Measuring the kinematic Sunyaev Zel’dovich Effect
with SPT and DES
Moriond March 24, 2016
Kyle Story Stanford University KIPAC
Measuring the kinematic Sunyaev Zel’dovich Effect
with SPT and DES
In collabora@on with: B. Soergel, S. Flender, L. Bleem, T. Giannantonio, G. Efstathiou, as well as the SPT and the DES collabora@ons
arXiv: 1603.03904
The South Pole Telescope Collaboration
The Dark Energy Survey Collaboration
The South Pole Telescope (SPT) 10-meter primary dish, observe the CMB with arcminute resolution SPT-SZ (2007) 960 detectors 95,150,220 GHz
SPTpol (2012) 1600 detectors 95,150 GHz +Polarization
SPT-3G (2017) ~16,000 detectors 95,150,220 GHz +Polarization Kyle Story
Recall talk by Amy Bender
4
Slide from T. Giannantonio
Slide from B. Benson
CMB + galaxy surveys: Powerful playground for testing the consistency of cosmology and probing open questions.
CMB Last-Scattering Surface “back-light” on the rest of the universe Recall talk by Nabila Aghanim
• • • •
CMB lensing Galaxies Galaxy weak lensing kSZ (image modified from WMAP/NASA)
Primer: the thermal SZ effect The Sunyaev-Zel’dovich effect is caused by CMB photons inverse Compton scattering off of electrons in hot intra-cluster gas. CMB illuminates galaxy clusters from behind: “back light”
Kyle Story
8
If the cluster is moving with respect to the CMB rest frame, the spectrum will contain an additional Doppler shift.
(Carlstrom 2002)
thermal SZ
kinematic SZ
Sunyaev & Zel’dovich 1980 Kyle Story
9
Why is the kSZ signal interesting? The kSZ probes the cosmic velocity field on ~100 Mpc scales electron line-of-sight velocity electron number density
Using the kSZ, we can study: • “Missing Baryon” problem • Probe of Large Scale Structure • Probe of cluster physics (electron density) • Probe of gravity on ~100 Mpc scales Kyle Story
Local cosmic velocity field (Courtois et. al. 2013)
How do we measure the kSZ? 90
o
o
1
Angular Scale 5′ 2′
10′
Planck
SPT 220 GHz
SPT - S13
SPT 150 GHz
ACT 148 GHz ACT 220 GHz
SPT 95 GHz
Power
1000
1′
100
10 10
500
1500
Angular Scale
2500
6000
10000
thermal SZ
kinematic SZ
The SZe contributes to the CMB anisotropy power spectrum. • tSZ component is well measured (e.g., George et al 2015) • kSZ component is very small! ApJ 779 (2013), arXiv:1210.7231 (KTS et al.) ApJ 799 (2015), arXiv:1408.3161 (George, … KTS et al.)
Kyle Story
11
arXiv:1511.02843 (Flender et al.)
There is another way to detect the kSZ: the pairwise estimator Pairwise Estimator:
where
(Ferreira 99)
Key Concept: On average, clusters move towards each other under their mutual gravitational attraction. Method: subtract CMB temperatures at the location of pairs of galaxy clusters. Kyle Story
12
Figure credit: T. Giannantonio
There is another way to detect the kSZ: the pairwise estimator Pairwise Estimator:
where
(Ferreira 99) First measurement: ACT + BOSS (spec-z) (Hand et al. 2012)
Probes a combination of cluster physics and cosmology:
where b: cluster bias, τ: cluster optical depth, f: growth rate, σ8: normalization of matter power spectrum
~3σ (PTE 0.2%)
Kyle Story
13
Figure credit: T. Giannantonio
Analytic template for the pairwise kSZ (B. Soergel) Pairwise kSZ probes mass-weighted mean pairwise velocity v12(r)
(Keisler & Schmidt 2012, Schmidt 2010)
b: mass-weighted mean bias ξδv : density-velocity correlation function, ξ : density auto-correlation function
Pairwise kSZ amplitude:
In linear regime:
arXiv:1603.03904 (Soergel, Flender, KTS et al.)
Slide from B. Soergel
Ingredients to the pairwise kSZ estimator:
2h
22
h
6h
4h
0h
-40° -50° -60°
Catalog of clusters from DES
Kyle Story
CMB temperatures from SPT
DES Y1 Cluster Catalog: redMaPPer
(E. Rozo, E. Rykoff)
• redMaPPer Y1 mask: ~1,200 deg2 •~6,700 clusters with N>20 gal. • Photoz: σz/(1+z) ~ 0.010 -‐ 0.015
Number of gal. in each cluster Kyle Story
Redshift uncertainties arXiv:1603.03904 (Soergel, Flender, KTS et al.)
SPT CMB Temperature data
~10°
Apply a matched filter to the CMB data. • The filter is a combination of the expected cluster profile and the instrumental beam. • This filter is convolved with the CMB map • Each pixel in the filtered map is the amplitude of the best-fit profile centered at that pixel.
Unfiltered CMB map Kyle Story
Filtered map
Haehnelt & Tegmark 96
Results: 4.2σ detection! First pairwise detection with photo-z’s
Pairwise Estimator:
where
• • •
DES redMaPPer clusters 6,700 clusters (20 < Ngal < 60) SPT Temperature data
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
18
Null Tests all pass
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
19
We validate our analysis method and study systematics effects using simulations. (S. Flender)
• Full-sky lightcone out to z~1 from N-body simulations.
8
• Gas prescription for hot intra cluster gas • account for astrophysical affects (star formation, feedback , non-thermal pressure) -8 300
100
-300
-100
arXiv:1511.02843 (Flender et al.) Kyle Story
20
Studying Systematics with simulations
Redshift (photo-z) uncertainty reduces the signal at low separations, but the fit remains unbiased.
Contamination from primary CMB and tSZ does not bias results
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
21
Studying Systematics with simulations
Redshift (photo-z) uncertainty reduces the signal at low separations, but the fit remains unbiased.
Contamination from primary CMB and tSZ does not bias results
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
22
Studying Systematics with simulations
Redshift (photo-z) uncertainty reduces the signal at low separations, but the fit remains unbiased.
Contamination from primary CMB and tSZ does not bias results
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
23
kSZ measurements are in the regime of “first detections.” Measurements to date: Pairwise Estimator: 1. ACT x BOSS (spec-z), Hand et al. 2012: PTE = 0.002 (X σ) (rejection of null hypothesis) 2. Planck x SDSS-DR7 CGC (spec-z), Planck Int XXXVII 2015: 1.8 - 2.5 σ 3. SPT-SZ x DES-Y1 (photo-z), Soergel et al. 2016: 4.2 σ (template fit)
Other Estimators: 1. Planck x SDSS-DR7 CGC (spec-z), Planck Int XXXVII 2015: 3.0 - 3.7 σ (cross-corr with velocity reconstruction) 2. ACTpol x BOSS (spec-z), Schaan et al. 2015: 2.9 - 3.3 σ (stack on velocity reconstruction) 3. Planck, WMAP x WISE (photo-z), Hill et al. 2016: 3.8 - 4.5 σ (T^2 cross LSS) Kyle Story
24
Future kSZ measurements will be high S/N. Challenging interpretation, but large potential payoff. 4.2σ detection (statistical significance) of the kSZ from DES clusters and SPT CMB data (first detection from photometric data) Systematics limiting the interpretation: • redshift uncertainties • mis-centering of clusters • cluster profile • theory (halo bias, velocity model) • tSZ contamination • astrophysics (constrain τ and b, measure f?)
S/N Projections for future surveys
tSZ and foregrounds completely removed
arXiv:1603.03904 (Soergel, Flender, KTS et al.) Kyle Story
25
arXiv:1511.02843 (Flender et al. 2015)
Thanks for listening!