MATTEO VIEL

THE LYMAN- FOREST AS A COSMOLOGICAL PROBE

Contents and structures of the Universe – La Thuile (ITALY), 19 March 2006

80 % of the baryons at z=3 are in the Lyman- forest

baryons as tracer of the dark matter density field

IGM ~ DM at scales larger than the

Jeans length ~ 1 com Mpc

optical depth: ~ (IGM )1.6 T -0.7

GOAL: the primordial dark matter power spectrum from the observed flux spectrum

Tegmark & Zaldarriaga 2002

CMB physics z = 1100 dynamics

Lyphysics z < 6 dynamics + termodynamics

CMB + Lyman Long lever arm

Constrain spectral index and shape

Relation: P FLUX (k) - P MATTER (k) ??

Continuum fitting

Temperature, metals, noise

Maximum sensitivityof WMAP

RESULTS: POST- WMAP I LYMAN-

CMB

3000 LOW RESOLUTION LOW S/N 30 HIGH RESOLUTION HIGH S/NSDSS LUQAS

SDSS vs LUQAS

McDonald et al. astro-ph/0405013 Kim, Viel et al. 2004, MNRAS, 347, 355

DM

STARS

GAS

NEUTRALHYDROGEN

m = 0.26 = 0.74 b=0.0463 H0 = 72 km/sec/Mpc - 60 Mpc/h 2x4003

GAS+DM 2.5 com. kpc/h softening length

GADGET –II code COSMOS computer – DAMTP (Cambridge)

DATA 27 high res z~2.125 QSO spectra 3000 QSO spectra z>2.2

THEORY: SIMULATIONS full hydro simulations `calibrated’ simulations

THEORY: METHOD 3D flux inversion 1D flux modelling

ADVANTAGES band power large redshift range

DRAWBACKS only z=2.1 (and 2.72) and 34 parameters modelling larger error bars

RESULTS no large running, scale invariant, `high’ power spectrum amplitude

LUQAS SDSS

Cosmological implications: combining the forest data with WMAP

SDSS Seljak et al. 2004

8= 0.93 ± 0.07 n=0.99 ± 0.03

8= 0.90 ± 0.03 n=0.98 ± 0.02 nrun = -0.003 ± 0.010 nrun=-0.033± 0.025

d ns/ d ln k

Viel, Weller, Haehnelt, MNRAS, 2004, 355,L23

Viel, Haehnelt, Springel, MNRAS, 2004, 354, 684

Seljak et al. 2004

In the Seljak analysis WMAP+Ly- constrains running three times better than WMAP+all the rest

SDSS Seljak et al. 2005

r < 0.45 (95% lim.)

r = 0.50 ± 0.30

T/S =

T/S

V = inflaton potential

Viel, Weller, Haehnelt, MNRAS, 2004, 355, L23

Inflation and dark energy

RESULTS: POST- WMAP II LYMAN-

CMB

WMAP I + Lyman- high-res WMAP III + Lyman- high-res

(credit: Antony Lewis)

Tension between WMAP3 and high resolution Lyman-a… 1.5 discrepancy

for SDSS it could be worse…

WMAP + LUQAS (high res) sample

New COSMOMC version with WMAP3 and Lyman- downloadble http://cosmologist.info/

8= 0.87 ± 0.04

n = 0.97 ± 0.04

n run (0.002 Mpc-1)= 0.005 ± 0.030

SDSS analysis Seljak et al. 2005, PRD

best fit amplitude from SDSS Lyman-a and WMAP1

is 3.5 off the new WMAP3 value

WMAP + SDSS flux power constraints

Best fit WMAP3

SMALL SCALE POWERWarm dark matter ? New observational results

from dwarf galaxies

Neutrinos ?

Isocurvature ?

Cosmological implications: Warm Dark Matter particles

CDMWDM 0.5 keV

30 comoving Mpc/h z=3

MV, Lesgourgues, Haehnelt, Matarrese, Riotto, PRD, 2005, 71, 063534

k FS ~ 5 Tv/Tx (m x/1keV) Mpc-1

k FS ~ 1.5 (m x/100eV) h/Mpc

In general if light gravitinos

Set by relativistic degrees of freedom at decoupling

Cosmological implications: WDM, gravitinos, neutrinos

Set limits on the scale ofSupersymmetry breaking

susy < 260 TeV

m WDM > 550 eV m grav< 16 eV > 2keV sterile neutrino

m (eV) < 0.95 (95 %C.L.) WMAP + 2dF + LY

WDM CWDM(gravitinos)

neutrinos

Can sterile neutrinos be the dark matter ?

Seljak et al. a-ph/0602430

Abazaijan a-ph/0512631

Biermann & Kusenko PRL 2006,96, 091301 - Molecular hydrogen and reionization

- X-Ray backgrounds

- Flux power evolution

Flux p

ow

er

2 < m (KeV) < 8 m (KeV) > 14

Increasing z

M sterile neutrino > 10 KeV 95 % C.L.

SDSS data only

8 = 0.91 ± 0.07n = 0.97 ± 0.04

Fitting SDSS data with GADGET-2 this is SDSS Ly- only !!

Isocurvature perturbations

Beltran, Garcia-Bellido, Lesgourgues, Viel, 2005, PRD, D72, 103515

:isocurvature fraction at k=0.05 Mpc-1

:quantifies correlation between ISO andADIABATIC modes

No iso Only iso

Fully correlated

Anti correlated

n iso=1.9 ± 1.0 95% C.L.

This is a model in agreement with WMAP3!

n ad=0.95n iso =3

n ad = 0.95n iso =2

SUMMARY

Cosmological parameters: Lyman- points to 8 = 0.9 n = 0.98 little running

substantial agreement between SDSS and LUQAS but SDSS has smaller error bars (factor ~ 2), due to the different theoretical modelling and wider range of redshift probed by SDSS Constraints on inflationary models, isocurvature, neutrinos and WDM have been obtained

From high res Lyman-: gravitinos + Dark matter and isocurvature Models are in agreement with WMAP3 Other measurements using Lyman- based on different methods give the same answer (Zaroubi et al. 2005, Viel & Haehnelt 2006), systematics are now in better control

Lyman- forest is a complementary measurement ofthe matter power spectrum and can be use to constrain cosmologytogether with the CMB – TENSION with WMAP 3

Mean flux

Effective optical depth

<F> = exp (- eff) Power spectrum of F/<F>

Low resolution SDSS like spectra

High resolution UVES like spectra

Thermal state

T = T0 ( 1 + )

Statistical SDSS errors on flux power

Thermal histories Flux power fractional differences

Numerical modelling: HPM simulations of the forest

Full hydro 200^3 part. HPM NGRID=600 HPM NGRID=400F

LUX

PO

WE

R

MV, Haehnelt, Springel, astro-ph/0504641

Hydro-simulations: what have we learnt?Many uncertainties which contribute more or less equally

(statistical error seems not to be an issue!)

Statistical error 4%

Systematic errors ~ 15 %

eff (z=2.125)=0.17 ± 0.02 8 %

eff (z=2.72) = 0.305 ± 0.030 7 %

= 1.3 ± 0.3 4 %

T0 = 15000 ± 10000 K 3 %

Method 5 %

Numerical simulations 8 %

Further uncertainties 5 %

ERRORS CONTRIBUTION TO FLUCT. AMPL.

Constraints on the evolution of the effective optical depth

8 = 0.7

8 = 1

8 = 0.925

Spergel et al. 2003

Peiris et al. 2003

Lyman- does most of the job in constraining running!!!

Lack of power in theRSI model: haloesform later. Contrastwith the large inferred(Yoshida et al. 2003)

The LUQAS sample

Kim, MV, Haehnelt, Carswell, Cristiani, 2004, MNRAS, 347, 355

Large sample Uves Qso Absorption Spectra(LP-UVES program P.I. J. Bergeron)

high resolution 0.05 Angstrom, high S/N > 50low redshift, <z>=2.25, z = 13.75

#

s

pect

ra

redshift