CLASSE NEWS |
10 Jul 2014
New Tools for Seeking Dark Matter at the LHC
The dark matter that holds our universe together remains a deep mystery: we know it's there but have no idea what it is.
An intense search for dark matter is going on at the Large Hadron Collider (LHC) at the CERN Laboratory in Switzerland. This accelerator produces the highest energy particle collisions ever made in laboratory conditions, and thereby provides an important route to understanding dark matter. And the key to finding dark matter in an accelerator experiment is to observe its
absence.
Cornell graduate student and CMS collaborator Nathan Mirman presented a new technique at the 2014 LHCP Conference for sifting through the missing energy and momentum at the collider to find those cases -- the absences -- that could be due to escaping dark matter. The essence of the method is to compare the missing momentum with a calculation of how much momentum
might have been missed simply because of accumulated normal measurement error. The new part is to assess the accumulated error without restricting oneself to the oversimplified assumption of Gaussian errors. While the accumulated Gaussian errors can by summed easily with a simple formula, the accumulated non-Gaussian errors must be laboriously summed by doing successive, nested, convolution integrals. The CPU time needed is prohibitive when there are tens-to-hundreds of convolutions to be performed in each of tens-to-hundreds of millions of events.
Unless you have a better idea.
The better idea is to exploit Fourier transforms: when the functions describing the error distributions are Fourier transformed, the convolutions become trivial. Undergraduate physics major Yimin Wang developed the code to exploit Fast Fourier Transform (FFT) techniques to carry out the convolutions in a manageable time. The result now allows physicists to winnow out more dark matter fakes from the data to distill a purer sample of possible candidates.
No, the true dark matter culprit has not appeared yet; but the search techniques are getting stronger, the data sets are growing, and the energy available at the LHC will soon take a jump. We may find it yet!
