MW A LFD Epoch of R eionization P ow er Spectr um Obser vations · 2007-09-18 · MW A !LFD Epoch...

Preview:

Citation preview

MWA–LFD Epoch of Reionization

Power Spectrum Observations

Miguel F. Morales for the MWA–LFD EOR power spectrum collaboration

December 16, 2006 @ Melbourne University

Overview

• EOR power spectrum signal

• Building an EOR machine: MWA–LFD

• Foreground subtraction

0 (now)~6CMBBigBang

EOR

redshift

EOR power spectrum signal

0 (now)~6CMBBigBang

EOR

redshift

H& during EOR

100

200

300

400

500

7

7.58

8.59

Mpc/h

z

sv_z8_zeta10_ME.fts_0

Oliver Zahn

Statistical EOR detection

f

iy

ImageCube

ix

v

f

Visibilities

u

v

h

FourierRepresentation

u

FT SkyCoordinates

FTFrequency

Morales & Hewitt (2004)

Spherical symmetry

v

u

h

Morales & Hewitt (2004)

EOR power spectrum

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

Furlanetto, Zaldarriaga, Hernquist (2004a,b)Bowman, Morales & Hewitt (2005)

EOR power spectrum

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

Furlanetto, Zaldarriaga, Hernquist (2004a,b)Bowman, Morales & Hewitt (2005)

EOR power spectrum

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

Furlanetto, Zaldarriaga, Hernquist (2004a,b)Bowman, Morales & Hewitt (2005)

EOR power spectrum

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

Furlanetto, Zaldarriaga, Hernquist (2004a,b)Bowman, Morales & Hewitt (2005)

EOR power spectrum

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

Furlanetto, Zaldarriaga, Hernquist (2004a,b)Bowman, Morales & Hewitt (2005)

Kaplinghat (2005)

Building an EOR machine: the MWA–LFD

Challenges 1: Foregrounds• Faint point sources

• Smooth galactic emission

• Galactic radio recombination lines

• RFI

• Others!

Challenges 2: Ionosphere & Polarization

325 MHz polarized flux, 6° x 6°,4’ beam, 5 K peaks (de Bruyn)

Lonsdale (2004)

Step 1: increase sensitivity

Survey speed with compact antenna distribution

1

Table 1.

At|dA At|NAdA|At B |u| t

Power Spectrum S/N A2t A

3/2t (dA)!1/2 ! FOV B1/2 |u| n(|u|) t

Note. — This table lists the scaling relationships of the key equations. In order, the variables in each column are: total array areaholding the size of each antenna constant At|dA (adding antennas), total array area holding the number of antennas and distributionconstant At|NA

(increasing antenna size), the size of each antenna with the total array area held constant dA|At (dividing area into moresmall antennas), the total bandwidth B, the sensitivity as a function of wavenumber length |u|, and the total observing time t.

1

Table 1.

At|dA At|NAdA|At B |u| t

Power Spectrum S/N A2t A

3/2t (dA)!1/2 ! FOV B1/2 |u| n(|u|) t

Note. — This table lists the scaling relationships of the key equations. In order, the variables in each column are: total array areaholding the size of each antenna constant At|dA (adding antennas), total array area holding the number of antennas and distributionconstant At|NA

(increasing antenna size), the size of each antenna with the total array area held constant dA|At (dividing area into moresmall antennas), the total bandwidth B, the sensitivity as a function of wavenumber length |u|, and the total observing time t.

Morales (2005)

Step 2: systematics

0

10

20

30

40

50

60

0

10

20

30

40

50

60

!0.5

0

0.5

1

1.5

2

2.5

010

2030

4050

60

010

2030

4050

60−0.5

0

0.5

1

1.5

2

2.5

What we know about fitting algorithms is goodness-of-fit surface quality ö robustness

• Excellent instantaneous point spread function

• Data-rich observations & computation

Realtime software team

S. Doeleman, J. Kasper, D. Mitchell, M. Morales, R. Sault, R. Wayth

R. Cappallo, L. Greenhill, C. Lonsdale, M. Tegmark, M. Zaldarriaga

Core:

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

(Tegmark, Zaldarriaga, Morales)

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

true sky

(Tegmark, Zaldarriaga, Morales)

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

true skyionosphericdistortion

(Tegmark, Zaldarriaga, Morales)

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

true skyionosphericdistortion

Fouriertransform

(Tegmark, Zaldarriaga, Morales)

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

true skyionosphericdistortion

integration overFourier modes to

measured visibilities

Fouriertransform

(Tegmark, Zaldarriaga, Morales)

Laying down the math

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

true skyionosphericdistortion

integration overFourier modes to

measured visibilities

Fouriertransform

noise

(Tegmark, Zaldarriaga, Morales)

Data analysis

m(v) = B(v,u!)F(u!, !!)A(!!, !)I(!) + n(v)

Observation:

I(!) =!

OTN

!1O

"!1

AT (!, !")FT (!",u")BT (u",v)

# $% &

OT

N!1

m(v)

Data reduction, v1:

measure-ment

noise covariancematrix

inversion/deconvolution

integrationpoint

Ionospheric correctionA(!!, !) = "D[!! ! ! ! "!(!, t)]

A(u!,u) =

!d2!ei(u"u

!)·!+iu·"!(!,t)

A(u!,u) !

!d2!!ei(u"u

!)·!!

(1 + iu · "!(!!, t) + · · · )

! (2!)2"D[u " u!] + iu ·

!d2#ei(u"u

!)·!"#(#, t)

A(u!,u) = !D[u ! u!] + i

!

m

amu!D[u ! u!! bm] + cross terms

I(!) =!

OTN

!1O

"!1

AT (u,u")BT (u",v)

# $% &

OT

N!1

m(v)

Data reduction, v2:(Zaldarriaga & Morales; Morales et al.)

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

BT (u!

,v)

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

BT (u!

,v)

AT (!, !!)FT (!!,u!)

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

BT (u!

,v)

AT (!, !!)FT (!!,u!)

AT (u,u

!)

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

BT (u!

,v)

AT (!, !!)FT (!!,u!)

AT (u,u

!)

!

OTN

!1O

"

!1

Correlator

Receiver

Node

Correlator Boards

Realtime Computer

Visibility Binner

Calibrate &

Integrate

Visibilities to

2-8 sec. in

baseline rings

(2 Gvis, 8 Gops)

Bin subset of

baselines at full

freq. & time

resolution to 2°

FOV (100 M u,v,f,p

cells, 250 Mops)

Combine vis.

in freq. to 32

kHz resolution

(500 Mvis, 1/2

Gops)

125 Mvis/s 0.5 GB/s

50 MB/s

EOR Realtime

System –

integrates to 10

min. and adds to

database

Sort into per

antenna sets [500]

& beamform each

in direction of 100

calibrators (25

Gops)

Mapper

Ionospheric

Calibration

ASM v1

Survey

Instrumental

Calibration

Faraday Rotation

Solar Imaging,

2° FOV, 32 kHz

every 8 sec.

(75 Mops)

1 GB/sec.

1.7

GB

/s~

1/2

Gvis

/s

125 M

vis

/s

Oversampled

map of primary

FOV, every 8

sec. (4 pol., 32

kHz, 3.6 Gcells

3.5 Gops)

Map to RA Dec

at array res.

(96 kHz, I pol.,

every 8 sec.,

0.5 Gops)

Map to RA Dec

at array res.

(32 kHz, 4 pol.,

every ~100

sec. 0.5 Gops)

GSM

Prediction

EOR Imaging

EOR Power

Spectrum

SurveyPostage

stamps towards

calibrators (4

pol. 32 kHz, <<

1 Mops)

Solar Imaging

Legend

Software Block

Hardware Block

Parallel Software

Block

Output Broadcast

IPS Power

Detection

Pulsar

Interface

16 GB/s – 2 Gvis every 1/2 sec.

Voltage

Beamformer

16 voltage beams, 2 pol., 1 GB/s

Monitor &

Control

Antenna

Beamformer

Bin to primary

beam FOV & 32

kHz channels

(1.8 G u,v,f,p cells,

~10 Gops)

Bin towards 100

calibrators, image

postage stamps, 1

pol. (outside

primary FOV) 25

Gops

Clock

IPS

Pulsars

Data Transport Layer

32 GB/s

Faraday Rotation

Solar Imaging

Realt

ime S

cie

nce

P

ack

ag

es

Offl

ine S

cie

nc

e

Pa

ck

ag

es

MWA-LFD Dataflow Diagram

Realtime Common Data Processing

Shadow indicates

workpackage

GPS

GSM

v2

System Diagram

m(v)

BT (u!

,v)

AT (!, !!)FT (!!,u!)

AT (u,u

!)

!

OTN

!1O

"

!1

Calibration

Instrumental calibration

Instrument:

Gain from one antenna to rest of array, simultaneously for all antennas & 100 sources

Ionospheric calibration

Up to 1500 sources with known location, fit every 8 seconds ➙ rubber sheet

(Doeleman, Ting)

Ionospheric calibration

Up to 1500 sources with known location, fit every 8 seconds ➙ rubber sheet

(Doeleman, Ting)

1995MNRAS.274..324G

EOR foreground subtraction

EOR collaboration

J. Hewitt, M. Morales, R. Webster, F. Briggs, C. Carilli, S. Furlanetto, L. Greenhill,

L. Hernquist, A. Loeb, C. Lonsdale, A. de Oliveira-Costa, M. Tegmark, S. Wyithe, M. Zaldarriaga

Stages of astrophysical foreground subtraction

• Bright source & RFI subtraction

• Spectral fitting

• Parameter estimation, using power spectrum templates

Spectral fitting

Morales, Bowman & Hewitt (2006)

Spectral foreground errors

!Ps(!)" = 2!d"B2

!

"2a

#2!4+

"2

b

#2!2+ "2

c!$k(!)

"

!S(f) = !a df2 + !b df + !c

Parameter fitting

v

u

h

v

u

h

• Use templates to separate EOR signal from residual foreground subtraction errors

• Fit both local parameters & their ensemble errors

Morales, Bowman & Hewitt (2006)

MWA-LFD power

spectrum sensitivity

Bowman, Morales & Hewitt (2005)

z = 8, 360 hours of integration

0

1

2

3

4

5

6

7

8x 108

P [J

y2 Hz2 ]

0

1

2

3

4

5

6

7

8x 108

10-3 10-2 10-1 100 10110-1

100

101

102

103

k [Mpc-1 ]

(k3 P

/ 2π

t2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

MWA-LFD sensitivity vs. redshift

!

"

#!

#"$%#!

&

'%()*+%,-+.

!

"

#!

#"$%#!

&

-%/%0

#!!+

#!!

#!!#

#!!

#!#

#!+

1%(234!#.

516%'%7%+!+8#7+%(234!67+%)*%,-.

#!!6

#!!+

#!!#

#!!

#!!#

#!!

#!#

#!+

!

#

+

6

9

"

0

:

&$%#!

&

!

#

+

6

9

"

0

:

&$%#!

&

-%/%&

#!!+

#!!

#!!#

#!!

#!#

#!+

1%(234!#.

#!!6

#!!+

#!!#

#!!

#!!#

#!!

#!#

#!+

!

!;"

#

#;"

+

+;"

6$%#!

&

!

!;"

#

#;"

+

+;"

6$%#!

&

-%/%#!

#!!+

#!!

#!!#

#!!

#!#

#!+

1%(234!#.

#!!6

#!!+

#!!#

#!!

#!!#

#!!

#!#

#!+

!

"

#!

#"$%#!

:

!

"

#!

#"$%#!

:

-%/%#+

#!!+

#!!

#!!#

#!!

#!#

#!+

1%(234!#.

#!!6

#!!+

#!!#

#!!

#!!#

#!!

#!#

#!+

Bowman, Morales & Hewitt (2005)

MWA-LFD vs. MWA+,,,

0

1

2

3

4

5

6

7

8x 108

P [J

y2 Hz2 ]

0

1

2

3

4

5

6

7

8x 108

10-3 10-2 10-1 100 10110-1

100

101

102

103

k [Mpc-1 ]

(k3 P

/ 2π

2 )1/2 [M

pc-3

/2 J

y Hz

]

10-3 10-2 10-1 100 10110-1

100

101

102

103

angu

lar o

nly

MWA-LFD

http://web.haystack.mit.edu/arrays/MWA/index.html

• ~$., million instrument, funded

• Data taking starting in late /,,0

• Premier EOR instrument• Strict attention to systematic errors

Recommended