Stellar Intensity Interferometry: The Background John Davis Sydney Institute for Astronomy School of Physics University of Sydney NSW, Australia 29 January 2009 Intensity Interferometry Workshop 1
Personal Notes I regret that I cannot make this presentation in person but an outline of 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 of 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 Intensity Interferometry Workshop 2
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 Intensity Interferometry Workshop 3
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 Intensity Interferometry Workshop 4
The Entry of Richard Tw iss • 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, once 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 Intensity Interferometry Workshop 5
The Radio Intensity Interferometer Equipment (Hanbury Brown, Jennison, & Das Gupta, Nature, 170, 1061, 1952) Antenna Systems Heterodyne receivers tuned to 125 MHz with Δ f of 200 kHz Square-law detectors Low-frequency filters Δ f from 1 to 2 kHz Radio link Correlator Delay line 29 January 2009 Intensity Interferometry Workshop 6
The Radio Intensity Interferometer Experiment • The output signal from the correlator is proportional to the visibility 2 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 Intensity Interferometry Workshop 7
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 Intensity Interferometry Workshop 8
The Start of Optical Intensity Interferometry • RHB and RQT envisaged an optical analogue of the radio intensity interferometer The radio intensity The envisioned optical interferometer 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 Intensity Interferometry Workshop 9
Optical Intensity Interferometry Laboratory Experiment (Hanbury Brown & Twiss, Nature, 177, 27, 1956) • 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 C 1 laterally with the slide, no correlation was observed 29 January 2009 Intensity Interferometry Workshop 10
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 >10 11 years and Brannen & Ferguson for >1000 years to achieve a S/N ratio of 3! 29 January 2009 Intensity Interferometry Workshop 11
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 of it being the result of a random noise fluctuation was <1 in 10 15 • 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 Intensity Interferometry Workshop 12
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 Intensity Interferometry Workshop 13
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 observations 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 Intensity Interferometry Workshop 14
Recommend
More recommend
