The Cosmic Radio Background

Douglas Scott+ Tessa Vernstrom & Jasper Wall

Astrophysics from radio to submillimetreBologna, February 2012

EUROPE: A probe designed to delve into the“Big Bang” that created the cosmos hasuncovered an enigmatic fog of microwaveradiation in the centre of our galaxy.European astronomers reported on Monday

PS Why don’t you get on with your talk!

The Cosmic Radio Background

Douglas Scott+ Tessa Vernstrom & Jasper Wall

Astrophysics from radio to submillimetreBologna, February 2012

The Cosmic Microwave BackgroundThe Cosmic Radio Background

Best blackbody in the Universe

(better than you can buy at Bob’sBetter Blackbody Boutique)

The Cosmic Microwave BackgroundCosmic Backgrounds

BLAST/SPIRE

ARCADE 2

Radio countsARCADE 2 − CMB

Absolute Radiometer for Cosmology, Astrophysics, and Diffuse Emission (PI: Al Kogut)Cooled system, compares sky with internal reference6 frequencies, 3−90 GHz (3−100 mm)Balloon payload, ARCADE 2 flight in 2006Found excess over CMB + CRB from known sources

ArXiv: 0901.0546; 0555; 0559; 0562

ARCADE 2 results

!T ! 1.2 K

!

!

1 GHz

"

!2.6

ARCADE 2 results

Excess measured over CMB

Most significant at 3 GHz

Could it be genuinely diffuse gas - probably not!

Dark matter signature - not!

What kinds of sources could produce it?

Let’s look at what’s known about radio counts

May as well do if for each frequency band

→ Vernstrom, Scott & Wall (2011)

(Also ground-based TRES measurements in 2008)

The Cosmic Radio Background

Vernstrom, Scott & Wall arXiv:1102.0814

Euclidean normalization(flat = no evolution)

One less factor of √S(contribution to CRBper log interval in S)

The Cosmic Radio Background

Vernstrom, Scott & Wall arXiv:1102.0814

Euclidean normalization(flat = no evolution)

One less factor of √S(contribution to CRBper log interval in S)

(higher frequency bands)

The Cosmic Radio Background

Vernstrom, Scott & Wall arXiv:1102.0814

4 Vernstrom et al.

Table 2. !2 values for best fits at each of the frequencies

" !2 Degrees of A0 A1 A2 A3 A4 A5

MHz Freedom

150 68 45 6.58 0.36 !0.65 !0.19 0.26 0.099325 59 34 5.17 0.029 !0.11 0.36 0.17 0.20408 66 44 4.13 0.13 !0.34 !0.003 0.035 0.01610 75 59 3.02 0.71 0.97 0.91 0.28 0.0281400 4230 196 2.53 !0.052 !0.020 0.051 0.010 !0.00134800 32 47 1.95 !0.076 !0.15 0.020 0.0029 !0.000798400 41 29 0.79 !0.10 !0.23 !0.051 !0.019 !0.0029

Table 3. Values of the integrated sky brightness and tempera-ture contribution from radio source count for di!erent frequencybands. The uncertainties are 1# limits determined from Markovchain polynomial fits to the data. The high and low extrapolationsare discussed in the text.

" "I! T $T Extrapolated THigh Low

MHz W m!2 sr!1 mK mK mK mK

150 1.8 "10!14 17800 300 29400 18100325 2.1 "10!14 2800 600 5040 3100408 2.9 "10!14 1600 30 3000 1850610 4.2 "10!14 710 90 1200 7401400 7.5 "10!14 110 20 180 1104800 8.0 "10!14 3.2 0.2 10.8 6.78400 9.6 "10!14 0.59 0.05 3.0 1.9

the temperatures relative to the 1.4GHz count. This takesthe form of

T (!) = A! !1.4GHz

"", (4)

where A is the power-law amplitude, and " the index. Weset A to the 1.4GHz value of 0.110 K, while Monte CarloMarkov Chains (reference section 3.2) were used to find thebest value of " = –2.28 ± 0.02. The results of this power-lawfit can be seen in Fig. 2. Once we have this normalizationto the 1.4GHz curve we can extend the limits of integrationfor the 1.4GHz data, constraining the end behaviour of thepolynomial. We make the assumption that the counts fall o!after the end of the available data, using a choice of eithera steeper or shallower slope, to obtain both a high and lowestimate. To achieve this, artificial data points were addedpast where data are available, and the positions of thesepoints were varied until fits were obtained with the desiredend behaviour with reasonable slopes, while still making surethe curve fit the shape of the data, i.e. peaking in the ap-propriate place. These slopes were chosen to be the mostreasonable steep and shallow estimates, with the #2s beinga factor of 5 and 7 greater than the best fit to the data alone.The higher estimate could have been allowed to have an evenshallower slope, therefore allowing for an even higher back-ground estimate; however anything much shallower than thechosen fit would have #2 values several times larger than thebest fit to just the data. This fact makes any shallower fitsan unreasonable choice. The best fits for the extrapolationscan be seen in Fig. 3. In this figure the dashed line (the

Figure 2. Integration results and best fit power-law from Equa-tion 4.

higher estimate) is the shallower slope which falls o! moreslowly after the end of the data, while the solid line is asteeper slope fit to the data.

With these high and low extrapolation estimates for the1.4GHz data, we use Equation 4 to obtain estimates for theother frequencies. The steep slope estimate for the 1.4GHzdata ends up giving nearly the same result for the back-ground temperature as the unextrapolated estimate, due tothe fact that, while the limits of integration have been ex-tended, we controlled the end behaviour such that it falls o!steeply after the available data, whereas in the unextrapo-lated estimate the end of the curve is allowed to rise. Thelower extrapolation is thus essentially what would happenat the other frequencies if the shape of the curve were thesame as at the 1.4GHz fit.

The results of the extrapolated estimates are also givenin Table 3. Since even these reasonable extrapolations canchange the background estimates by about a factor of 2, thisshows how important it is to continue to push the counts tofainter limits.

3.2 Uncertainty – Monte Carlo Markov Chains

To investigate the uncertainties thoroughly, we carried outour fits with Monte Carlo Markov chains for each of the datasets, using CosmoMC (Lewis & Bridle, 2002) as a genericMCMC sampler. The #2 function was sampled for each setusing the polynomial in equation 1, which was then fed tothe sampler to locate the #2 minimum. Each of the six pa-

c# 2011 RAS, MNRAS 000, 1–9

Radio count lower limitsto background

About the right slopebut low by ~3−7

The Cosmic Radio Background

Vernstrom, Scott & Wall arXiv:1102.0814

Radio count lower limits

ARCADE (extrapolated)

ARCADE

What kind of sources?

Vernstrom, Scott & Wall arXiv:1102.0814

8 Vernstrom et al.

Figure 7. The 1.4 GHz source count data. The solid line givesthe best fit to the data while having a moderate slope at thefaint end, while the thick solid line is our best fit to the 1.4 GHzdata from Section 2. The other three lines are bumps peaking at7.9 (dotted), 5.0 (dashed), and 3.1 (dot dash) µJy which producethe background temperature necessary to match the ARCADE 2results. On this plot the height of such a bump is proportional to

S1/2peak

necessary contribution to the background. We saw that abump could exist in this range, peaking at fluxes as brightas 8 µJy, and could integrate up to the excess emission of ±320mK, with a height that is consistent with the data.

We still have no direct evidence that such a new popu-lation exists, and so further investigation into the faint endof the counts is needed. The infrared and radio connectioncould be used to test this idea through use of signal stackingand by examining di!erent possible luminosity functions tolook at the evolution of such a population. The final answermay only be reached when source count data available in theµJy range, perhaps in the era of the EVLA and eventuallythe SKA.

ACKNOWLEDGMENTS

This research was supported by the Natural Sciences andEngineering Research Council of Canada.

REFERENCES

Altschuler D. R., 1986, Astron. Astrophys. Suppl. Ser., 65,267

Benn C. R., Grue! G., Vigotti M., Wall J. V., 1982, MN-RAS, 200, 747

Blake C., Wall J., 2002, MNRAS, 337, 993Bondi M., Ciliegi P., Schinnerer E., Smolcic V., Jahnke K.,Carilli C., Zamorani G., 2008, ApJ, 681, 1129

Bondi M., Ciliegi P., Venturi T., Dallacasa D., Bardelli S.,Zucca E., Athreya R. M., Gregorini L., Zanichelli A., LeFevre O., Contini T., Garilli B., Iovino A., Temporin S.,Vergani D., 2007, Astronomy & Astrophysics, 463, 519

Bondi M., Ciliegi P., Zamorani G., Gregorini L., VettolaniG., Parma P., de Ruiter H., Le Fevre O., Arnaboldi M.,Guzzo L., Maccagni D., Scaramella R., Adami C., Bardelli

S., Bolzonella M., Bottini D., Cappi A., 2003, Astronomy& Astrophysics, 403, 857

Brandt W. N., Hasinger G., 2005, Ann.Rev.Astr.Ap., 43,827

Bridle A. H., Davis M. M., Fomalont E. B., Lequeux J.,1972, A.J., 77, 405

Ciliegi P., McMahon R. G., Miley G., Gruppioni C.,Rowan-Robinson M., Cesarsky C., Danese L., Frances-chini A., Genzel R., Lawrence A., Lemke D., Oliver S.,Puget J., Rocca-Volmerange B., 1999, MNRAS, 302, 222

Condon J. J., Mitchell K. J., 1984, A.J., 89, 610de Zotti G., Massardi M., Negrello M., Wall J., 2010, As-tronomy & Astrophysics Reviews, 18, 1

Donnelly R. H., Partridge R. B., Windhorst R. A., 1987,ApJ, 321, 94

Feretti L., Burigana C., Enßlin T. A., 2004, New Astron-omy Reviews, 48, 1137

Fixsen D. J., 2009, ApJ, 707, 916Fixsen D. J., Kogut A., Levin S., Limon M., Lubin P., MirelP., Sei!ert M., Singal J., Wollack E., Villela T., WuenscheC. A., 2009, ArXiv e-prints

Fomalont E. B., Kellermann K. I., Cowie L. L., Capak P.,Barger A. J., Partridge R. B., Windhorst R. A., RichardsE. A., 2006, Ap.J. (Suppl), 167, 103

Fomalont E. B., Kellermann K. I., Partridge R. B., Wind-horst R. A., Richards E. A., 2002, A.J., 123, 2402

Fomalont E. B., Kellermann K. I., Wall J. V., Weistrop D.,1984, Science, 225, 23

Garn T., Green D. A., Riley J. M., Alexander P., 2008,MNRAS, 387, 1037

Gervasi M., Tartari A., Zannoni M., Boella G., Sironi G.,2008, ApJ, 682, 223

Gregory P. C., Scott W. K., Douglas K., Condon J. J.,1996, Ap.J. (Suppl), 103, 427

Grue! G., 1988, Astronomy & Astrophysics, 193, 40Gruppioni C., Ciliegi P., Rowan-Robinson M., Cram L.,Hopkins A., Cesarsky C., Danese L., Franceschini A., Gen-zel R., Lawrence A., Lemke D., McMahon R. G., Miley G.,Oliver S., Puget J., Rocca-Volmerange B., 1999, MNRAS,305, 297

Haarsma D. B., Partridge R. B., 1998, ApJl, 503, L5+Hales S. E. G., Baldwin J. E., Warner P. J., 1988, MNRAS,234, 919

Hauser M. G., Dwek E., 2001, Ann.Rev.Astr.Ap., 39, 249Henkel B., Partridge R. B., 2005, ApJ, 635, 950Hopkins A. M., Afonso J., Chan B., Cram L. E., Geor-gakakis A., Mobasher B., 2003, A.J., 125, 465

Ibar E., Ivison R. J., Biggs A. D., Lal D. V., Best P. N.,Green D. A., 2009, MNRAS, 397, 281

Katgert J. K., 1979, Astronomy & Astrophysics, 73, 107Katgert P., Oort M. J. A., Windhorst R. A., 1988, Astron-omy & Astrophysics, 195, 21

Kuehr H., Witzel A., Pauliny-Toth I. I. K., Nauber U.,1981, Astron. Astrophys. Suppl. Ser., 45, 367

Lagache G., Puget J., Dole H., 2005, Ann.Rev.Astr.Ap.,43, 727

Lewis A., Bridle S., 2002, Phys. Rev.D, 66, 103511Longair M. S., 1966, MNRAS, 133, 421Longair M. S., Sunyaev R. A., 1969, Ap.J. (Letters), 4, 65Madau P., Pozzetti L., 2000, MNRAS, 312, L9McGilchrist M. M., Baldwin J. E., Riley J. M., TitteringtonD. J., Waldram E. M., Warner P. J., 1990, MNRAS, 246,

c! 2011 RAS, MNRAS 000, 1–9

Lots of scatter!Increasing?

Flat?

Decreasin

g?Bumps like these

would fit ARCADE

1.4 GHz

What kind of sources?

Extra sources fainter than ~10μJy

Could be just below current limits

But have to break the far-IR/radio correlation

Far-IR/radio correlation

1991ApJ...376...95C

Condon, Anderson & Helou 1991

What kind of sources?

Extra sources fainter than ~10μJy

Could be just below current limits

But have to break the far-IR/radio correlation

Search for them using stacking of optical/IR sources?

Improve the counts using careful P(D) approach

Pick 3GHz, where ARCADE is most discrepant

New EVLA data at 3 GHz

Single ~14 arcmin field

In “Lockman Owen” region (with ultra-deep 1.4GHz)

C configuration, FWHM ~ 7.5 arcsec

Expect ~1μJy rms

Use P(D) fitting

PI: Jim Condon

Observations start any day now

Karl G Jansky