Computational and Theoretical Astrophysics

IF I was to give myself a professional tag, I'd say I'm a computational astrophysicist with mild theoretical tendencies, working on the field of galaxy formation and evolution, somewhere at the interface between observations and simulations. If you are not confused yet enough, allow me to give you some context.

One of the greatest challenges of the human mind is to understand how the Universe evolved from a fairly uniform state shortly after its 'birth' (big bang), to the highly structured entity it is today. This curiosity has lead to the development of two rapidly growing research disciplines, Cosmology, which studies the Universe as a whole, and of Astrophysics, to study individual astronomical systems (planets, stars, galaxies, ...). Within these disciplines, three different, but complementary, approaches are followed to tackle the aforementioned challenge. On the one hand, astronomers (or 'observers') collect data about the Universe with high-tech instruments such as telescopes (optical, X-ray, ultra-violet, etc), radio antennas, and other kind of detectors. On the other, theorists using basically pencil and paper make an attempt to interpret the observations from first principles, this is, using the fundamental laws of physics. Finally, when the analytic methods of a theorist become no longer feasible due to the complexity of a system, computational astrophysicists (or 'simulators') create artificial universes and generate data using computers. Observers characterise astrophysical systems by measuring their properties. Theorists try to makes sense of these numbers with analytic models. Simulators try to recreate astrophysical systems and their observed properties using numerical models. All three face difficulties in accomplishing their respective tasks, and either can do without the other. Needles to say, it is not uncommmon to find two, or even all three, of these approaches embodied in a single individual

The task of a simulator could be stated as follows: To provide a working model of a given astrophysical system (which could be the entire Universe!). It can be broken down into to the following steps: 0) identify (isolate) the system to make it simple enough; 1) set a model that, ideally, captures all the relevant physics (gravity, thermo- and hydrodynamics, etc.) of the system; 2) translate (i.e. code) this model into a computer program; 3) use this program to perform calculations, and compare the outcome to the observed properties of the system (that's what the picture above is all about). These steps constitute a loop which is completed when, in the likely case that the model fails to agree with reality, it has to be modified (goto step 1, revisit, and redo).

Each of the steps above implies a lot of work. Usually, the system of choice is a matter of professional 'taste'. The fundamental physical laws are generally in place, and which ones to include is dictated mostly by the overarching principle of 'keep things as simple as possible'. Thus, steps 0 and 1 are the easy ones. Translating a theoretical model into a working computer program that actually performs the desired calculations in a reasonable amount of time is a whole different story. This has required the development of powerful computers and sophisticated numerical methods. Also, efficient handling of a large volume of simulation data and the comparison to observations demands the existence of analysis and visualisation tools specifically tailored for the problem at hand.

I'm interested in particular in the problem of how galaxies have come to be the way they are. Although I'm sure I'd have enjoyed being an observer, I chose to attack this problem with theory and numerical models, thus becoming a simulator.