ICIC

Among the many other things I haven’t had time to blog about, this term we opened the new Imperial Centre for Inference and Cosmology, the culmination of several years of expansion in the Imperial Astrophysics group. In mid-March we had our in-house grand opening, with a ribbon-cutting by the group’s most famous alumnus.

Statistics and astronomy have a long history together, largely growing from the desire to predict the locations of planets and other heavenly bodies based on inexact measurements. In relatively modern times, that goes back at least to Legendre and Gauss who more or less independently came up with the least-squares method of combining observations, which can be thought of as based on the latter’s eponymous Gaussian distribution.

Our group had already had a much shorter but still significant history in what has come to be called “astrostatistics”, having been involved with large astronomical surveys such as UKIDSS and IPHAS and the many allowed by the infrared satellite telescope Herschel (and its predecessors ISO, IRAS and Spitzer). Along with my own work on the CMB and other applications of statistics to cosmology, the other “founding members” of ICIC include: my colleague Roberto Trotta who has made important forays into the rigorous application of principled Bayesian statistics to problems cosmology and particle physics; Jonathan Pritchard who studies the distribution of matter in the evolving Universe and what that can teach about its constituents and that evolution; and Daniel Mortlock, who has written about some of his work looking for rare and unusual objects elsewhere on this blog. We are lucky to have the initial membership of the group supplemented by Alan Heavens, who will be joining us over the summer and has a long history of working to understand the distribution of matter in the Universe throughout its history. This group will be joined by several members of the Statistics section of the Mathematics Department, in particular David van Dyk, David Hand and Axel Gandy.

One of the fun parts of starting up the new centre has been the opportunity to design our new suite of glass-walled offices. Once we made sure that there would be room for a couple of sofas and a coffee machine for the Astrophysics group to share, we needed something to allow a little privacy. For the main corridor, we settled on this:
IMG 2932
The left side is from the Hubble Ultra-Deep field (in negative), a picture about 3 arc minutes on a side (about the size of a dime or 5p coin held at arm’s length), the deepest — most distant — optical image of the Universe yet taken. The right side is our Milky Way galaxy as reconstructed by the 2MASS survey.

The final wall is a bit different:
IMG 2926
The middle panels show part of papers by each of those founding members of the group, flanked on the left and right side with the posthumously published paper by the Rev. Thomas Bayes who gave his name to the field of Bayesian Probability.

Of course, there has been some controversy about how we should actually refer to the place. Reading out the letters gives the amusing “I see, I see”, and IC2 (“I-C-squared”) has a nice feel and a bit of built-in mathematics, although it does sound a bit like the outcome of a late-90s corporate branding exercise (and the pedants in the group noted that technically it would then be the incorrect I×C×C unless we cluttered it with parentheses).

We’re hoping that the group will keep growing, and we look forward to applying our tools and ideas to more and more astronomical data over the coming years. One of the most important ways to do that, of course, will be through collaboration: if you’re an astronomer with lots of data, or a statistician with lots of ideas, or, like many of us, somewhere in between, please get in touch and come for a visit.


Unfortunately we don’t yet have a webpage for the Centre..