Quadrant asymmetry in the Larissa Santos angular distribution of - - PowerPoint PPT Presentation

“The map of what we call reality is an ever-shifting mosaic of ideas.” Marcelo Gleiser

Quadrant asymmetry in the angular distribution of the CMB: what do we know so far?

Larissa Santos University of Science and Technology of China, USTC The early universe, cosmology and fundamental physics

• Beijing. September 16th, 2015

Outline

❖ A very, very brief introduction to CMB ❖ Anomalies in the CMB temperature distribution ❖ The quadrant asymmetry in CMB data ❖ Results for WMAP and Planck ❖ Influence of different masks in the previous results

A very brief introduction

❖ 50th anniversary of the article

that revealed the discovery of CMB

❖ May 1965: Penzias and

Wilson

❖ COBE satellite: discovery of the

CMB temperature anisotropies

2009 Planck

A very brief introduction

❖ 50th anniversary of the article

that revealed the discovery of CMB

❖ May 1965: Penzias and

Wilson

❖ COBE satellite: discovery of the

CMB temperature anisotropies

2009 Planck

Is the universe really isotropic?

❖ In the LCDM model the

universe is homogeneous and isotropic

❖ We analyze the temperature

fluctuations of the CMB

❖ Finally, we must compare

the observations with the theoretical model

Anomalies in the CMB temperature distribution

Low quadrupole amplitude

Concordance model Cosmic variance Low quadrupole amplitude

Spergel et al. (2003), Tegmark et al. (2003)

The alignment between the quadrupole and octopole

❖ The quadrupole and

• ctupole are aligned

❖ The difference between

these two axes is of only 10 degrees.

Tegmark et al. (2003), Oliveira-Costa et al. (2004)

North-south asymmetry and the cold spot

Systematics? Not statistically significant?

Bennett et al. (2011), Bennett et al. (2013) Vielva et al. (2004), Cruz et al. (2007) Eriksen et al. (2004), Hansen et al. (2004)

❖ It was found a cold spot in the sky with a diameter of

approximately 10 degrees

❖ The probability of finding such feature on LCDM

simulations is of approximately 2% according to Planck 2015 results

❖ The north-south asymmetry was discovered between

the Galactic north and south hemispheres

❖ The asymmetry is in disagreement with the standard

cosmological model not exceeding 99% C.L. (Planck collaboration, 2015)

On one hand

❖ Planck data confirms most anomalies found in WMAP data ❖ astro-ph arXiv:1303.5083 ❖ astro-ph arXiv:1506.07135 ❖ Other authors also confirm different anomalies ❖ North-south asymmetry: Bernui et al. (2014), astro-ph arXiv:

1404.2936

❖ Quadrupole/octupole alignment: Polastri el al. (2015), astro-ph

arXiv:1404.2936

❖ Cold spot: Gurzadyan et a. (2014), astro-ph arXiv:1404.6347

On the other hand

❖ Anomaly claims are not statistically significant and there is no

compelling evidence for deviations from the LCDM model - 9 year data release from WMAP team

❖ astro-ph arXiv:1001.4758 ❖ No evidence for significant hemispherical anomalies (Quartin et al.,

2015)

❖ astro-ph arXiv:1408.5792 ❖ After subtraction of astrophysical and second oder effects, only the low

quadrupole may still be considered anomalous (Rassat et al., 2014)

❖ astro-ph arXiv:1405.1844

The quadrant asymmetry

❖ Comparing quadrants in the sky using the two point

correlation function (TPCF)

( )

〉 〈 = ) ( ) (

q p

n n T T C γ

CMB sky vs CDM simulations

Λ

Residual Galactic foregrounds: WMAP KQ85 mask

❖ Available sky fraction = 85%

Step by step

❖ 1000 Monte Carlo simulations were generated using CMB

LCDM best fit power spectrum

❖ We divided the CMB sky foreground-cleaned maps and

each simulated map in 4 quadrants

❖ To avoid residual foregrounds in the CMB maps, we mask

the contaminated pixels

❖ We then calculate the TPCF for each quadrant of every map ❖ Finally, we compare the results between observations and

the model

Results obtained for WMAP data

Northwestern Quadrant (NWQ) Northeastern Quadrant (NEQ) Southeastern Quadrant (SEQ) Southwestern Quadrant (SWQ)

Planck first released mask

❖ Available sky fraction (mask-rulerminimal) = 83.65%

Comparing WMAP and Planck first data release

Statistics for Planck1

∑

=

=

bins

N i i bins

f N

1 2

1 σ

Mean value for the simulations Value for the Planck map SMICA1

❖ The excess of power in the SEQ occurs in 19.3% of the

MC simulations and the lack of power in the NEQ

• ccurs with a only 0.5% chance considering the LCDM

model

❖ The probability that the asymmetry SEQ/NEQ found in

the data happen in the simulations is of 0.8%

Planck 2015 data release (SMICA2)

❖ The excess of power found in the CMB SEQ occurs in

27.1% of the MC simulations and the lack of power in the NEQ occurs with a only 0.6% chance considering the LCDM model

❖ The probability that the asymmetry SEQ/NEQ found in

the data happen in the simulations is of 1.4%

❖ The new result is consistent with previous ones

What are these anomalies?

❖ Are there any explanations for these anomalies? ❖ Could we still think about systematics? ❖ Maybe foregrounds? ❖ New physics? ❖ Are they statistically independent or do they have a

common origin?

❖ Are they actually even statistically significant?

Quadrant asymmetry and the low quadrupole amplitude

❖ We generated simulations with a modified LCDM seed

power spectrum assuming C2 = C2WMAP and C3 = C3WMAP

❖ The probability that the asymmetry SEQ/NEQ found in

the data happen in this new set of simulations was calculated in approximately 3%

❖ We conclude that it is unlikely that the quadrant

asymmetry is related to the low quadrupole amplitude

Quadrant asymmetry and the cold spot

R = 50 R = 100 R = 150

Masking the cold spot

❖ The excess of power in the SEQ is not due to the

presence of the cold spot

Rotating the axis with respect to z direction

Anti-clockwise rotation performed in the SEQ Clockwise rotation performed in the SEQ

❖ The TPCF reaches its highest value for a 5-degree

clockwise rotation

❖ Its value increases 3% in comparison to the previous

chosen SEQ quadrant

❖ The excess of power on the TPCF starts to decrease as

we rotate the sky above 10 degrees

❖ The rotations in the anti-clockwise direction show a

more evident dissipation of the excess of power as the rotation angle increases

σ

Foregrounds: testing different sky cuts

❖ A more severe mask (U73): fsky = 74.83

Results

❖ Overall results seem consistent with the ones previously

found using WMAP masks and Planck mask- rulerminimal

❖ We can still see an excess of power in the SEQ for

CMB observations

❖ The lack of power in the NEQ CMB observations is

still there

The new Planck mask

❖ Planck 2015 mask UT78: fsky = 78.67

Unexpected results using UT78

Consistency check with other 2015 CMB Planck maps

❖ The excess of power in the SEQ becomes bigger and

more anomalous, decreasing its probability to occur in the LCDM model to 10.9%

❖ On the other hand, using UT78 mask, the NEQ is no

longer anomalous.

❖ Can we considere UT78 a good mask? It is in opposition

with all the other tested masks.

Mask-rulerminimal+UT78

❖ Consistent results with mask-rulerminimal alone and U73

alone

❖ fsky = 74.6

❖ We combined UT78 with the other masks, calculated the

TPCF for each resulting quadrant and its correspondent value

❖ The new results are in agreement with the previous ones

using mask-rulerminimal and U73 alone

❖ The lack of correlation in the NEQ is still present for the

combined masks

Performing some more tests

❖ Calculating the histograms in the regions of CMB map

which is covered by the other masks and uncovered ny UT78

UT78-mask-rulerminimal UT78-U73

❖ We calculated the statistics of the histograms (kurtosis and

skewness) for each quadrant

❖ We found that the value for the kurtosis in the NEQ in these

regions is always above 3 for the data and in average not bigger than 2.42 for the simulations

❖ We found a particularly high kurtosis for the case UT78-U73: 4.94 ❖ We did not find any simulated map among 1000 with such value

• f kurtosis in the same region of the sky

❖ The results suggests that UT78 leaves residual foregrounds in the

data unmasked, if so, being unsuitable for cosmological analysis

Is the power asymmetry dependent of our previous choice of quadrants?

❖ We considered circular regions in CMB map as well as

in simulations

❖ We tested different radius for each choose region ❖ For a direct comparison with the previously chosen

quadrants (in number of pixels), we restricted ourselves to radius that run from from 60 to 80 degrees.

φ,θ

( ) = 270!,135!

( )

φ,θ

( ) = 270!,45!

( )

φ,θ

( ) = 225!,45!

( )

❖ We found that the biggest excess of power falls in the

region centered at and radius of 80 degrees

❖ We found significant lack of power for regions centered

at and for radius 60 and 70 degrees, respectively.

❖ Both the excess of power and the lack of power shown

above have low probability to occur in the simulations (between 1% and 5%).

φ,θ

( ) = 270!,135!

( )

φ,θ

( ) = 270!,45!

( )

φ,θ

( ) = 225!,45!

( )

❖ These results are in agreement with the previous ones

when we divided the sky in quadrants

❖ Finally, as expected, we concluded by choosing different

regions in the CMB sky that the power asymmetry is not dependent of our previous choice of quadrants

So, what is the explanation?

❖ Attempt to explain these features in terms of systematics or

emission of local astrophysical sources have not been successful.

❖ Some phenomenological models have been suggested to

account for the observations

❖ So far, they don’t provide a complete and satisfactory

explanations for the so called anomalies in the CMB temperature distribution

Thank you!