Large-Scale Structure: Next Frontier for Tests of Inflation Dragan - PowerPoint PPT Presentation

Large-Scale Structure: Next Frontier for Tests of Inflation Dragan Huterer University of Michigan

f NL = 0 f NL = -5000 f NL = -500 f NL = +500 f NL = +5000 Using publicly available NG maps by Elsner & Wandelt

Constraints from Planck . ISW-lensing subtracted KSW Binned Modal SMICA Local . . . . . . . . . . . . . . . . 2.7 ± 5.8 2.2 ± 5.9 1.6 ± 6.0 Equilateral . . . . . . . . . . . . − 42 ± 75 − 25 ± 73 − 20 ± 77 Orthogonal . . . . . . . . . . . . − 25 ± 39 − 17 ± 41 − 14 ± 42 NILC Local . . . . . . . . . . . . . . . . 4.5 ± 5.8 3.6 ± 5.8 2.7 ± 6.0 Equilateral . . . . . . . . . . . . − 48 ± 76 − 38 ± 73 − 20 ± 78 Orthogonal . . . . . . . . . . . . − 53 ± 40 − 41 ± 41 − 37 ± 43 SEVEM Local . . . . . . . . . . . . . . . . 3.4 ± 5.9 3.2 ± 6.2 2.6 ± 6.0 Equilateral . . . . . . . . . . . . − 36 ± 76 − 25 ± 73 − 13 ± 78 Orthogonal . . . . . . . . . . . . − 14 ± 40 − 9 ± 42 − 2 ± 42 C-R Local . . . . . . . . . . . . . . . . 6.4 ± 6.0 5.5 ± 5.9 5.1 ± 5.9 Equilateral . . . . . . . . . . . . − 62 ± 79 − 55 ± 74 − 32 ± 78 Orthogonal . . . . . . . . . . . . − 57 ± 42 − 41 ± 42 − 42 ± 42 Planck collaboration XXIV, 2013

Constraints from Planck: modal expansion X B ( k 1 , k 2 , k 3 ) = α prs q p ( k 1 ) q r ( k 2 ) q s ( k 3 ) p,r,s 10.0 5.0 Mode coefficieints 0.0 -5.0 NILC SEVEM -10.0 SMICA 0 50 100 150 200 250 300 Mode number Planck collaboration XXIV, 2013

Current upper bound on NG is ~1000 times smaller than this: P ( δ T/T ) δ T/T

Next Frontier: Large-Scale Structure CMB LSS dimension 2D 3D ∝ l max2 ∝ k max3 # modes systematics & relatively relatively selection func. clean messy temporal evol. no yes λ only λ , z, M, bias... can slice in

▲ Harvard-Cfa survey (1980s)

Dark Energy Survey (2013)

Dark Energy DESI (~2017) Survey (2013)

Dark Energy DESI (~2017) Survey (2013) LSST (~2020)

Dark Energy DESI (~2017) Survey (2013) Euclid and WFIRST (~2025) LSST (~2020)

21cm mapping Dark Energy DESI (~2017) Survey (2013) Euclid and WFIRST (~2025) LSST (~2020)

Non-Gaussianity papers 10 years of Primordial non-Gaussianity in the past 10 years WMAP7 WMAP9 Planck # of articles with 200 WMAP5 140 “Non-Gaussian” ? in the title WMAP7 Planck on the ADS data base (1.5 σ ) Dalal et al. 120 Yadav & Wandelt (2.8 σ ? ) WMAP3 150 100 WMAP5 (1.7 σ ) WMAP1 80 WMAP3 COBE (0.7 σ ) Large-Scale Structure 100 60 WMAP1 CMB (0.8 σ ) COBE ( ≪ 1 σ ) Inflation / Theory 40 50 non-primordial NG 20 0 0 2000 2002 2004 2006 2008 2010 2012 2000 2002 2004 2006 2008 2010 Produced by Emiliano Sefusatti Thursday, September 15, 2011

Large-Scale Structure in Three Easy Steps:

Step 1: Produce theory predictions (including from simulations)

Simulations with non-Gaussianity (f NL ) f NL = -5000 f NL = -500 Under-dense region evolution decrease with f NL f NL = 0 Over-dense region evolution increase with f NL f NL = +500 80 Mpc/h f NL = +5000 375 Mpc/h Same initial conditions, different f NL Slice through a box in a simulation N part =512 3 , L=800 Mpc/h Dalal et al. 2008

...and now f NL =0 f NL =100 f NL =1000 with baryons! Zhao, Li, Shandera & Jeong, arXiv:1307.5051

Step 2: Use multiple LSS probes in dataset

Using LSS (and CMB) tracers - correlation functions Giannantonio et al. 2013

1 2MASS 0.8 0.6 SDSS 0.4 0.2 LRG 0 Final constraints: -0.2 NVSS -0.4 HEAO QSO 2MASS SDSS LRG NVSS HEAO QSO Covariance matrix Giannantonio et al. 2013

Covariance of weak lensing probes results from numerical simulations Shear peaks Shear 2-pt Shear field Marian, Smith et al. 2013

Step 3: Control the Systematic Errors

Systematic errors ‣ Already a limiting factor in measurements ‣ Will definitely be limiting factor with Stage-IV quality data ‣ Quantity of interest: (true sys. − estimated sys.) difference ‣ Self-calibration : measuring systematics internally from survey

Example I: photometric redshift errors z phot -z spec Requirements from “training set” log (Scatter) 50% error degradation log (Bias) C. Cunha Ma, Hu & Huterer 2006

Example II: LSS calibration errors (a) Stellar density (b) Extinction (c) Airmass (d) Seeing (e) Sky brightness • dominate on large angular scales • can be measured, removed using same or other data Leistedt et al 2013 see also Ho et al 2012; Huterer et al 2013

Conclusion: LSS has a lot to offer; many handles on both physics (NG/DM/DE) and systematics 10 2 no nuisance parameters 5 Coupon HOD parameters 5 piecewise HOD parameters 10 1 σ (w a ) σ (w p ) 10 0 1-halo dominated 10 − 1 p k NL 10 − 2 10 − 3 10 − 1 10 0 10 1 k max Cunha, Huterer & Doré 2010 Heidi Wu