Download presentation
Presentation is loading. Please wait.
Published byCaroline Hensley Modified over 9 years ago
1
Simulations of the galaxy population constrained by observations from z=3 to present day: implications for galactic winds and fate of their ejecta Bruno Henriques Simon White, Peter Thomas, Raul Angulo, Qi Guo, Gerard Lemson, Volker Springel
2
Motivation Because we should. The physics of galaxy formation are complex but observations suggest they must obey simple relations. Why use a phenomenological approach to study galaxy formation? Still, we do not have a good understanding and cannot work from first principles, so models must be observationally based. Fast method to compute the evolution of the galaxy population across cosmic time for samples as large as modern surveys.
3
The evolution of the stellar mass function The model fits the present day distribution of masses but predicts the dwarf galaxy population to build up too early.Guo2011
4
Massive galaxies are too blue and dwarf galaxies too red Dwarf galaxies are too old and are not forming enough stars Dwarf galaxies are too clustered clustering colour SSFR
5
Semi-analytic modelling MCMC Complex galaxy formation physics Choose parameters to sample Star formation, SN feedback, AGN feedback efficiency, Metals yield 3. Self-consistent model of galaxy formation across cosmic time Henriques et al. (2009), Henriques & Thomas (2010), Henriques et al. (in prep.) Large Volume Across Cosmic Time Robust statistical method to explore the allowed likelihood regions in high- dimensional parameter spaces
6
Constrain the model at multiple redshifts Stellar Mass Function, K-band & B-band Luminosity Functions Wide and narrow surveys combined to achieve good statistics and large dynamical range. Maximum and minimum observational errors used to estimate systematic uncertainties.
7
Time varying parameters Reincorporation of gas after ejection by SN feedback needs to increase towards low redshift All other parameters have consistent regions at all redshifts Reincorporation time scaling with M vir, similar to Oppenheimer et al. (2008, 2010)
8
Strong ejection + no reincorporation set the low mass end at high-z Strong reincorporation at later times produces the required build up for z<1
9
Colors and SFR The delayed reincorporation of gas shifts star formation in dwarfs to lower redshifts. Low mass galaxies have higher star formation rates and younger ages. A population of low mass galaxies with blue colours remains down to z=0
10
Satellite galaxies in massive halos have lower mass, hence reducing clustering at fixed mass Galaxy formation physics, and not just cosmology/merging, have a strong impact on galaxy clustering.
11
Conclusions MCMC methods can be used to learn exactly how specific descriptions of a physical process affect galaxy observables at different epochs in a self-consistent way. The allowed likelihood regions in parameter space can be explored for any combination of observations at multiple epochs. In order to explain the observed evolution of the number density of intermediate/low-mass galaxies, the reincorporation of ejected gas should scale approximately with Mvir, being negligible at low mass at z>2 and rapid for most galaxies at low redshift. Low-galaxies form later and are significantly younger at z=0 Evolution of the massive end is reproduced across all redshifts Phenomenological models provide a fast method to describe the formation and evolution of galaxies in a cosmological volume, with high resolution and across cosmic time.
12
8 August 2015 12 No feedback The halo mass function is much steeper at both ends than the galaxy stellar mass function Supernova feedback has the right scale to make star formation sufficiently inefficient in small haloes The reheated gas would eventually cool in massive haloes producing an excessive number of bright galaxies Luminosity Function low mass Observations Supernova feedback high mass low mass
13
8 August 2015 13 Massive galaxies have more gas fuel than small ones No ongoing star formation Older populations than small galaxies Z=2.0 Z=1.0 Z=0.0
15
Stars Cold Gas Hot Gas Ejected Gas Recycling Star Formation Cooling Reheating Ejection Reincorporation Stars The Munich Model
16
different supernova feedback (increased efficiency) Merger treatment 1.The Munich Model Guo et al. 2011 Henriques et al. 2011, 2012 different stellar populations Croton et al. 2006 De Lucia & Blaizot 2007 AGN feedback model (suppression of cooling) dust model SN feedback model - reheating + ejection + reincorporation
18
Extended MCMC Capabilities Observational constraints at multiple redshifts Time-evolution of parameters (pre-processing step) Stellar mass and luminosity functions constraints from z=3 to z=0 Takes full advantage of the self-consistent evolution of galaxies If not needed, the current parametrisation is not ruled out by observations If needed, a different parametrisation is required (it rules out any others) If a good fit can not be found, the current model is ruled out
19
M05 vs BC03
20
Gas
21
TB-AGB TB-AGB + RHeB
22
Web-based, modeler & observer friendly semi-analytic model Combine the most robust set of dark matter numerical simulations available Stellar Mass resolution of 10 8 M with a large enough volume to sample BAO MS, MII & MXXL Monte Carlo Markov Chain optimization + Fit physical and cosmological parameters Modular implementation of the physics “Observer friendly” outputsChoose IMF, SPS, Bands, Dust modelGALFORMOD
23
Chemical Enrichment 0.8 M8 M SN Ia + Stellar WindsSN II Metals return timescale <100 Myr Rob Yates, Peter Thomas, Simon White, Guinevere Kauffmann, Bruno Henriques
24
Far-Infrared Emission Peter Thomas, Sorour Shamshir, Bruno Henriques, Qi Guo + Sussex Infrared Use empirical templates from Herschel to get an emission spectra for the light re-emitted by dust Full radiative transfer code
25
Cosmology Sampling with MCMC Incorporate the Angulo & White 2010 formalism into the semi-analytic model. Include cosmological parameters in the sampling. Bruno Henriques, Marcel Van Daalen, Raul Angulo, Simon White, Volker Springel, Fabio Fontanot, Qi Guo
26
Colours
27
Somerville
28
Henriques, Maraston, Monaco, et al. (Astro-ph: 1009.1392) Light – Weighted AgesMass – Weighted Ages TP-AGB M05 BC03 Average!!! Ages of Galaxies
29
SSP 8 August 2015 29 Van der Wel, Franx, Wuyts, et al. 2006 Chandra Deep Field - South ACS+IRAC+J&H filters
30
Older then the Universe!Undetected in MIPS! 8 August 2015 30 Maraston, Daddi, Renzini, et al. 2006 What are the implications for galaxy formation models?
31
MUSYC – Gawiser et al. 2006 GOODS – Giavalisco et al. 2004 Optical to mid-infrared data 8 August 2015 31 Marchesini 2009
32
CB07 8 August 2015 32 BC03 CB07 M05 Henriques, Maraston, Monaco, et al. (Astro-ph: 1009.1392)
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.