The CO Mapping Array Pathfinder (COMAP)

Kieran Cleary (Caltech) AME Workshop, June 23, 2016

1

CO

MA

P

What is COMAP? • Aim: ultimately, to trace LSS at EoR. In short term, focus on lower-redshift science

• CO

• Common star-formation tracer at low redshift

• Bright, even at high z

• Traces photon sources responsible for the ionization

• Emits in line ‘ladder’

• Mapping• Intensity mapping:

• aggregate line emission from unresolved galaxies

• Measure spectra: trace evolution with redshift - ‘tomography’

• Array

• Single-dish, focal-plane array

• All-sky, 1000-element arrays, broadband backends ultimately required

• Pathfinder

• Test technology and focus on intermediate redshift, z~2-3. 2

CO

MA

P

Fig. 1.—(a) Correlation between HCN and IR luminosities in 65 galaxies. Some limits in HCN luminosities are indicated with arrows. (b) Correlation betweenLCO and LIR for the same HCN sample. The sample is divided into luminous infrared galaxies (LIGs) and ultraluminous infrared galaxies (ULIGs) withLIR ! 1011 L" and less luminous ‘‘normal’’ spiral galaxies. The solid lines are the fits to the less luminous galaxies with a slope fixed at unity. A single slope fitsHCN data for both low and high IR luminosities, but not for the CO data.

Fig. 1a

Fig. 1b

Gao & Solomon (2004)

58 Carilli & Walter

Table 2: Factors to calculate L′CO(1−0) from higher–J transitions up to J=5. For

the SMGs and QSOs average values are quoted based on all available literatureestimates. For the CSG the ratio is taken from Dannerbauer et al. (2009) andAravena et al. (2010); their LVG model 1. The values for the Milky Way and thecentre of M82 are from Weiß et al. (2005b) ‘–’ indicates that no well–constrainedvalue is available. We assume the same excitation as for the CSG for the LBGs,SFRGs. We adopt QSO excitation for the RGs and SMG excitation for the 24µmand ERO sources.source SMG QSO CSG MW M82L′CO(2−1)/L

′CO(1−0) 0.85 0.99 0.97 0.5 0.98

L′CO(3−2)/L

′CO(1−0) 0.66 0.97 0.56 0.27 0.93

L′CO(4−3)/L

′CO(1−0) 0.46 0.87 0.2 0.17 0.85

L′CO(5−4)/L

′CO(1−0) 0.39 0.69 – 0.08 0.75

Figure 1: Redshifted frequencies νobs of CO transitions (left) and other keytracers of the starforming ISM (right) as a function of redshift z, followingνobs = νrest/(1 + z). The shaded areas indicate the frequency bands coveredby various telescopes. Highlighted are the ALMA frequency bands as well as the‘high–frequency’ bands of the JVLA. The colored points indicate detection of allhigh redshift (z> 1) lines. The color of the points refer to the different sourcetypes, as explained in the left panel.

Carilli & Walter (2013)

�⌫ �⌫Observed frequency <-> distance

COMAP collaboration

3

CO

MA

P

K. Cleary (PI) Tony Readhead

Tim Pearson James Lamb David Woody

Sarah Church Risa Wechsler

Tony Li Dongwoo Chung

Andrew Harris Joshua GundersenTodd Gaier Charles Lawrence Brandon Hensley

Clive Dickinson Stuart Harper

Hans Kristian Eriksen Ingunn Wehus

COMAP Phase I

4

CO

MA

P

MMIC$HEMT$$amplifiers$

polarizer$

feed

horn$

LO$

IQ$m

ixers$

EX$

EY$

AMPLIFIER$MODULE$ MIXER/IF$BOARD$

BandEpass$filters$ IF

$am

plifiers$

QY$

IX$

QX$

IY$

900$

900$

USBX$

LSBX$

USBY$

LSBY$

polarizer)feedhorn) AMPLIFIER))

MODULE)

MIXER/IF)BOARD)

Observefor2years10mtelescopeatOVRO

NSF AAG

NSF MRI

COMAP Phase I

26-34GHzFocal-planearray19single-pol

pixels

10mtelescope

19x8GHzbackendCASPER-based

COMAP

COMAP Phase I

5

CO

MA

P

MMIC$HEMT$$amplifiers$

polarizer$

feed

horn$

LO$

IQ$m

ixers$

EX$

EY$

AMPLIFIER$MODULE$ MIXER/IF$BOARD$

BandEpass$filters$ IF

$am

plifiers$

QY$

IX$

QX$

IY$

900$

900$

USBX$

LSBX$

USBY$

LSBY$

polarizer)feedhorn) AMPLIFIER))

MODULE)

MIXER/IF)BOARD)

Observefor2years10mtelescopeatOVRO

NSF AAG

NSF MRI

COMAP Phase I

26-34GHzFocal-planearray19single-pol

pixels

10mtelescope

19x8GHzbackendCASPER-based

COMAP

COMAP LNA modulesCO

MA

P

0.0#

5.0#

10.0#

15.0#

20.0#

25.0#

30.0#

26# 28# 30# 32# 34# 36# 38# 40#

Noise&Tem

perature&(K

)&

Freqency&(GHz)&

40 LNA modules already built (JPL R&TD) Using existing MMIC LNA designs (100 nm) Evaluating new 35 nm designs (<12 K noise from 26-34 GHz noise) Nominal Tsys=44 K (assumes 16 K amplifier noise)

Experiment Parameters

7

CO

MA

P

COMAP Parameters Angular resolution (arcmin) ~4’

Number of pixels 19

System temperature (K) 44

Frequency coverage (GHz) 26-30 30-34

Spectral resolution (R) ~800

Final map sensitivity per 4 GHz band in single field (μK) 1.33

-Four fields, each 2.5 sq deg -2,000 hrs observing -Chosen to overlap with galaxy surveys, e.g. COSMOS, DES

Phase I sensitivity

8

CO

MA

P

10�2 10�1 100

k [Mpc�1]

10�2

10�1

100

101

102

k3 P(k

)/2⇡

2[µ

K2 ]

26-30 GHzCO(1-0)z ⇠ 2.8 � 3.4

COMAP 10m

Fiducial⇠ Pullen+2013, B⇠ Righi+2008Pullen+2013, AVisbal+Loeb 2010

10�2 10�1 100

k [Mpc�1]

0

5

10

15

20

P(k

)/�

P(k

)1�

3�

5�

26-30 GHzCO(1-0)z ⇠ 2.8 � 3.4

COMAP 10m (detection significance)FiducialRighi+2008 (approx.)Visbal+Loeb 2010Pullen+2013, APullen+2013, B (approx.)

Tony Li

Single 4-GHz band

Phase I sensitivity

9

CO

MA

P

10�2 10�1 100

k [Mpc�1]

10�2

10�1

100

101

102

k3 P(k

)/2⇡

2[µ

K2 ]

26-30 GHzCO(1-0)z ⇠ 2.8 � 3.4

COMAP 10m

Fiducial⇠ Pullen+2013, B⇠ Righi+2008Pullen+2013, AVisbal+Loeb 2010

10�2 10�1 100

k [Mpc�1]

0

5

10

15

20

P(k

)/�

P(k

)1�

3�

5�

26-30 GHzCO(1-0)z ⇠ 2.8 � 3.4

COMAP 10m (detection significance)FiducialRighi+2008 (approx.)Visbal+Loeb 2010Pullen+2013, APullen+2013, B (approx.)

Tony Li

Single 4-GHz band

Keating et al.

Phase II

10

CO

MA

P

26-34GHzFocal-planearray19single-pol

pixels

COMAP Phase II

26-34GHzFocal-planearray19single-pol

pixels

26-34GHzFocal-planearray19single-pol

pixels

26-34GHzFocal-planearray19single-pol

pixels

26-34GHzFocal-planearray19single-pol

pixels

10mtelescope

10mtelescope

10mtelescope

10mtelescope

10mtelescope

Observefor3years4GHzbackend

5x10mtelescopeatOVRO

Phase II sensitivity

11

CO

MA

P

10�2 10�1 100

k [Mpc�1]

10�2

10�1

100

101

102

k3 P(k

)/2⇡

2[µ

K2 ]

30-34 GHzCO(1-0)z ⇠ 2.4 � 2.8

COMAP 10m: x5 dishes

Fiducial⇠ Pullen+2013, B⇠ Righi+2008Pullen+2013, AVisbal+Loeb 2010

10�2 10�1 100

k [Mpc�1]

0

5

10

15

20

P(k

)/�

P(k

)1�

3�

5�

30-34 GHzCO(1-0)z ⇠ 2.4 � 2.8

COMAP 10m: x5 dishes (detection significance)FiducialRighi+2008 (approx.)Visbal+Loeb 2010Pullen+2013, APullen+2013, B (approx.)

COMAP Phase II

Tony Li

Tony Li

COMAP Phase III

12

CO

MA

P

Observefor2years10mtelescopeatOVRO

NSF AAG

NSF MRI

COMAP Phase I

26-34GHzFocal-planearray19single-pol

pixels

10mtelescope

19x8GHzbackendCASPER-based

COMAP

Ka

KaCOMAP Ku

Foregrounds

13

CO

MA

P

•Continuum foregrounds •Free-free •CMB

•Synchrotron •Point sources

•Foreground continuua fitted out as low-order modes

•Line foregrounds•Other lines redshifted into 26-34 GHz band

•HCN (1-0), CN (1-0), CS (1-0), HCO+

•Brightest line foreground sub-dominant

Foreground

Foreground Galactic synchrotron from the Global Sky model (de Oliveira-Costa et al. 2008).Free-Free emission from the Planck Sky Model (Delabrouille 2013)Point sources (PSM model with limit between strong and faint sources = 100 mJy)CMB (Commander-Ruler CMB intensity map)

Mollview projectionSky at 30 GHz

Galactic free-freeGalactic synchrotron

Faint point sourcesCMB

K_CMB 3

Foregrounds

14

CO

MA

P

•Continuum foregrounds •Free-free •CMB

•Synchrotron •Point sources

•Foreground continua fitted out as low-order modes

•Line foregrounds•Other lines redshifted into 26-34 GHz band

•HCN (1-0), CN (1-0), CS (1-0), HCO+

•Brightest line foreground sub-dominant

10�1 100

10�310�210�1100101102103104105106

�2 (k)

(µK2 )

model A

10�1 100

k (Mpc�1)

10�1100101102

(CO

+HCN

)/CO

Figure 5: 3D spherically-averaged power spectra (left panels) and averageC` over all frequency channels (right panels), for varied HCN line luminos-ity models. Median spectrum values and 95% interval at each ` are shownfor each model, with log-space residuals between uncontaminated and con-taminated CO spectra shown below each spectra plot.

10

COHCNCO + HCN

Figure 5: 3D spherically-averaged power spectra (left panels) and averageC` over all frequency channels (right panels), for varied HCN line luminos-ity models. Median spectrum values and 95% interval at each ` are shownfor each model, with log-space residuals between uncontaminated and con-taminated CO spectra shown below each spectra plot.

10

Dongwoo Chung

AME Science

15

CO

MA

P

COMAP Parameters Angular resolution (arcmin) ~4’

Number of pixels 19

System temperature (K) 44

Frequency coverage (GHz) 26-30 30-34

Spectral resolution (R) ~800

Final map sensitivity per 4 GHz band in single field (μK) 1.33

Mapping of extended areas: - Galactic: NCP, compact targets - Nearby galaxies

1 MHz spectral resolution (higher possible?)

Schedule

• COMAP Phase I fully funded by NSF (AAG + MRI), KISS, Caltech, Stanford, JPL, Miami, Manchester, Oslo

• 4-year project: 2-year build, 2-year observation

• Start observing October 2017!