Talks

Durham University

OCW017 | The Ogden Center for Fundamental Physics (West) | Department of Physics | South Road | Durham | DH1 3LE | United Kingdom

Galaxy clustering encodes rich information about the physical processes that govern the occupation of dark matter haloes - yet despite its sensitivity to these processes, clustering is rarely used as a direct constraint on physical models of galaxy formation. A key reason for this is computational: measuring clustering statistics for the large, simulated catalogues produced by semi-analytical and empirical models is expensive, making it impractical to include clustering in parameter calibration pipelines that must explore large parameter spaces. We present a general method with which clustering can be recovered accurately from a sparsely sampled galaxy catalogue, provided a suitable correction is applied to account for the reduced sampling rate. We demonstrate this by applying this method to GALFORM, run on the P-Millennium N-body simulation. The reduction in computational cost is sufficient to make galaxy clustering a viable statistic for use in the calibration of semi-analytical model parameters for the first time. Finally, we discuss the dependence of the required sampling rate on scale and galaxy population and outline the regime in which the sparse-sampling approximation is reliable.

University of Heidelberg

Heidelberg University, Grabengasse 1, 69117 Heidelberg, Germany