1 On the road to discovery of relic gravitational waves: From cosmic microwave background radiation Wen Zhao Department of Astronomy University of Science.

1

On the road to discovery of relic On the road to discovery of relic gravitational wavesgravitational waves: From cosmic microwav: From cosmic microwav

e background radiatione background radiation

Wen ZhaoWen ZhaoDepartment of Astronomy

University of Science and Technology of China

The 6th KIAS Workshop on Cosmology and Structure Formation

2

Question: Why do RGWs Exist? In the curved space-time, the vacuum state at “time 1” naturally

corresponds to the multi-particle state at “time 2”.

Assuming the Universe had a vacuum state of graviton at the initial stage, with the expansion of the Universe, but the radiation-dominant stage, the gravitons were naturally generated. (Grishchuk, 1974)

If the Universe had an exact de-Sitter expansion in the inflationary stage, the generated primordial power spectrum of RGWs is scale-invariant. The amplitude of the spectrum directly depends on the energy scale of inflation: . Chaotic inflations predict r~O(0.1), (Linde, 1986)

3

4

CMB power spectraDensity perturbations Relic gravitational waves

r=1

5

Detection in the CMB

Method a: BB channel But B-polarization is very small. When the noise is large, this channel is useless.

Method b: TT+EE+TE channels Total TT = TT (dp,+) + TT (gw,+) Total EE = EE (dp,+) + EE (gw,+) Total TE = TE (dp,+) + TE (gw,-) This method is limited by cosmic variance. When r<0.05, these three channels will be useless.

6

Detection in the CMB

1. BICEP2

2. WMAP & Planck

3. Forecasts 3.1 Planck B-mode polarization 3.2 Ground-based experiments 3.3 Next-generation satellites

7

BICEP2: Constructed E & B

8

BICEP2: BB Spectrum & RGWs

Alternative Explanations

Recently, Colley & Gott (arXiv:1409.4491) cross-correlated genus statistic

in BICEP2 and Planck 353GHz data, and got:

12

Relic gravitational waves in WMAP (Planck) data

WMAP9 gives: r<0.13 at 95% C.L. (Komatsu et al 2012)

This constraint comes from TT+TE observations.

(Planck)TT + (WMAP9)TE + highL even give: r<0.11 at 95% C.L. (Planck Collaboration, 2013)

BB only gives r<2.1 at 95% C.L. (WMAP Collaboration, 2010)

Why?

1. All TT+TE data, the multipoles till to around 10,000

2. Power-law form for density perturbations is correct at least for four orders in scale

15

WMAP low-multipole TE data

16

Low-multipole TT+TE data (l<100)

1. power-law forms for d.p. and g.w. : (WMAP/Planck case)

2. power-law d.p. and g.w., but l<100 data (Our case)

17

WMAP TT+TE data (l<100)

1. power-law forms for d.p. and g.w.

2. d.p. with running

5. piece-wise form for d.p. and g.w.

3. power-law d.p. and g.w., but l<100 data

4. power-law d.p. and g.w., but l<100 data + SNIa + SDSS

FORECAST:

•Planck B-mode Polarization

•Ground-based and balloon-borne experiments

•Next generations (CMBPOL, COrE, LiteBird et al.)

19

Detecting “B-mode” by “Planck”1. Planck can detect B-mode, if r>0.05.

2. Planck can only detect the reionization peak at l<10.

However, 1) “cosmic variance” is too large:

2) low-multipole TT data 3) depends on the nearly "full-sky"

show a lot of “anomalies”; map (but most regions are dirty)

3. It seems that Planck CANNOT directly confirm or exclude the BICEP2 results at recombination peak (multipole l~80), because the instrumental noises in this region are too large!

4. Indirect comparison may be possible. For example, Planck HF data dust template dust map at 150GHz (BICEP2 region)

comparing with BICEP2 map

20

21

Other CMB experiments

Ground-based experiments: KECK, BICEP3, QUIET, POLARBEAR,

SPTPOL, ABS, QUIJOTE, ACTPOL, QUBIC, CLASS, et al.

Balloon-borne experiments: EBEX, PIPER, Spider, et al.

Survey plan: a small part of full sky for a long time.

Noise level: cosmic lensing limit.

Frequency channels: multi-frequency channels.

http://en.wikipedia.org/wiki/List_of_cosmic_microwave_background_experiments

22

Noise levels

23

Detection limits

24

Fourth Generation: Planned CMBPOL (COrE, LiteBird, PRISM, PIXIE) experiments

25

Testing inflationary consistency relation: nt=-r/8------ key issue for inflation

26

Summary, but not conclusion! Detections of CMB B-mode polarization are mainly at two peaks: 1. recombination peak at l~80; 2. reionizatin peak at l<10. It is the unique window for the low-frequency GW detection.

BICEP2 claimed r=0.2, but other explanation is also possible; WMAP (Planck) low-multipole data favor r~O(0.1) (but large uncertainty).

Forthcoming observations: 1. Planck (alone): recombination peak seems difficult to be detected (limited by instrumental noises); reionization peak is possible (limited by CV, contamination in Galactic plane); cross-correlation with BICEP2 is helpful. But if it can say "YES" or "NO" for BICEP2? 2. Ground-based experiments (KECK, BICEP3, SPTPOL, POLARBEAR, et al.): recombination peak is possible, so long as r>0.01. 3. Next generations (CMBPOL, COrE, LiteBird et al.): detection is possible for r>0.001; inflationary consistency relation is possible to be tested, if r>0.1.

27

THANKS!