativistic modeling, and there have been almost as many computer scientists as physicists that I’ve run into (regarding large projects) because at the end of the day, a physicist will be better at physics and a computer scientist will be better at computer science, and you need both in large amounts on these projects.

My advisor has been working on a numerical code for gravitational wave modeling, and the group of people working in the project are just two physicists and two computer scientists, because the code is absolutely massive, with so many moving parts, and while my advisor is quite good at coding, at the end of the day a lot of physicists end up being self taught as far as coding goes (or at the very least, with very little formal instruction) and so the perspectives of computer scientists become so valuable in these contexts.

I also have a friend who’s background is entirely in mathematical computing who just got into his PhD with an advisor doing numerical relativity as well, with little to no physics background.

At the end of the day, while all of these experiences are super anecdotal, I think if you’re serious about wanting to change your career to this, you should consider pursuing a PhD, because while a lack of physics background might not disqualify you from a (CS related) position, a lack of a PhD might (big emphasis on might) based on the exact nature of the job you’re seeking. The closer you want to be on the front line of doing cool research, the more likely you’ll need a PhD.