Clusters of Galaxies (SWG+OULE3) simulaLon requirements LAURO - - PowerPoint PPT Presentation

LAURO MOSCARDINI

(on behalf of CG‐SWG and OULE3 Clusters)

• DIP. FISICA E ASTRONOMIA, UNIBO

LAURO.MOSCARDINI@UNIBO.IT

Clusters of Galaxies (SWG+OULE3) simulaLon requirements

Euclid CSWG meeLng, Barcelona, 4th ‐5th December 2014

Conclusions

The majority of our ac0vi0es, mostly at the level of OU‐LE3 (but also of SWG), will have strong criLcaliLes if we don’t have soon a new ‘good’ set of mock catalogues!

Why do we need simulaLons?

•  Galaxy Cluster detecLon 
• Purity and completeness of Cluster catalogues
• Richness
• Mean redshiI
• Velocity dispersion
• Weak lensing masses
• Cluster luminosity func0on
• Cluster stellar mass func0on
• Radial profile
• 2‐point correlaLon funcLon of clusters
• Power spectrum of clusters
• 3‐point correlaLon funcLon
• Bispectrum of clusters
• Covariance matrix of the cluster 2‐point corr. funct.
• Covariance matrix of the cluster power spectrum

OU‐LE3 Clusters of Galaxies ImplementaLon WPs

Similar to what done for Galaxy Clustering

Why do we need simulaLons?

• Sample selec0on
• Mass modelling
• Likelihood
• StaLsLcs on cluster samples: including CalibraLon of

halo mass and bias: effects of baryons and non‐ standard DM and cosmological models

• Mass‐observable relaLon
• ValidaLon
• Astrophysics of Galaxy Clusters
• External Data

Clusters of Galaxies SWG WPs

Cluster detecLon with Euclid

Given the amount and quality of the Euclid data, there are at least three main ways available:

• from the photometric survey
• from the spectroscopic survey
• from the cosmic shear survey

(but also from strong lensing features, etc. …) and, of course, from a combinaLon of the previous methods and/or including informa0on from external data in other wavelengths

Cluster idenLficaLon:

this is at moment our bo`leneck

Cluster idenLficaLon makes quite strong requirements on mocks: we require galaxies to be “painted” in such a way that

• bserved environmental dependence of the galaxy popula0on is

decently reproduced (unlike for WL, and, to first approxima0on, for GC). For our immediate needs, we don’t want to study galaxy forma0on  HOD mocks are probably fine for us. BeZer to have a fully phenomenological (but realisLc) and flexible descripLon

• f how galaxies are distributed, rather than a physically

mo0vated descrip0on of galaxy forma0on from “first principles”.

What has been done up to now?

Seven different cluster finder algorithms were applied to the same set of mock catalogues built to reproduce the forthcoming Euclid photometric survey.

2 different rounds of the Cluster finder challenge

(see the internal report EC_CFC1_V1.0.pdf)

The mock used: Durham

Extracted by A. Cappi from the 500 degree2 H‐band light‐ cone limited to HAB ≤24 provided by C. Baugh and A. Merson. A photometric redshie has been assigned from a Gaussian distribu0on with a precision mimicking the expected Euclid

• performances. Catastrophic failures also included.

Matching has been done w.r.t. the catalogue of halos having at least M≤1013 Msun and 5 members.

Completeness(M,z)

RedshiI Mass

Courtesy: S. Farrens et al.

z<0.9 z>0.9

Completeness Completeness

Purity vs. Completeness

Courtesy: A. Iovino et al.

“Problems” with the Durham mocks

• cluster galaxies are not red enough: the sequence is too

blue and there are small differences w.r.t. to field galaxies

• galaxy density profiles are not well reproducing real

data

• unrealis0c tails in the velocity distribuLon inside high‐z

clusters  no way to use them to study velocity dispersion

• no Halpha emi`ers in clusters (SAM kill star‐forming

galaxies in clusters)  no way to use them to understand the spectroscopic selec0on func0on

• unclear defini0on of the halo properLes (mass, center,

…)

No red sequence at z>1

Black Points: a simulated cluster at z=1.4 with M=2 1014 Blue circles: a real high‐z cluster (XMMU J235‐25587)

Bardelli

Red sequence is necessary to strengthen detecLons!

Galaxy density profiles

Roncarelli/Bellagamba

Clusters with: mass around Mfof=1014 Msun 0.2<z<0.3 H<18

0.01 0.1 1 10 100 0.05 0.1 0.2 0.5 1 2 galaxy number density (Mpc/h)-2 radius (Mpc/h) SDSS profile MICE profile Durham profile

Halpha galaxies

Iovino/CucciaL

Orange/Red points: galaxies H‐band selected Blue/Black points: Halpha galaxies in ‘groups’ Cyan Points: Halpha galaxies in the field

The Halpha selected sample avoids the cosmic structures delineated by H‐band selected galaxies

6/27/14 Paris Clusters

Tests on the MICE simulaLons:

Counts in H band

too shallow: incomplete at H>21 and

• nly up to z=1.4

Contribu0ons from:

• S. Bardelli, F. Bellagamba, A.

Cappi, S. Maurogordato, M. Roncarelli, M. Vannier

Tests on the MICE simulaLons:

Red sequence at z=0.5 in the right posi0on

CLASH cluster, Mercurio et al.

Black Points: a simulated cluster at z=1.2 Red points: galaxies in clusters Green circles: central galaxies Blue circles: a real cluster (XMMU J235‐25587) Magenta star: expected color for a passive galaxy at z=1.4

Tests on the MICE simulaLons:

Red sequence at higher redshie Some BCG in the JH plane are too blue and there is a too large spread in J‐H

Our Christmas (?!) wishlist

i.e. What we urgently need (I)

1) A mock catalogue (past‐light cone) with all galaxies, selected down to the nominal limi0ng apparent magnitudes of Euclid (and possibly Pan‐STARSS and DES). To be on the safe side, one should include all the galaxies that sa0sfy at least

• ne of the Euclid photometric magnitude limits (preferred H=24?) and Halpha

limit. For these galaxies, we need the real redshie, the observed redshie, the photometric redshie P(z), the peculiar velocity and the possible membership in dark haloes. Different implementaLons of galaxy formaLon would allow to take into account our ignorance of cluster proper0es at high‐z. The mock must cover a sufficiently large volume (possibly 15000 deg2)

What we urgently need (II)

2) A catalogue of all dark ma`er halos above a given limi0ng mass (e.g. 10^13 solar masses) with: ‐ posi0on of the centre (as minimum of the gravita0onal poten0al) ‐ M_delta and r_delta at different delta overdensi0es (e.g. 2500, 500, 200, and delta=virial with respect to cri0cal density at the halo redshiI) ‐ velocity dispersions from DM par0cles along three cartesian axes ‐ mass density profiles ‐ possible galaxy membership 3) The corresponding Euclid shear maps for weak lensing detec0on/mass calibra0on/etc. 4) The corresponding X‐ray and SZ maps to test the possible improvement

• f detec0ons and mass es0mates with external data (eRosita, Planck, ACT,

etc.)  less urgent but important!

Conclusions

The majority of our ac0vi0es, mostly at the level of OU‐LE3 (but also of SWG), will have strong criLcaliLes if we don’t have soon a new ‘good’ set of mock catalogues!

QuesLon

Which SWG or OU is going to address

• ur requirements?