Astrophysics is the study of extra-terrestrial (not Earth bound) physical systems. Since the entire Universe (minus the Earth and its occupants) is extra-terrestrial, astrophysics covers an enormous range of topics, including everything from the tiniest grains of dust to entire galaxies to the large scale structure and composition of the Universe itself. Unlike other physicists, astrophysicists are unable to manipulate objects under study. For example, it's quite difficult to bring a galaxy into the lab, pick it apart, and see what makes it tick. Almost everything that we know about the Universe comes from capturing light from distant objects so we have to become experts at collecting and analyzing light. To gather light from space, we build ever larger and more sophisticated telescopes. We even put some of our telescopes in space to avoid the undesirable effects of the Earth's atmosphere.
To analyze light, we study the interactions between light and matter. By understanding those interactions, we can reconstruct the physical environments encountered by the light received by our telescopes.
Stars form from collapsing clouds of interstellar gas and dust. As the cloud collapses, angular momentum is conserved causing the cloud to spin faster. As the density increases, a star is formed in the center and the remaining material flattens into a very dusty circumstellar disk. It is from the dust in this disk that planets will eventually form, and it is thus called a proto-planetary disk. Understanding the detailed composition, geometry, and dynamics of these proto-planetary disks is critical to understanding the process of planetary formation. We use a variety of different computational models to understand the composition and geometry of proto-planetary disks.
Data mining in astronomical archives.
With the advent of space telescopes, astrophysics has entered a data-rich era. The enormous amount and quality of the available data has stimulated the creation of computational models of ever increasing complexity. Data analysis is complicated by both by the volume of data and the number and complexity of the available models. Using techniques from computer science and mathematics, we are finding ways to compare the various models, analyze large quantities of data in an automatic way, and search for previously unseen patterns in large data sets.