water vapour radiometry at alma: properties of the...

Water Vapour Radiometry at ALMA: Properties ofthe Atmosphere and First Tests of Phase Correction

B. Nikolic, R. Bolton & J. S. Richer

Cavendish Laboratory/Kavli InstituteUniversity of Cambridge

8 March 2011

(University of Cambridge) WVR phase correction at ALMA March 2011 1 / 52

Atacama Large Millimetre Array (ALMA)Artist’s vision in 2006


ALMA AOS in October 2010from http://wikis.alma.cl/bin/view/AIV/CommissioningStatusReview

Photo: N. Mizuno


http://wikis.alma.cl/bin/view/AIV/CommissioningStatusReview

ALMA AOS


ALMA Transporter in action


ALMA 7 m antenna


ALMA test images released by JAO

• This shows the well-known spiral NGC253, with an optical image of the whole galaxy on the left (credit: ESO). The ALMA test images show dense clouds of gas in the central regions of the galaxy: (middle) the CO J = 2-1 line at 230 GHz and (right) the continuum and CO J = 6-5 line at 690 GHz. The small ellipse in the bottom left hand corner of each of the ALMA test images indicates the angular resolution – that shows the size of the beam formed when the signals from the antennas are combined together. There were six antennas for the 230 GHz observations and only four for those at 690 GHz.



• An example of ALMA’s potential as a spectroscopic instrument: on the left is the map of the molecular “hot core” G34.26+0.15, which is unresolved with the short baselines that we are presently using, so the “image” is not very interesting whereas a section of the spectrum near 100 GHz shows a “forest” of molecular lines. A few of the chemical species that are responsible for the emission lines are identified on the plot.(University of Cambridge) WVR phase correction at ALMA March 2011 8 / 52


• This shows the emission from the disk of dust surrounding the star Beta Pictoris. On the left is an image at 70 microns wavelength made with Herschel, (Olofsson et al., SDP Presentations, Madrid, Dec 2009) and on the right is the ALMA test data at 870 microns showing the denser material in the central region of the disk. At this distance 5 arcseconds corresponds to 100 times the radius of the Earth’s orbit around the Sun, or about twice the radius of the “Kuiper belt” surrounding the Solar System, which contains many dwarf planets and also some dust, but much less than in the disk around Beta-Pic. The disk is very thin and we are viewing it edge on – in both observations the apparent thickness is a reflection of the angular resolution of the instrument.(University of Cambridge) WVR phase correction at ALMA March 2011 9 / 52

Introduction

Outline

1 Introduction

2 Phase correction software

3 First results at ALMAGood examplesPoor examplesEffect on the beam“Dry” fluctuations

4 Summary


Introduction

Atmospheric Phase Fluctuations

R(t)

The turbulenttroposphere

Astronomicalwavefront

Corruptedastronomicalwavefront


Introduction

Path fluctuations due to the atmosphere

−750

−500

−250

0

250

500

750

δL(µ

m)

δL(µ

m)

6 6.2 6.4 6.6 6.8 7 7.2

t (hours UT)t (hours UT)

0.009

0.0092

0.0094

0.0096

0.0098

|A|

|A|


Introduction

Phase closure (→ antenna based errors)Antenna 0 Vs 1 Antenna 0 Vs 2

−2000

−1500

−1000

−500

0

500

δL(µ

m)

δL(µ

m)

7.1 7.15 7.2 7.25 7.3 7.35 7.4 7.45


0

250

500

750

1000

1250

δL(µ

m)

δL(µ

m)

7.1 7.15 7.2 7.25 7.3 7.35 7.4 7.45


Antenna 1 Vs 2 Closure phase

−3500

−3000

−2500

−2000

−1500

−1000

δL(µ

m)

δL(µ

m)

7.1 7.15 7.2 7.25 7.3 7.35 7.4 7.45


−1000

−500

0

500

1000

δL(µ

m)

δL(µ

m)

7.1 7.15 7.2 7.25 7.3 7.35 7.4 7.45



Introduction

Atmospheric transmission + ALMA bands1 mm precipitable water vapour; absorption is dominated by H2O, O2

Band 2

Band 3 Band 4

Band 5

Band 6

Band 7

Band 8 Band 9

Band 10

0

0.2

0.4

0.6

0.8

1

T xT x

100 200 500 1000

ν (GHz)ν (GHz)


Introduction

Water Vapour cm/mm/sub-mm lines1 mm precipitable water vapour

0

50

100

150

200

250

300

Tb

(K)

Tb

(K)

200 400 600 800 1000

ν (GHz)ν (GHz)


Introduction

Water Vapour cm/mm/sub-mm lines1 mm precipitable water vapour

0

50

100

150

200

250

300

Tb

(K)

Tb

(K)

200 400 600 800 1000

ν (GHz)ν (GHz)

22 GHz Water Line –previous WVR systems

183 GHz Water Line –ALMA WVR system


Introduction

The 183 GHz Water Vapour LineBlue rectangles are the production WVR filters

0

50

100

150

200

250

T b(K

)T b

(K)

175 177.5 180 182.5 185 187.5 190

ν (GHz)ν (GHz)


Introduction

WVR in the ALMA receiver cabin


Introduction

ALMA 183 GHz Water Vapour RadiometersTechnical concept of the units

To Relay Optics& The Sky

Chop

Cold Load

Hot Load

LO Chain

Input 125 MHz ref-erence signal

Digital synth Amplifier Freq. Multiplier

90 GHzLO Signal

LNA

0.5-2

2-4.5

4.5-6.5

6.5-8.5

Det

ectio

n,Sa

mpl

ing

&R

eado

ut

Readout viaCAN Bus


Introduction

ALMA 183 GHz Water Vapour RadiometersTechnical concept of the units

Un-cooled mixer, double-sideband, with ≈ 1000 K receiver noiseTotal bandwidth ≈ 18 GHz split into four DSB channelsDicke-switched with a chopper wheel against loads at twotemperatures allowing continuous calibrationSpecifications:

Sensitivity: 0.08–0.1 K per channel RMSStability: 0.1 K peak-to-peak over 10 minutes + 10 degree tiltsAbsolute accuracy: 2 K maximum error

Hardware complete!All units delivered to ALMA


Introduction

Sky brightness observed by WVRs

50

100

150

200

250

T B(K

)T B

(K)

6 6.25 6.5 6.75 7 7.25


50

100

150

200

250

T B(K

)T B

(K)

6 6.25 6.5 6.75 7 7.25


50

100

150

200

250

T B(K

)T B

(K)

6 6.25 6.5 6.75 7 7.25


Observed brightness tem-peratures of WVR on thethree antennas involved inthis test observation. Thefour colours in each panelare the four channels of theWVRs.


Introduction

Correlation between WVR signals and path

−1

−0.5

0

0.5

1

∆TB

,1(K

)∆T

B,1

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8

−1.5

−1

−0.5

0

0.5

1

1.5

∆TB

,2(K

)∆T

B,2

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8

−1

−0.5

0

0.5

1

∆TB

,3(K

)∆T

B,3

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8

−0.5

0

0.5

1

∆TB

,4(K

)∆T

B,4

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8


Introduction

Convert WVR signals into path fluctuation estimatesThis allows to correct the phase!

−5000

−2500

0

2500

δL(µ

m)

δL(µ

m)

0 1000 2000 3000 4000

Time (s)Time (s)


Introduction

Raw & WVR corrected inferred path3-minute running mean removed to accentuate atmospheric effects

−300

−200

−100

0

100

200

300

resi

dual

δL(µ

m)

resi

dual

δL(µ

m)

6 6.2 6.4 6.6 6.8 7 7.2


−300

−200

−100

0

100

200

300

δL(µ

m)

δL(µ

m)


Phase correction software

Outline

1 Introduction



4 Summary



Stage I – Retrieval of atmospheric parameters

50

100

150

200

250

T B(K

)T B

(K)

6 6.25 6.5 6.75 7 7.25


−1

−0.5

0

0.5

1

∆TB

,1(K

)∆T

B,1

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8

TB: Four brightnesstemperatures at

one instant in time

p(T,P,c): Priors

Water vapour modelNested samplingBayesian analysis

Atmospheric condi-tions: distributions

for P, T and c

Bayesian Evidence

dT/dL: Phase cor-rection coefficients



Stage II – Conversion to path fluctuation

50

100

150

200

250

T B(K

)T B

(K)

6 6.25 6.5 6.75 7 7.25


−1

−0.5

0

0.5

1

∆TB

,1(K

)∆T

B,1

(K)

−200 −100 0 100 200 300

δL(µm)δL(µm)

0

2

4

6

8

−5000

−2500

0

2500

δL(µ

m)

δL(µ

m)

0 1000 2000 3000 4000

Time (s)Time (s)

dT/dL: Phase cor-rection coefficients

TB: Four brightnesstemperatures as

function of time

Noise estimates

Linear trans-formation (or

something close)Path error toeach antenna

Error estimates



Stage III – Application to visibilities

−5000

−2500

0

2500

δL(µ

m)

δL(µ

m)

0 1000 2000 3000 4000

Time (s)Time (s)

Observed visibilities

Path estimate

Other calibra-tion tables

CASA

Corrected visibilities

Map making,model fitting, etc



Software architecture

CASA dependent code Better I/OCore algorithms

LibAir package

GSL Boost CASA HDF5

BNMin1:inference /

optimisationLibAIR Core LibAIR I/O

for CASALibAIR I/Ofor HDF5

SWIGLibAIRPython

bindings

wvrgcal:main user-

facingapplication

msdump:Extract data

from MSto HDF5



wvrgcal – the ALMA/CASA phase correction applicationSee http://www.mrao.cam.ac.uk/˜bn204/alma/wvrsoft.html

This is the user-facing application that does WVR phasecorrectionFully compatible with official ALMA software:

Reads ALMA/CASA Measurement SetsWrites CASA Gain Calibration (“T”-Jones) tablesCallable and scriptable from CASAInter-operates with CASA facilities like applycal, accum,plotcal & browsetable

Built on top of:LibAIR, the phase correction libraryCASA-Core and CASA libraries for input/output

Public download, as source-code and Linux binariesWorks! – Now we are making it better


http://www.mrao.cam.ac.uk/~bn204/alma/wvrsoft.html


User documentationhttp://wikis.alma.cl/bin/view/AIV/Application

Also: mailing list hosted here in Cambridge: https://lists.cam.ac.uk/mailman/listinfo/mrao-wvrgcal


http://wikis.alma.cl/bin/view/AIV/Application

https://lists.cam.ac.uk/mailman/listinfo/mrao-wvrgcal

https://lists.cam.ac.uk/mailman/listinfo/mrao-wvrgcal

First results at ALMA

Outline

1 Introduction



4 Summary


First results at ALMA Good examples

Outline

1 Introduction



4 Summary



First application – February 2010Courtesy of Al Wooten

02:41:16.8 02:52:48.0 03:04:19.2 03:15:50.4 03:27:21.6

Time

0

50

100

150

Phase

of

Corr

ect

ed D

ata

(d

eg)

This was an observation switching between two quasars. Blue: uncorrectedphase; red: corrected phase



This morningCourtesy of Satoki Matsushita



Data-set A002 Xb9f5d X1: Short baselineRed: uncorrected phase; Blue: corrected phase

−75

−50

−25

0

25

50

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)



Data-set A002 Xb9f5d X1: Long baselineRed: uncorrected phase; Blue: corrected phase

−500

−400

−300

−200

−100

0

100

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)



Data-set A002 Xb9fce X1Similar conditions, shortly after in time compared to the results in the previous slide

−750

−500

−250

0

250

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)



Data-set A002 Xa0705 X1Another example, only short baselines

−60

−40

−20

0

20

Deg

rees

Deg

rees

0 250 500 750 1000 1250 1500

Time (s)Time (s)


First results at ALMA Poor examples

Outline

1 Introduction



4 Summary



Data-set A002 Xba2ed X1Short baselines, leak-through phase fluctuations (offset in coefficients due totime-constant cloud?)

−20

−15

−10

−5

0

5

10

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)

Very short baseline (A0-A1), essentially no phase fluctuations tocorrect




−30

−20

−10

0

10

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)

Also a very short baseline (A0-A2), some atmospheric-like phasefluctuation seen and corrected




−80

−60

−40

−20

0

20

40

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)

Slightly longer baseline (A0-A3): atmospheric phase fluctuationsclearly seen, corrected somewhat but clear “leak-through”




−80

−60

−40

−20

0

20

40

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)

Slightly longer baseline (A1-A3): again WVR correction helps but“leak-through”



Dataset A002 X9c46d X1Very good weather but dominated by instrumentals (BL A0-A1)

−300

−200

−100

0

100

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)



Dataset A002 X9c46d X1Very good weather but dominated by instrumentals (BL A0-A2)

−40

−30

−20

−10

0

10

20

Deg

rees

Deg

rees

0 500 1000 1500 2000

Time (s)Time (s)


First results at ALMA Effect on the beam

Outline

1 Introduction



4 Summary



Inner ALMA pad positionsFrom http://www.alma.cl/˜dbarkats/pad_position_plotter/plots/ALMA_pad_viewer_zoom2.html


http://www.alma.cl/~dbarkats/pad_position_plotter/plots/ALMA_pad_viewer_zoom2.html


Effect of WVR correction on the ALMA beam

No phase correction WVR phase correction

Short observation with very inhomogeneous uv distribution – one antennawas on a long north baseline and others were close together in a cluster


First results at ALMA “Dry” fluctuations

Outline

1 Introduction



4 Summary



Tropospheric phase fluctuations

Refractive index of air n 6= 1:

n − 1 ≈10−6[

αPd

T+ β

Pw

T+ γ

Pw

T 2

]

Pw : Partial pressure of the water vapourT : Temperature of the water vapourFurthermore, the refractive index is a function of frequency (i.e.,the atmosphere is dispersive), especially at sub-mm frequenciesand close to the edges of the bandsHorizontal and line of sight variation in atmospheric propertieslead to phase errors and phase fluctuations



Evidence for ‘dry’ fluctuationsWeak correlation between path and TB fluctuations

25 m baseline 90 m baseline

−0.4

−0.2

0

0.2

∆TB

,1(K

)∆T

B,1

(K)

−1550 −1525 −1500 −1475 −1450

δL(µm)δL(µm)

0

2

4

6

8

10

4.2

4.4

4.6

4.8

5

5.2

∆TB

,1(K

)∆T

B,1

(K)

−25 0 25 50 75 100 125

δL(µm)δL(µm)

0

2

4

6

8

Very dry conditions (0.4 mm PWV)



Evidence for ‘dry’ fluctuationsBaseline length ∼ 25 m

8.0 8.2 8.4 8.6 8.8 9.0 9.2�0.06

�0.04

�0.02

0.00

0.02

0.04

0.06

0.08



Evidence for ‘dry’ fluctuationsBaseline length ∼ 100 m

8.0 8.2 8.4 8.6 8.8 9.0 9.2�0.10

�0.05

0.00

0.05

0.10


Summary

Outline

1 Introduction



4 Summary


Summary

Summary

Working ALMA interferometer, working WVRs!Initial, but fully working, WVR phase correction software – usednow to do phase correction at ALMAPhase correction performance meets (or is very close) to thespecificationsImportant to obtain more test observations, especially:

On long baselinesMore consistent, long-term, measurements

Our work now focusing on:Improving the existing algorithmsUsing more information (e.g., weather stations, ancillaryinstruments)Implementing new algorithms (e.g., the ‘empirical’)Rolling out to users for testing and feed back


Summary

Challenges

Dealing with widely different conditions at different antennas∼15 km baselines∼300 m altitude differenceridges, other topographical features

Dealing with cloudy conditionsNon-linear effects in a single observation

Probably non-negligible on long baselines

Using WVR signals for amplitude calibrationPhase transfer between different sources

What is maximum angular distance/airmass change for transfer?Improve astrometric accuracy?


water vapour radiometry at alma: properties of the...

Documents