CROSS-CORRELATION IN THE HIGH-Z SKY FEDERICO BIANCHINI
THE QUEST FOR Λ Energy Density Dark Energy is z ∼ 1 z ‣ Learn about the Dark Energy/Gravity sector 2
THE QUEST FOR Λ Energy Density Dark Energy is TOMOGRAPHY z ∼ 1 z ‣ Learn about the Dark Energy/Gravity sector 3
THE QUEST FOR Λ Energy Density Dark Energy is TOMOGRAPHY z ∼ 1 z ‣ Learn about the Dark Energy/Gravity sector ‣ Investigate Astrophysics 4
THE QUEST FOR Λ Energy Density Dark Energy is TOMOGRAPHY z ∼ 1 z ‣ Learn about the Dark Energy/Gravity sector ‣ Investigate Astrophysics ‣ Isolate Systematics 5
A BRIEF HISTORY OF TIME 6
A BRIEF HISTORY OF TIME - Initial conditions - Particle content - Reionization - Cosmic acceleration 7
CMB LENSING IN A NUTSHELL CMB photons are weakly gravitationally deflected by the intervening matter distribution during their cosmic journey credits: ESA and Planck team ˜ X (ˆ n ) = X (ˆ n + r φ (ˆ n )) Z χ ∗ d χ f K ( χ ∗ − χ ) φ ( ˆ n ) = − 2 f K ( χ ∗ ) f K ( χ ) Ψ ( χ ˆ n ; η 0 − χ ) 0 Growth of Geometry structures 8
T(ˆ n ) ( ± 350 µK ) E(ˆ n ) ( ± 25 µK ) B(ˆ n ) ( ± 2 . 5 µK ) *no primordial B-modes
T(ˆ n ) ( ± 350 µK ) E(ˆ n ) ( ± 25 µK ) B(ˆ n ) ( ± 2 . 5 µK )
LENSING IS SMOOTH Lensing convolves the unlensed CMB power spectra with CMB lensing power spectrum 5 . 0 6000 4 . 5 5000 4 . 0 / 2 ⇡ [ µK 2 ] 3 . 5 4000 3 . 0 A L ` ( ` + 1) C TT 2 . 5 3000 ` 2 . 0 2000 1 . 5 1 . 0 1000 0 . 5 0 . 0 0 500 1000 1500 2000 ` 11
LENSING IS SMOOTH Lensing convolves the unlensed CMB power spectra with CMB lensing power spectrum IDEA 5 . 0 6000 4 . 5 5000 4 . 0 / 2 ⇡ [ µK 2 ] 3 . 5 4000 3 . 0 A L ` ( ` + 1) C TT 2 . 5 3000 ` 2 . 0 2000 1 . 5 1 . 0 1000 0 . 5 0 . 0 0 500 1000 1500 2000 ` 12
CMB LENSING RECONSTRUCTION Lensing introduces statistical anisotropy , i.e. correlates previously uncorrelated multipoles h X ( l ) Y ∗ ( l � L ) i = 0 h X ( l ) Y ∗ ( l � L ) i / φ ( L ) METHOD We can extract lensing by looking at the off-diagonal correlations between X and Y Hu&Okamoto02 Seljak&Zaldarriaga97 Lensing Potential Normalization Optimally-chosen Filtered (data) maps weight function X,Y ϵ [T,E,B] 13
APPLICATIONS OF CMB LENSING ‣ Neutrino masses { Das+11; Keisler+11; Planck XVII(13) ‣ Geometrical degeneracy & XV(15); Story+15, Sherwin+16,… ‣ Dark energy φ CMB ‣ Reionization A s e -2 τ δ g PlanckXIII(2015) x κ gal φ CMB T CIB y tSZ Azabajian+13 I C γ B 14
APPLICATIONS OF CMB LENSING ‣ Neutrino masses { Das+11; Keisler+11; Planck XVII(13) ‣ Geometrical degeneracy & XV(15); Story+15, Sherwin+16,… ‣ Dark energy φ CMB ‣ Reionization A s e -2 τ Smith+07, ‣ Tracers bias Bleem+12,Sherwin+12, ‣ Photo-z calibration Planck XVII(13), FB+15 &16 , δ g Giannantonio+15, … ‣ Primordial non-Gaussianities ‣ Multiplicative bias calibration Hand+15;Liu&Hill15; Kirk+16; x ‣ Intrinsic alignments Harnois-Déraps+16 & 17 κ gal φ CMB ‣ Star formation history Planck XVIII(13); Holder+13; ‣ Tracers bias Hanson+13; van Engelen+15 ‣ Redshift distribution estimation T CIB Hill&Spergel13 ‣ Warm-hot intergalactic medium Van Waerbeke+14 ‣ Clusters environment y tSZ ‣ Dark Matter studies Fornengo+15 ‣ Leverage for high-z astrophysical Feng+16 I C γ B contribution to gamma sky 15
TWO SIDES OF THE SAME COIN IDEA Hosts MATTER Lens CMB Host MATTER Lens CMB Galaxies DISTRIBUTION photons galaxies DISTRIBUTION photons Lensing is insensitive to matter’s nature Light is a biased tracer of matter Lensing is insensitive to nature of matter Light is a (biased) tracer of matter Galaxy Matter Galaxy Matter δ g = b δ fluctuations fluctuations fluctuations fluctuations Bias Bias 16
SIGNAL MODELING Observables trace DM fluctuations with different weightings Overlap between kernels means that there is a xc signal 0 . 8 0 . 8 0 . 6 0 . 6 dN/dz W κ 0 . 4 0 . 4 0 . 2 0 . 2 0 . 0 0 . 5 1 . 0 1 . 5 2 . 0 2 . 5 3 . 0 3 . 5 4 . 0 z Z z ∗ H ( z ) dz χ 2 ( z ) W � ( z ) b ( z ) dN ⇣ ⌘ C � g = b σ 2 = κ data dz P k, z δ data ∝ ⇥ 8 grad c g 0 Z z ∗ H ( z ) dz b ( z ) dN i 2 h = C gg � � ∝ b 2 σ 2 = P k, z δ data δ data � 8 χ 2 ( z ) c dz g g 0 17
MAGNIFICATION BIAS Lensing of background galaxies by foreground LSS induces an apparent clustering in the sky WL limit δ obs n ) = δ clust (ˆ (ˆ g (ˆ g (ˆ n ) + δ µ n ) n ) ∝ ( α − 1) → δ µ − − − − − − g g 5 Flux Limit = 33 mJy ‣ Modifies observed area 4 N ( > S ) ∝ S − α ‣ (De-)magnify observed fluxes log N ( > S ) α ' 3 3 2 1 0 1 . 4 1 . 6 1 . 8 2 . 0 2 . 2 2 . 4 2 . 6 2 . 8 3 . 0 3 . 2 log S 250 [mJy] 18
THE INGREDIENTS CMB LSS Planck TM Herschel TM ∗ 2009 − † 2013 ‣ 9 bands in [30,857] GHz ‣ SPIRE 250, 350, 500 µm ‣ Beams FWHM in [30’ - 5’] ‣ Beams FWHM 18”, 26”, 36” ‣ Noise ~30/40 µ K-arcmin ‣ Noise 5.8, 6.3, 6.8 mJy 19
PLANCK CMB LENSING MAPS To test the robustness of results we make use of CMB lensing maps from both 2013 and 2015 Planck data releases ˆ φ WF 2015 Planck 10 − 5 10 − 6 10 − 7 C ` 10 − 8 N 2015 ` N 2013 ` 10 − 9 10 1 10 2 10 3 ` -7e-05 6e-05 PlanckXVII (2013) T : φ = COBE : Planck PlanckXV (2015) 20
H-ATLAS GALAXY SAMPLE ‣ Selection criteria - Baseline (flux-based) S 250 µm > 35 mJy S 350 µm > 3 σ - Gonzalez-Nuevo+12 (flux + color) S 350 µm /S 250 µm > 0 . 6 S 500 µm /S 350 µm > 0 . 4 ‣ Tomographic photo-z bins z ph ≥ 1 . 5 1 . 5 ≤ z ph < 2 . 1 z ph ≥ 2 . 1 ‣ k f sky ' 0 . 01 N g ' 90000 21
SUB-MM MAGIC The sub-mm flux remains approximately constant for z > 1 Swinbank+10 Lapi+11 Photo-zs are estimated through template fitting assuming the SED of SMM J1235-0102 22
THE PATCHES Convergence NGP Convergence SGP Convergence G09 Convergence G12 Convergence G15 1.5 ’/pix, 1850x1850 pix 1.5 ’/pix, 700x700 pix 1.5 ’/pix, 600x600 pix 1.5 ’/pix, 600x600 pix 1.5 ’/pix, 600x600 pix (51,84) (0,-78) (-132,28) (-84.5,60) (-12,54) -0.411 0.522 -0.411 0.522 -0.411 0.522 -0.411 0.522 -0.411 0.522 Galaxies NGP Galaxies SGP Galaxies G09 Galaxies G12 Galaxies G15 1.5 ’/pix, 1850x1850 pix 1.5 ’/pix, 700x700 pix 1.5 ’/pix, 600x600 pix 1.5 ’/pix, 600x600 pix 1.5 ’/pix, 600x600 pix (51,84) (0,-78) (-132,28) (-84.5,60) (-12,54) -0.737 0.802 -0.737 0.802 -0.737 0.802 -0.737 0.802 -0.737 0.802 23
CORRELATION BY EYE Sub-mm galaxies trace the peaks of matter density field and are denser in regions where the CMB convergence is enhanced 0 . 06 0 . 04 0 . 02 0 . 00 h i � 0 . 02 � 0 . 04 z > 1 . 5 1 . 5 < z < 2 . 1 � 0 . 06 z > 2 . 1 � 0 . 08 � 0 . 4 � 0 . 2 0 . 0 0 . 2 0 . 4 0 . 6 0 . 8 h δ i Maps smoothed to ~ 1 degree scale (~ 30 Mpc at z = 2) 24
AND NOW DATA!
2013 VS 2015 DATA CROSS-SPECTRA ˆ � 0 Q � 0 L 0 ) − 1 P L 0 � ˜ X X C XY P L � M �� 0 B 2 C XY = ( Hivon+01 � L L 0 � �� 0 z ph > 1 . 5 ˆ Theory b = 2.79, A = 1.65 Theory b = 2.80, A = 1.62 7 . 5 2013 × g GN 12 M 2013 ` ( × 10 − 7 ) 2013 × g GN 12 M 2015 2015 × g GN 12 M 2015 5 . 0 h i = 2015 × g 35 mJy M 2015 C g 2 . 5 0 . 0 100 200 300 400 500 600 700 800 z ∼ 2 FB +16 ` ∼ 50 Mpc ∼ 6 Mpc 26
ERRORS & NULL-TESTS ‣ We use two sets of sims : - 500 correlated Gaussian galaxy and CMB lensing maps - 100 sims of CMB lensing released by Planck team ‣ Null test to validate power spectrum extraction pipeline ‣ Covariance matrices evaluation × 10 − 8 1 . 2 1 . 2 True H-ATLAS Planck Sims 0 . 8 1 . 0 Real Herschel C κ g 0 . 4 ℓ Real Planck ℓ ( × 10 − 7 ) 0 . 0 0 . 8 Analytical − 0 . 4 χ 2 / d . o . f . = 7 . 2 / 7 → P.T.E. = 41% − 0 . 8 0 . 6 × 10 − 8 ∆ C κ g 1 . 2 True Planck 0 . 4 0 . 8 C κ g 0 . 4 ℓ 0 . 2 0 . 0 − 0 . 4 χ 2 / d . o . f . = 5 . 9 / 7 → P.T.E. = 55% 0 . 0 − 0 . 8 100 200 300 400 500 600 700 800 100 200 300 400 500 600 700 800 ℓ ℓ FB +15 27
DATA TOMOGRAPHY Null hypothesis (= no correlation between fields) is rejected at a significance between 10 𝜏 to 22 𝜏 2015 × g 35 mJy g 35 mJy Theory b = 3.54, A = 1.45 Theory b = 3.54 Theory b = 2.89, A = 1.48 Theory b = 2.89 2 . 4 7 . 5 Theory b = 4.75, A = 1.37 Theory b = 4.75 ` ( × 10 − 7 ) ` ( × 10 − 6 ) z ph ≥ 1 . 5 ˆ z ph ≥ 1 . 5 ˆ 1 . 5 ≤ ˆ z ph < 2 . 1 1 . 5 ≤ ˆ z ph < 2 . 1 1 . 6 5 . 0 z ph ≥ 2 . 1 ˆ z ph ≥ 2 . 1 ˆ C g C gg 2 . 5 0 . 8 0 . 0 0 . 0 100 200 300 400 500 600 700 800 100 200 300 400 500 600 700 800 ` ` FB +16 28
CONSTRAINTS FROM JOINT ANALYSIS We introduce an amplitude parameter A that rescales theoretical cross- power spectrum and combine observed cross- and auto-power spectra z > 1 . 5 2 . 1 1 . 5 < z < 2 . 1 1 . 8 z > 2 . 1 L = AC κ g, th C κ g ˆ L A 1 . 5 ∝ Ab 1 . 2 ? A > 1 @ 2-3 s 0 . 9 2 . 4 3 . 0 3 . 6 4 . 2 4 . 8 5 . 4 b FB +16, Aversa+15 29
Recommend
More recommend