Stellar Intensity Interferometry: The Background John Davis Sydney - - PowerPoint PPT Presentation

29 January 2009 1 Intensity Interferometry Workshop

Stellar Intensity Interferometry: The Background

John Davis

Sydney Institute for Astronomy School of Physics University of Sydney NSW, Australia

29 January 2009 2 Intensity Interferometry Workshop

Personal Notes

I regret that I cannot make this presentation in person but an outline

• f my involvement in stellar interferometry may be of interest:
• I was a student at the University of Manchester when the

radio version of intensity interferometry was implemented

• Although I wasn’t involved, I was at Jodrell Bank as a PhD

student and then as a Postdoc throughout the development

• f the optical technique and the measurement of Sirius
• In 1961 Hanbury Brown invited me to work on the Narrabri

Stellar Intensity Interferometer and no young postdoc in his right mind would have said anything but “Yes please!”

• I have been involved in optical stellar interferometry ever

since.

29 January 2009 3 Intensity Interferometry Workshop

Outline of Presentation

• Robert Hanbury Brown’s original idea of intensity interferometry

and the role of Richard Twiss

• The radio astronomy experiment and scintillation
• Optical laboratory experiments
• The measurement of Sirius by intensity interferometry
• The Narrabri Stellar Intensity Interferometer
• Plans for a Very Large Stellar Intensity Interferometer (VLSII)
• The VLSII abandoned in favour of an amplitude interferometer
• The Sydney University Stellar Interferometer (SUSI)
• Some thoughts on the future of intensity interferometry

29 January 2009 4 Intensity Interferometry Workshop

The Origin of the Idea of Intensity Interferometry

• Circa 1949, the angular sizes of the two brightest radio sources, Cygnus A

and Cassiopeia A, were unknown and some thought they were “radio stars”. Robert Hanbury Brown (RHB) was determined to measure them

• If these sources were galaxies their angular sizes would be of the order of

a minute of arc and easy to measure with a conventional interferometer but, if they were stars, extremely long baselines would be needed and RHB concluded that this was impossible with the available technology

• RHB worried about it and had the following thought, in his own words, “If

the radiation from a discrete source in the sky is picked up at two different places on Earth, is there anything besides the phase and amplitude of the signals which we can compare to find the mutual coherence?”

• He then visualised the “noise-like” signal seen by two separated observers

and realised that the noise corresponded to low-frequency fluctuations in the intensity of the signal and convinced himself that the correlation between the intensity fluctuations was a measure of their mutual coherence.

29 January 2009 5 Intensity Interferometry Workshop

The Entry of Richard Twiss

• Although RHB had convinced himself with a simple analysis that his

idea of intensity interferometry was sound, he was not able to develop the mathematical theory to establish the sensitivity himself

• He sought help from a friend who put him in touch with Richard Twiss

(RQT), a gifted mathematician

• After his initial analysis RQT announced to RHB “This idea of yours is

no good, it doesn’t work!”

• It turned out that RQT had made a simple mistake in an integral and,
• nce corrected, he produced a rigorous and quantitative theory of

intensity interferometry

• The next step was to develop a radio intensity interferometer to test

the technique and to measure Cygnus A and Cassiopeia A

29 January 2009 6 Intensity Interferometry Workshop

The Radio Intensity Interferometer Equipment

Antenna Systems Heterodyne receivers tuned to 125 MHz with Δf of 200 kHz Square-law detectors Low-frequency filters Δf from 1 to 2 kHz Correlator Delay line Radio link

(Hanbury Brown, Jennison, & Das Gupta, Nature, 170, 1061, 1952)

29 January 2009 7 Intensity Interferometry Workshop

The Radio Intensity Interferometer Experiment

• The output signal from the correlator is proportional to the visibility2

that would be observed with a Michelson type interferometer

• Four baselines of different lengths and orientations were used to

determine the angular dimensions of Cygnus A and Cassiopeia A

(Hanbury Brown, Jennison & Das Gupta, Nature, 170, 1061, 1952)

• Cygnus A was elongated with dimensions of approx. 0.5′ x 2′ and

Cassiopeia A was roughly symmetrical with a diameter of approx. 3.5′

• It turned out that Graham Smith at Cambridge and Bernie Mills in

Australia had also measured these sources with conventional radio interferometers and obtained similar results

(Mills, Nature, 170, 1063,1952; Smith, Nature, 170, 1065, 1952)

• In RHB’s words they had built “a steam roller to crack a nut” because

the sources were clearly not stars – but that is another story

29 January 2009 8 Intensity Interferometry Workshop

Lessons from the Radio Intensity Interferometer

• Matching paths in the arms of the interferometer was much easier

than for a conventional amplitude interferometer: the tolerance was set by the maximum frequency of the filtered low-frequency signals whose correlation was being measured, and not by the frequency of the radio signal

• It was observed that the correlation from Cassiopeia A was constant

in spite of violently scintillating signals due to the ionosphere

• RQT found that they had overlooked, in the theoretical development,

perhaps the most astonishing and valuable feature of intensity interferometry – it can work perfectly through a turbulent medium

• RHB and RQT wondered at that point if intensity interferometry

could be made to work at optical wavelengths and measure the angular diameters of stars

29 January 2009 9 Intensity Interferometry Workshop

The Start of Optical Intensity Interferometry

• RHB and RQT envisaged an optical analogue of the radio intensity

interferometer The radio intensity interferometer The envisioned optical analogue

• For the optical analogue to work the time of arrival of photons had to

be correlated at the two photocathodes for coherent incident light

• This had never been observed and experimental proof was needed

29 January 2009 10 Intensity Interferometry Workshop

Optical Intensity Interferometry Laboratory Experiment

• A simplified diagram of the experimental apparatus
• A measurement was made with the photocathodes optically

superimposed and correlation was observed in close agreement with the theoretical prediction

• When the photocathodes were optically separated, by translating

C1 laterally with the slide, no correlation was observed

(Hanbury Brown & Twiss, Nature, 177, 27, 1956)

29 January 2009 11 Intensity Interferometry Workshop

The Controversy

• These experimental results set the cat amongst the pigeons!
• Experimentalists set out to check the results and concluded that

the RHB & RQT results were wrong

• Adám, Jánossy & Varga (Acta Hungarica, 4, 301, 1955) and Brannen &

Ferguson (Nature, 178, 481, 1956) carried out experiments and did not detect correlation – the latter went as far as stating that the existence of correlation would call for “a major revision of some fundamental concepts of quantum mechanics”

• Neither group had evaluated the theoretical predictions for the

conditions of their experiments . RHB & RQT did the calculations (Nature, 178,1447, 1956) and showed that Adám et al. would have needed to integrate for >1011 years and Brannen & Ferguson for >1000 years to achieve a S/N ratio of 3!

29 January 2009 12 Intensity Interferometry Workshop

More on the Controversy

• The two negative experiments used a photon coincidence counting

technique and RHB & RQT showed that they were simply too insensitive to record correlation

• Twiss, Little & Hanbury Brown (Nature, 180, 324, 1957) repeated the

Brannen & Ferguson coincidence counting experiment with a brilliant light source of narrow spectral bandwidth They not only measured their predicted correlation but showed that the chance

• f it being the result of a random noise fluctuation was <1 in 1015
• Purcell (Nature, 178, 1449,1956) adopted a different approach to the

theory and his analysis of the three experiments came to the same conclusions as RHB & RQT

• In parallel with dealing with the controversy RHB decided to

measure the angular diameter of a main-sequence star (Sirius) to demonstrate the astronomical potential of intensity interferometry

29 January 2009 13 Intensity Interferometry Workshop

The Measurement of Sirius - I

Simplified diagram of the apparatus The starlight collectors – two World War II 1.56 m diameter searchlights

29 January 2009 14 Intensity Interferometry Workshop

The Measurement of Sirius - II

• The measurements were made in conditions of severe scintillation in

the winter of 1955-6 – wet, cold & muddy!

• The maximum elevation of Sirius at transit was 20 degrees and the
• bservations were made at elevations between 15.5 and 20 degrees

Although the measured angular diameter of 7.1±0.55 mas is larger than the currently accepted value of ~6.0 mas it was a remarkable achievment

(Hanbury Brown & Twiss, Nature, 178, 1046, 1956)

This measurement showed beyond doubt the potential of stellar intensity interferometry and led to the Narrabri Stellar Intensity Interferometer

29 January 2009 15 Intensity Interferometry Workshop

Simplified diagram of the Narrabri Stellar Intensity Interferometer (NSII)

Output to printer the correlator (some details later) P1 and P2 are Photomultipliers f1 and f2 are identical wide-band filters

29 January 2009 16 Intensity Interferometry Workshop

The Narrabri Stellar Intensity Interferometer (NSII)

The general layout of the NSII

• The maximum baseline, set by the track diameter of

188 m, was chosen to resolve the O5 star ζ Puppis

• Circular track so no delay compensation as the baseline

was kept perpendicular to the star’s direction

29 January 2009 17 Intensity Interferometry Workshop

The Narrabri Stellar Intensity Interferometer (NSII)

Stellar Observations (1964-1972)

188 m diameter track Reflectors Catenaries (Signal & power cables) Control Building Reflector Garage Baseline

29 January 2009 18 Intensity Interferometry Workshop

• The individual mirrors were spherical and were mounted on paraboloidal frames.
• Given the tolerance on the radius of curvature, the shorter radius mirrors were

used at the centre and the longer radius ones towards the edge of the paraboloids.

The NSII Reflectors

Diameter = 6.5 metres 252 individual hexagonal mirrors 251 aligned on the signal detector and 1 on a separate star guidance detector

29 January 2009 19 Intensity Interferometry Workshop

The Control Desk of the NSII (circa 1967)

JD in control!

29 January 2009 20 Intensity Interferometry Workshop

The Evolution of Control Desks!

The Analogue Control Desk of the NSII (circa 1967) The Digital Control Desk of SUSI (2008)

29 January 2009 21 Intensity Interferometry Workshop

The Correlator

• The function of the correlator was to measure the correlation

between the fluctuations in the anode currents of the photomultiplier detectors at the foci of the two reflectors

• The r.m.s. signal to noise ratio at the output of the multiplier

was very small (<1 in 105 for a bright unresolved star)

• The correlator multiplied the fluctuations in the two channels

together and the correlation was a unidirectional output superimposed on random noise

• The output of the correlator was extremely sensitive to DC

gain drifts, pick up, and cross-coupling. Phase switching techniques were essential to overcome these problems as shown on the next slide

29 January 2009 22 Intensity Interferometry Workshop

Block diagram of the correlator

10 second phase switch to minimise the effect of drift in circuits and counter false correlation due to pick up and coupling between circuits 5 kHz phase switch to counter stability problem of high gain DC ampllifiers

29 January 2009 23 Intensity Interferometry Workshop

The NSII Correlator

The output printer Originally all thermionic valves – transistors were not an option in 1961!

Note the size!

Transistorised sequence timer that replaced a mechanical/microswitch system circa 1965

29 January 2009 24 Intensity Interferometry Workshop

The Correlator Output

• The integrated correlation was printed every 100 seconds and plotted

by the observer as shown in the example below for 20th May 1965 for

• bservations of β Crucis

29 January 2009 25 Intensity Interferometry Workshop

The Correlator Output

• The integrated correlation was printed every 100 seconds and plotted

by the observer as shown in the example below for 20th May 1965 for

• bservations of β Crucis at baselines of 32.7 m and 94.2 m

Number of 100 second integration cycles

“Dummy” run β Cru: 32.7m β Cru: 94.2m “Dummy” run

“Dummy” Runs Between observations the detectors were exposed to the same light levels as from the star using incoherent artificial sources to monitor any drift in the system Correlation

29 January 2009 26 Intensity Interferometry Workshop

Some Key Points

• The measured correlation depends on the gain of the system.

Calibrated with a standard source of wide band noise fed simultaneously into both channels in place of the photomultiplier

• utputs before and after each night of observations
• The scale of the correlation depends on instrumental parameters

and the scale changes if, for example, the detectors are changed

• Hence it is necessary to measure both short and long baselines

with the same instrument parameters – also identifies binaries

• Large reflectors may partially resolve the source and this must

be taken into account as we did

• The NSII had several changes in instrumental parameters and

reducing all results to a consistent scale was a tedious exercise

• Providing sensitivity is adequate, serious consideration should be given

to using calibration sources, as is done for amplitude interferometry

29 January 2009 27 Intensity Interferometry Workshop

The Achievements of the NSII

• Measured the angular diameters of 32 stars for spectral types from O5

to F8 resulting in the effective temperature scale for early-type stars

(Hanbury Brown, Davis & Allen, MNRAS, 167, 121, 1974; Code, Davis, Bless & Hanbury Brown, ApJ, 203, 417, 1976)

• Made the first interferometric-spectroscopic study of a double-lined

spectroscopic binary (α Vir) (Herbison-Evans et al., MNRAS, 151, 161, 1971)

• Carried out a number of exploratory experiments including:
• Detected previously unsuspected binary stars
• Measured the angular size of the emission envelope around the Wolf-

Rayet star γ2 Vel in ionised carbon lines

• Measured the effects of Cerenkov light pulses on the NSII
• Attempted to detect a corona around β Ori in polarised light
• Attempted to measure limb-darkening for Sirius
• Attempted to measure the rotational distortion of Altair
• The signal-to-noise was insufficient to obtain astrophysically

significant results for the last three experiments but they illustrated the potential of high angular resolution stellar interferometry

29 January 2009 28 Intensity Interferometry Workshop

The Hiatus at the Conclusion of the NSII Programme

• RHB decided to close the NSII stellar programme in 1972 when further
• bservations would have been of low weight (due to low S/N) and

would not have added significantly to the results already obtained

• He planned to build a 2 m telescope and use it to demonstrate the

capabilities of the new detectors that were being developed at that time - but I persuaded him that we should build on our experience and develop a Very Large Stellar Intensity Interferometer (VLSII)

• We carried out a detailed study of the science that we would want to

do, including measuring the pulsations of Cepheids, and concluded that we would need to reach a visual magnitude of >+7

• As an aside, in parallel with this we used the NSII in a collaboration

with the CfA to detect atmospheric Cerenkov light from extensive air showers (Grindlay et al., ApJ, 197, L9, 1975; Grindlay et al., ApJ, 201, 82, 1975)

29 January 2009 29 Intensity Interferometry Workshop

A Proposal for a VLSII - I

• There were no unknowns except the achievable sensitivity
• The sensitivity parameters at our disposal were:
• The light collecting area (A)
• The quantum efficiency of the detectors (α)
• The radio frequency bandwidth of the signals (Δf)
• The number of multiplexed optical channels (N)

S/N ∝ A.α √ Δf.N

• α and Δf were set by the detectors and A and N were

at our disposal – the latter limited by crude optics

• Based on our study of the sensitivity required for the science we

were interested in, we developed a proposal for a Very Large Stellar Intensity Interferometer (VLSII)

29 January 2009 30 Intensity Interferometry Workshop

A Proposal for a VLSII - II

• Initially we started with extremely optimistic predictions about sensitivity

based, in part, on predictions by RCA about future photomultipliers and

• n optimistic numbers of optical channels at the foci of the reflectors

N was based on a study of polarising and dichroic beamsplitters and was limited by the crude optics

• f the reflectors. The

specifications of the latter were set by the maximum funds we estimated might be achievable. Parameter NSII Optimistic VLSII A (m2) 30 160 Δf (MHz)
 80 1000 N 1 10 α (%) 25 30 Gain 1 72 mlimit +2.5 +7.1

29 January 2009 31 Intensity Interferometry Workshop

A Proposal for a VLSII - III

The basic configuration of the proposed VLSII

Aligned to observe a star in the zenith Aligned to observe a star at ~70o elevation

29 January 2009 32 Intensity Interferometry Workshop

A Model of the Proposed VLSII

• Two 10m diameter siderostats

in each arm

• Multi-spectral channels at the

foci of fixed paraboloids

• 1 km long railway tracks

JD and RHB with the model of the VLSII

29 January 2009 33 Intensity Interferometry Workshop

The Model of the Proposed VLSII

29 January 2009 34 Intensity Interferometry Workshop

• As we developed the design we revised our estimates of the

sensitivity to represent more realistically what we believed could be achieved in practice within a reasonable timescale

Second Thoughts on the Proposal for a VLSII

A limiting magnitude of +5.8 did not meet our needs Parameter NSII Optimistic Realistic VLSII VLSII A (m2) 30 160 160 Δf (MHz)
 80 1000 200 N 1 10 6 α (%) 25 30 25 Gain 1 72 21 mlimit +2.5 +7.1 +5.8

29 January 2009 35 Intensity Interferometry Workshop

Second Thoughts on the Proposal for a VLSII

• As noted on the previous slide, a limiting magnitude of +5.8 did not

meet our needs

• I was concerned that a modernised form of Michelson’s classical stellar

interferometer might be more sensitive and I persuaded RHB that we should make a comparison of the two techniques

• We carried out a detailed comparison and consulted RQT who had

been developing a small scale modern Michelson interferometer in Italy (I had worked on an early version with RQT in England during a sabbatical)

• Our study showed that a modern Michelson (amplitude) interferometer

promised greater sensitivity and we decided to abandon the VLSII and RHB left me to develop a prototype modern amplitude interferometer

• Little did we realise how long it would take amplitude interferometry

to reach the sensitivity we were after!

29 January 2009 36 Intensity Interferometry Workshop

The Sydney University Stellar Interferometer 12.4 m Prototype

150 mm diameter southern siderostat & relay mirrors. (The northern siderostat is hidden by the building)

29 January 2009 37 Intensity Interferometry Workshop

The Sydney University Stellar Interferometer 12.4 m Prototype

• The prototype was used to develop and test the various sub-systems

needed for a large long-baseline amplitude interferometer

• At the time, in the 1970s and early 1980s, we were pushing the

boundaries of technology and some aspects were only just possible

• We successfully demonstrated the feasibility of our approach with a

measurement of the angular diameter of Sirius that was in good agreement with the NSII value but achieved in a fraction of the

• bserving time (Davis & Tango, Nature, 323, 234, 1986)
• These included wavefront tip-tilt detection & correction, dynamic optical

path length compensation, and rapid signal sampling & processing

• Based on this success we designed and raised the funds to build

the Sydney University Stellar Interferometer (SUSI)

29 January 2009 38 Intensity Interferometry Workshop

The Sydney University Stellar Interferometer (SUSI)

Seen from the northern end of its 640 m North-South baseline array

29 January 2009 39 Intensity Interferometry Workshop

SUSI

An input station & siderostat Optical Path Length Compensator Blue beam combination system Red beam combination system

29 January 2009 40 Intensity Interferometry Workshop

SUSI Parameters and Status

• Baselines: 5 m, 10 m then in ~√2 steps to 640 m (5-160 m fully
• perational)
• Apertures: 20cm diameter siderostats, beam diameter 14cm
• Spectral Range: 430 nm < λ < 950 nm
• Blue beam-combination system: 430-530 nm (Δλ: 1-4 nm) Blimit ~ +2.5
• Early-type stars, early-type binaries
• Red beam-combination system: 530-950 nm (Δλ: 5-10%) Rlimit ~ +5
• Late-type stars, binaries, Cepheids

The following beam-combination systems have been used for the scientific programme to date but are in the process of being replaced: SUSI is being upgraded for remote operation with a new beam-combination system (PAVO) developed in a collaboration for both SUSI and CHARA

• SUSI PAVO has 10 parallel spectral channels, spatial filtering & mlimit ~+7

29 January 2009 41 Intensity Interferometry Workshop

• Measurement of the outer scale of turbulence
• Accurate stellar angular diameter measurements (<1%)
• Combined interferometric-spectroscopic studies of binary

systems

• First direct mass determinations of masses of β CMa stars

(in spectroscopic binaries)

• Measurement of the angular diameter variations and, in

combination with spectroscopic radial velocities, determination of distances and mean radii of Cepheids

• First interferometric spectropolarimetry
• First combined interferometric-asteroseismological study

to determine the mass of a single star

Some Highlights of the SUSI Science Programme

29 January 2009 42 Intensity Interferometry Workshop

Some Random Thoughts on II

• Any thoughts I have are not profound, and have no doubt already

been considered by others

• The following are some topics for consideration that I will expand
• n in the next slides:

1. What science is possible with II that cannot be done with AI? 2. What is the sensitivity limit for a modern Intensity Interferometer and how does it compare with current Amplitude Interferometers? 3. Methods of calibrating measurements of correlation

29 January 2009 43 Intensity Interferometry Workshop

Some Random Thoughts on II - 1

• 1. What science is possible with II that cannot be done with AI?

Some possible stellar programmes:

• Angular diameters and limb-darkening of single stars
• Binary stars, particularly double-lined spectroscopic binaries
• Cepheid variables
• Oblateness of rapidly rotating stars
• Emission line observations

All these programmes have been, and are being addressed by current amplitude interferometers and I do not see where II could make a significant contribution

An often quoted advantage of II is that the problems were solved with the NSII and that there are no unknowns Although it has taken a long time to achieve, I believe that the same can now be said of AI First, a comment:

29 January 2009 44 Intensity Interferometry Workshop

Some Random Thoughts on II – 1 (cont.)

There are non-stellar programmes that may be possible with II but not with AI – I have not had time to study these possibilities but they must be carefully evaluated taking into account what is possible with current and future amplitude interferometers It has been suggested that AI cannot operate at the short wavelengths and long baselines needed for the hottest stars because of increasing seeing effects. I do not believe this is true for the following reasons:

• Blimit ~ +2.5 for SUSI but this was mainly due to the detection

technique.

• I had long-term plans to upgrade the blue system and would have

easily reached Blimit +7 - even with the small SUSI beam diameter.

• Furthermore, we have shown that the outer scale of turbulence is
• nly a few tens of metres and seeing effects will remain constant

beyond that (Davis et al., MNRAS, 273, L53, 1995)

29 January 2009 45 Intensity Interferometry Workshop

Some Random Thoughts on II - 2

2. What is the sensitivity limit for a modern II and how does it compare with current AIs?

• In spite of the earlier comment regarding “no unknowns”, there are

uncertain factors entering the calculations if the large light collectors developed for very high energy gamma-ray observations are to be used, including:

• Point–spread function – large, admitting sky background and adding

noise, reducing mlimit

• Although not directly a sensitivity factor, the pointing accuracy of the

reflectors must be good enough for II

• Shape of the tesselated surface that may give a significant spread in

path lengths and hence limit the bandwidth that can be used

• For these reasons I have not attempted to carry out independent

sensitivity calculations but it would appear that with existing arrays, the limiting magnitude would be in the range +6 to +8

29 January 2009 46 Intensity Interferometry Workshop

Some Random Thoughts on II – 2 (cont.)

• A dedicated instrument such as that proposed by Erez Ribak would have

a much fainter mlimit, but would be hard to fund in the face of the achievements of AI, unless a compelling scientific case could be made A summary of selected long-baseline optical/IR amplitude interferometers

Acronym Location Number Aperture Maximum Wavelength Instrument Limiting Notes Status

• f

Diameter Baseline Range Magnitude * Apertures (m) (m) (mm) SUSI Narrabri, Australia 2 0.14 640 0.43-0.53 Blue system B ~+2.5 Superseded W 0.53-0.95 Red system R ~+5 Superseded W 0.6-0.9 PAVO R ~ +7 C ISI

• Mt. Wilson, USA

2 1.65 70 10

• ?

W NPOI Flagstaff, USA 6 (4) 0.12 (0.35) 437 (38) 0.45-0.85

• ?

W CHARA Mt. Wilson, USA 6 1.0 330 0.45-2.4 Several other W 0.4-0.9 VEGA R ~ +8 instruments W 0.62-0.9 PAVO R ~ +10

• inc. imaging

W Keck Mauna Kea, Hawaii 2 (4) 10 80 2.2-10

• ?

Only 1 baseline W VLTI Cerro Paranal, Chile 4 (4) 8 (1.8) 130 (200) 1.0-10 MIDI (8m) N ~+4 W MIDI (1.8m) N ~+0.7 W AMBER (8m) H & K ~ +7

• Imaging. Fainter

W AMBER (1.8m) H & K ~ +5 with PRIMA tracking W PRIMA K ~ +8-+11 C MRO New Mexico, USA 6 (+4) 1.4 340 0.6-2.4

• ~ +14 #

Imaging B # Spectral band not given * W = Working; C = Commissioning; B = Being built

29 January 2009 47 Intensity Interferometry Workshop

Some Random Thoughts on II – 2 (cont.)

• Although the entries in the table are incomplete, inspection shows that

magnitudes of +7 and fainter are being achieved by several instruments including imaging

• Short wavelengths and very long baselines are not as well represented

but are feasible

Acronym Location Number Aperture Maximum Wavelength Instrument Limiting Notes Status

• f

Diameter Baseline Range Magnitude * Apertures (m) (m) (mm) SUSI Narrabri, Australia 2 0.14 640 0.43-0.53 Blue system B ~+2.5 Superseded W 0.53-0.95 Red system R ~+5 Superseded W 0.6-0.9 PAVO R ~ +7 C ISI

• Mt. Wilson, USA

2 1.65 70 10

• ?

W NPOI Flagstaff, USA 6 (4) 0.12 (0.35) 437 (38) 0.45-0.85

• ?

W CHARA Mt. Wilson, USA 6 1.0 330 0.45-2.4 Several other W 0.4-0.9 VEGA R ~ +8 instruments W 0.62-0.9 PAVO R ~ +10

• inc. imaging

W Keck Mauna Kea, Hawaii 2 (4) 10 80 2.2-10

• ?

Only 1 baseline W VLTI Cerro Paranal, Chile 4 (4) 8 (1.8) 130 (200) 1.0-10 MIDI (8m) N ~+4 W MIDI (1.8m) N ~+0.7 W AMBER (8m) H & K ~ +7

• Imaging. Fainter

W AMBER (1.8m) H & K ~ +5 with PRIMA tracking W PRIMA K ~ +8-+11 C MRO New Mexico, USA 6 (+4) 1.4 340 0.6-2.4

• ~ +14 #

Imaging B # Spectral band not given * W = Working; C = Commissioning; B = Being built

29 January 2009 48 Intensity Interferometry Workshop

Some Random Thoughts on II - 3

3. Methods of calibrating measurements of correlation

• i. Split the light with a beamsplitter to two detectors at the focus of a

reflector and measure the correlation between the signals – corresponding to zero baseline. This would lose a factor of 2 in S/N.

• ii. Use the technique adopted for AI of alternating observations of the

target with observations of calibrators - sources of relatively small angular size or of known angular size In all cases the effects of partial resolution by the large reflectors would need to be taken into account

• iii. Do as was done with the NSII and measure long and short baselines

and fit the expected transform. Provided no changes are made to the instrument the zero baseline correlation value can be established for single stars and used to detect binary systems etc.

29 January 2009 49 Intensity Interferometry Workshop

• I suspect that it is unlikely that II would contribute significantly to

stellar studies

SUMMARY

• In spite of technical advances it is not obvious that II will

achieve greater sensitivity than amplitude interferometers that exist or, in the case of the MRO, are under construction

• AI is not necessarily limited in resolution by seeing and, if the

existing longer baselines (up to 640 m) of SUSI are brought on line, SUSI-PAVO will be capable of measuring some of the hottest stars

• The future of II as an astronomical technique is dependent on

achieving the sensitivity for programmes that it can do and that AI cannot The following are my personal conclusions - but I don’t expect everyone to agree with them!

29 January 2009 50 Intensity Interferometry Workshop

The End