Subhaloes going Notts Julian Onions University of Nottingham

Subhaloes going Notts Julian Onions – University of Nottingham With Frazer Pearce, Alex Knebe, Stuart Muldrew, Hanni Lux, et al.

What is it? • Comparison of substructure detection in a • Milky Way sized object. Follow on from “Haloes going Mad”. Why do it? • Dark matter detection via annihilation. • Lensing via substructure. • Is 1% accuracy realistic? • Many large, single halo simulations each with own subhalo detector: are these comparable?

Finder Contact Method Level Adaptahop Dylan Tweed 3 -space particle 4 AHF Alex Knebe Adaptive mesh 1 Grasshopper Doug Potter 3 -space particle 2 HBT Jiaxin Han 3 -space+tracker 2 HOT 3 D Yago Ascasibar 3 -space particle 4 HOT 6 D Yago Ascasibar Phase space 4 HSF Michal Maciejewski Phase space 2 Mendieta Andres Ruiz 3 -space particle 3 Rockstar Peter Behroozi Phase space 1 STF Pascal Elahi Velocity/Phase space 3 Subfind Volker Springel 3 -space particle 1 Voboz Mark Neyrinck Voronoi 3

Common Processing • Pipeline – Assign duplicates to smallest halo – refine centre of mass iteratively based on 50% Process up to M 200 Calculate statistics – –

Results

Sub. Haloes Going Notts 14 -18 th May, 2012, Derbyshire

A Productive Cosmo. Comp workshop - Overview paper (Knebe et al. ) Spin paper (Onions et al. submitted) Shape parameter Spin parameter 2 Subhalo Correlation function (Pujol et al. ) Performance in Major Mergers (Behroozi et al. ) Shape of Subhalo population (Hoffmann et al. ) Galaxies going MAD (Knebe et al. submitted) GHalo Comparison (Potter et al. ) Streams in Subhalofinders (Elahi et al. ) Lagrangian Regions (Neyrinck et al. )

Performance

Where does the scatter come from? Several possibilities • Processing inconsistencies • Unbinding steps – do some not remove as many as others • Particle gathering – different strategies for gathering.

Own Unbinding

Common Unbinding

No Unbinding

Own Unbinding

Common Unbinding

Different particles

No subfind

40 -80, 000 particles ~10%

Conclusions • Several reliable, efficient subhalo finders exist • For recovering substructure, phase space finders indistinguishable from real space finders • Important to use good unbinding • Good to about 10% accuracy, more seems hard. • Still scatter – driven largely by the definition of subhalo being used.