I’m on a bus to Sydney for a mini-meeting on wide-field astronomy done by a Chinese collaboration from Dome A in Antarctica. The bus doesn’t have Internet, but in about 15 minutes my phone will, so I’m keeping this quick and simple.
On Thursday we had a very well-delivered talk from Phil Hopkins (currently UC Berkeley) on the subject of galaxy formation. It was so well laid out that even I could understand most of it, and remember stuff afterwards. The subject was the effect of “feedback” on galaxy formation.
Background: The naive expectation is that when you gather a lot of gas together in one place, it’ll start to collapse under its own gravity and condense to form stars. While that is true, the process is much less efficient than the naive expectation would lead us to believe, on two counts: First, we observe that only about 2% of the gas in molecular clouds actually forms stars. Second, even accounting for this fact, there are fewer really massive galaxies — in fact, fewer galaxies at all mass scales — than we would naively expect. “Feedback” is the name given to a set of physical processes which act to regulate the rate at which stars are formed, so that galaxies don’t use up all their gas at once.
Phil went on to explain what “feedback” really meant here, dividing the sources into “stellar feedback”, which happens on small length scales and is directly related to stars forming, and “SMBH feedback”, which is related to processes involving supermassive black holes (SMBHs) at the center of each galaxy. Stellar feedback can include things like the pressure exerted by radiation from very bright stars, or by the blast waves of supernovae (exploding young stars). The crucial advance in understanding how these things work is that the simulations Phil and his collaborators have been doing account explicitly for the momentum generated by these processes, instead of simply considering the energy they dump into the surrounding dense gas. While hot gas has more pressure than cold gas and therefore won’t collapse as readily to form stars, dense gas in galaxies will cool off too quickly for that to make a difference, so considering ways of actually pushing the gas outwards and dispersing it has the potential to really change the game.
My favorite part of the talk was where Phil showed that you can get the right order of magnitude for stellar feedback through a back-of-the-envelope calculation based on the momentum flux, using arguments inspired by the simulation results. Using just a few ingredients you can reproduce the well-known Kennicutt-Schmidt relation, which expresses the star formation rate as a function of the observed surface density of cold gas on the sky, with the right slope. The slope of the relation is also independent of the details of how the stars form. Getting these details right matters most for low-mass galaxies, but it also helps the discrepancies at the high-mass end, so their importance can’t be (or haven’t been) stressed enough.
Starting soonish I’m going to try (for a while) posting one- or two-paragraph summaries of what I work on each day — besides being educational for readers, it should help me monitor the arc of how I’m spending my time and what I find interesting, which will become progressively more important as I continue my work here.