Radio interferometry at millimetre and sub-millimetre wavelengths Bojan Nikolic 1 & Fr ´ ed´ eric Gueth 2 1 Cavendish Laboratory/Kavli Institute for Cosmology University of Cambridge 2 Institut de Radioastronomie Millim ´ etrique Grenoble ERIS 2009 Oxford, September 2009 Rev 33 B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 1 / 62
71
Embed
Radio interferometry at millimetre and sub-millimetre wavelengthsbn204/publications/2009/NikolicERIS... · 2018. 12. 14. · Radio interferometry at millimetre and sub-millimetre
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Radio interferometry at millimetre and sub-millimetrewavelengths
Bojan Nikolic1 & Frederic Gueth2
1 Cavendish Laboratory/Kavli Institute for CosmologyUniversity of Cambridge
2 Institut de Radioastronomie MillimetriqueGrenoble
ERIS 2009Oxford, September 2009
Rev 33
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 1 / 62
Introduction
Outline
1 IntroductionScientific differences from the cm/m-wave bandObservational differences from the cm/w-wave bandScience examples
Introduction Scientific differences from the cm/m-wave band
Some of the fundamental science from other bands
Most science targets ‘cool’ and close to thermal equilibriumRotational lines of molecules, dust continuum, atomic carbon
Emission mechanisms are energetically significant for starformation both on local and galaxy-wide scalesRelatively low opacity except in the strongest molecular linesStrong positive cosmological ‘K-correction’
Continuum from star-forming galaxies does not dim from z = 1 toz ∼ 8
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 4 / 62
Introduction Scientific differences from the cm/m-wave band
Some of the fundamental science from other bands
Most science targets ‘cool’ and close to thermal equilibriumRotational lines of molecules, dust continuum, atomic carbon
Emission mechanisms are energetically significant for starformation both on local and galaxy-wide scalesRelatively low opacity except in the strongest molecular linesStrong positive cosmological ‘K-correction’
Continuum from star-forming galaxies does not dim from z = 1 toz ∼ 8
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 4 / 62
Introduction Scientific differences from the cm/m-wave band
CO emission line ladder in Milky WayFixsen et al. (1999)
Lines show models S ∝ ν4 exp[−E/kT ]
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 5 / 62
Introduction Scientific differences from the cm/m-wave band
Some of the fundamental science from other bands
Most science targets ‘cool’ and close to thermal equilibriumRotational lines of molecules, dust continuum, atomic carbon
Emission mechanisms are energetically significant for starformation both on local and galaxy-wide scalesRelatively low opacity except in the strongest molecular linesStrong positive cosmological ‘K-correction’
Continuum from star-forming galaxies does not dim from z = 1 toz ∼ 8
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 6 / 62
Introduction Scientific differences from the cm/m-wave band
Spectral energy distribution of galaxiesLagache et al. (2005)
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 7 / 62
Introduction Scientific differences from the cm/m-wave band
Some of the fundamental science from other bands
Most science targets ‘cool’ and close to thermal equilibriumRotational lines of molecules, dust continuum, atomic carbon
Emission mechanisms are energetically significant for starformation both on local and galaxy-wide scalesRelatively low opacity except in the strongest molecular linesStrong positive cosmological ‘K-correction’
Continuum from star-forming galaxies does not dim from z = 1 toz ∼ 8
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 8 / 62
Introduction Scientific differences from the cm/m-wave band
Dust extinction model: UV to near-IRDraine (2003)
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 9 / 62
Introduction Scientific differences from the cm/m-wave band
Dust extinction model: near-IR to mm-waveDraine (2003)
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 10 / 62
Introduction Scientific differences from the cm/m-wave band
Some of the fundamental science from other bands
Most science targets ‘cool’ and close to thermal equilibriumRotational lines of molecules, dust continuum, atomic carbon
Emission mechanisms are energetically significant for starformation both on local and galaxy-wide scalesRelatively low opacity except in the strongest molecular linesStrong positive cosmological ‘K-correction’
Continuum from star-forming galaxies does not dim from z = 1 toz ∼ 8
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 11 / 62
Introduction Scientific differences from the cm/m-wave band
Positive cosmological K-correctionLagache et al. (2005)
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 12 / 62
Introduction Observational differences from the cm/w-wave band
Outline
1 IntroductionScientific differences from the cm/m-wave bandObservational differences from the cm/w-wave bandScience examples
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 13 / 62
Introduction Observational differences from the cm/w-wave band
Fundamental observational differences from cm/mwave
Small field of viewFiner resolution (yet to be fully realised)High cost per element of the arrayLack of zero-spacing (‘total-power’) and short-spacing informationSky has a small dynamic range, low surface brightness of typicalsourcesMechanical effects on antennas are importantTroposphere gets seriously in the way, ionosphere not importantLarge absolute – but small fractional – bandwidthsCurrent arrays do not have good instantaneous uv -coverage
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 14 / 62
Introduction Observational differences from the cm/w-wave band
Contrast vs single dish
Single dish continuum surveys are confusion limited –interferometry essential for really deep surveysMuch easier to integrate down:
Atmospheric brightness fluctuations are rejectedStanding waves are rejectedGain fluctuations of the receivers less important
Better astrometryGood surface brightness sensitivity is expensive
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 15 / 62
Introduction Science examples
Outline
1 IntroductionScientific differences from the cm/m-wave bandObservational differences from the cm/w-wave bandScience examples
(Sub-)mm radiation ispredominantly affected by thetroposphere→ (Sub-)mm telescopes aresited at high elevations(Mauna Kea, Chajnator, SP),airborne observatories(SOFIA) or spaceLittle effect on polarisation
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 28 / 62
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
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 32 / 62
Fundamental uncertainties1 Atmospheric transparency varies with time and with frequency2 Receiver gain is variable3 Antenna gain is difficult to measure and sometimes variable
Difficult to inject a signal of known strengthQuasars are highly variable at (sub-)mm wavelengthsSolar system bodies (e.g., Mars, Neptune) also variable, but canbe modelled
Accurate models for the radiometric brightness are requiredMay be resolved, especially at sub-mm wavelengths
→ A calibration chain is required
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 45 / 62
Build reasonably stable receiver systemsCalibrate receiver gain using hot and ambient load
every ∼ few to tens of minutesCalibrate atmospheric absorption through a combination of:
Tipping scans (once ∼ 1 hour)Atmospheric models and WVRs (could go as short as ∼ 1 second)Total power atmospheric emission[Quasar observations (once ∼ 3 mins)]
Calibrate antenna gains using primary calibration standardsonce per session – once a yearat short wavelengths only planets may be suitable
Calibrate antenna primary beam shape through directinterferometric measurement
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 46 / 62
Offline calibration/imaging
Outline
1 IntroductionScientific differences from the cm/m-wave bandObservational differences from the cm/w-wave bandScience examples
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 47 / 62
Offline calibration/imaging
Bandpass calibration
Principle:Frequency-dependence of gain is independent of time
Calibration steps:Observe a strong quasar at beginning of each session(Need high SNR since can not combine the channels)Fit (complex) gain vs frequency
If SNR is high solve for each channel individuallyOtherwise fit a smooth function of frequency
Apply this bandpass solution to all other data in the session
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 48 / 62
Offline calibration/imaging
Bandpass calibration: PdB example – amplitude
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 49 / 62
Offline calibration/imaging
Bandpass calibration: PdB example – phase
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 49 / 62
Offline calibration/imaging
Bandpass calibration: SMA Data + CASAAmplitude – two out of 24 spectral windows shown
Antennas 1&2 Antennas 1&3 Antennas 1&4
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
0
5
10
15
20
25
Vect
or
Avera
ge A
mplit
ude
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
5
10
15
20
25
Vect
or
Avera
ge A
mplit
ude
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
5
10
15
20
25
Vect
or
Avera
ge A
mplit
ude
Antennas 1&5 Antennas 1&6
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
5
10
15
20
25
Vect
or
Avera
ge A
mplit
ude
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
5
10
15
20
25
Vect
or
Avera
ge A
mplit
ude
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 50 / 62
Offline calibration/imaging
Bandpass calibration: SMA Data + CASAPhase – two out of 24 spectral windows shown
Antennas 1&2 Antennas 1&3 Antennas 1&4
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
-150
-100
-50
0
50
100
150
Phase
Avera
ge
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
-160
-150
-140
-130
-120
-110
Phase
Avera
ge
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
-160
-150
-140
-130
-120
Phase
Avera
ge
Antennas 1&5 Antennas 1&6
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
-60
-50
-40
-30
-20
-10
0
10
20
Phase
Avera
ge
-12400 -12350 -12300 -12250 -12200
Channel Velocity (km/s)
-100
-90
-80
-70
-60
-50
-40
Phase
Avera
ge
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 51 / 62
Offline calibration/imaging
Phase calibration
Principles:Observed phase of a point source at phase centre should be zeroThe phase response of telescope will change very little for smallangular changes on the skyMost causes of errors are antenna-based and independent ofbaseline
Calibrations steps:Observe quasars every 10 seconds to 20 minutesFit (complex) gain vs time to estimate phase variation
If SNR is low and you think phase should be varying slowly, fitsmooth functionsIf SNR is high or there are jumps, use linear interpolation
Apply phase solution to science data
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 52 / 62
Offline calibration/imaging
Phase calibration: PdB Example – smooth phasevariation
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 53 / 62
Offline calibration/imaging
Phase calibration: PdB Example – jump ignored bysmooth fit
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 53 / 62
Offline calibration/imaging
Phase calibration: PdB Example – use higher orderfunction
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 53 / 62
Offline calibration/imaging
Phase transfer
Principles:Low frequency receivers are more sensitive, atmosphere moretransparent, telescopes more efficientQuasar spectra often roughly ∝ ν−0.7
→ easier to make phase calibration observations at lowerfrequency
Calibration steps:Observe phase calibration targets at 3 mmScale phase solutions to science bands and apply to data
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 54 / 62
Offline calibration/imaging
Phase transfer: PdB Example
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 55 / 62
Offline calibration/imaging
Phase transfer: PdB Example – with transferredcorrection
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 55 / 62
Offline calibration/imaging
Amplitude/Flux calibration
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 56 / 62
Offline calibration/imaging
Amplitude/Flux calibration
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 56 / 62
Offline calibration/imaging
Amplitude/Flux calibration
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 56 / 62
Offline calibration/imaging
Imaging
Imaging generally tractable with established techniquesAdding short/zero-spacing one important challenge but nowalmost routine in some systems
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 59 / 62
Summary
Summary
The physics of (sub-)mm emission means observing it allowsunique scienceInterferometers open the possibility of high-resolution and deepobservationsTroposphere has a big effect on (sub-)mm radiation butcombination of excellent sites and new techniques can/will resolvemost of theseMany of the techniques are the same as traditional cm-waveinterferometryIRAM mm-interferometry summer schools: next year
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 60 / 62
Summary
Resources on the web
Brogan et al: CASA training pages:http://casa.nrao.edu/casatraining.shtml
Belloche A., Andre P., 2004, A&A, 419, L35Draine B. T., 2003, ARA&A, 41, 241Fixsen D. J., Bennett C. L., Mather J. C., 1999, ApJ, 526, 207Lagache G., Puget J.-L., Dole H., 2005, ARA&A, 43, 727Riechers D. A., Walter F., Bertoldi F., Carilli C. L., Aravena M., Neri R.,
Cox P., Weiss A., Menten K. M., 2009, ArXiv e-printsWalter F., Riechers D., Cox P., Neri R., Carilli C., Bertoldi F., Weiss A.,
Maiolino R., 2009, Nature, 457, 699Younger J. D., Omont A., Fiolet N., Huang J.-S., Fazio G. G., Lai K.,
Polletta M., Rigopoulou D., Zylka R., 2009, MNRAS, 394, 1685
B. Nikolic & F. Gueth (Cambridge/IRAM) (Sub-)mm Wave Interferometry ERIS 2009 62 / 62