Date:  Wednesday, February 14, 1996
Time:  4:15 PM ­ Refreshments at 4:00
Location:  ERL 126


DETECTION OF NONSOLAR PLANETS BY SPINNING INFRARED INTERFEROMETER

Prof. Ronald Bracewell,
STARLab, Deptt. of Electrical Engineering,
Stanford University

Abstract

Back in the 70s NASA held two workshops at Ames, `Search for Extraterrestrial
Life' and `Extrasolar Planetary Detection.'  Barney Oliver worked out an
apodized optical telescope with such faint sidelobes that it might just be
possible to see a Jupiter-like planet against the dazzle of the parent star.
The light from Jupiter would be two billionths of the sunlight if our solar
system was viewed from afar. The telescope would have to be in orbit and be
very carefully freed from scattered and diffracted light.  At the same time I
reasoned that in the infrared the low temperature of Jupiter is favorable but
of course the Sun emits much more total power. Nevertheless, we are 100,000
times better off.  Of course, the wanted IR radiation is still only two
ten-thousandths of the unwanted.

A further advantage would accrue if the collecting area was split into two
apertures, say one meter in diameter, and if the radiation received was
combined in antiphase so as to produce zero response in the median plane of
the two elements.  This null line could be pointed at the center of a suitable
star. The sinusoidal interference pattern associated with a two-element
interferometer would peak up at an angular distance depending inversely on the
spacing of the elements.  At a distance of 30 light years, the angular
separation of star and planet is half a second of arc, suggesting an element
spacing of 7.7 m at a wavelength of 40 microns.  An attractive feature of this
conclusion was that such an interferometer could be constructed on a rigid
beam that would fit into the space-shuttle bay.

Although the stellar radiation is rejected along one diameter of the star, the
power receptivity builds up as the square of distance towards opposite limbs
of the star.  So the star is not nulled out completely, but the rejection
factor is calculable.

There is a question of pointing accuracy here.  The star will have an angular
diameter measured in milliseconds of arc, and unless the median plane of the
interferometer is pointed to corresponding accuracy, the rejection factor will
deteriorate.

An additional design feature is now introduced.  Spin the structure at a fixed
rate around the Earth-star axis and synchronously detect the modulated signal
from the planet, thus eliminating the dc signal from the star.  Error analysis
tells you how accurate the pointing must be.  But a good thing is working for
the spinning idea; the modulation of starlight due to mispointing varies with
the spin frequency, whereas that of the planetary signal varies at twice the
spin frequency and can be sorted out.

Signal-to-noise ratio under the above circumstances can be estimated if the
sky background is taken into account; the major contribution is from the cool
zodiacal-light particles.  It would help to place the interferometer out near
Jupiter or above the zodiacal plane.  For the situation at the time see Nature,
vol. 274, pp. 780-781, 1978 and Bracewell and MacPhie, Icarus, vol. 38, 99.
136-147, 1979.  An objection that was raised in print held that Cherenkov
radiation generated in the mirrors would be fatal, but that was disposed of
in Moon and Planets, vol. 30, pp. 75-77, 1984.

Of course no-one thought that this concept would be acted on. Even if NASA
gave the go-ahead there would be a 10 or 15 year wait until launch, so no
proposal was made, though Lockheed performed a study.

With the passage of time, NASA Administrator Dan Goldin had a vision of a
planetary image being obtained that would have the same electrifying effect on
the public as that beautiful blue and white image of the Earth in space had.
A competition was announced, in consequence of which a team led by Roger Angel
of the Lunar and Planetary Laboratory of the University of Arizona, Tucson,
has been selected to push ahead with a concept.  Lo and behold, the spinning
IR null interferometer has risen like a phoenix from the ashes.

Two important new features have been added by Angel and his chief collaborator
Nick Woolf.  The first is to add a second pair of collectors at twice the
spacing with half the area, also having a median null.  But the slope of the
field response versus angle, at the null, is made equal and opposite to that
of the original pair.  The power receptivity thus is very flat over the disc
of the star; signal rejection and pointing tolerance both benefit calculably.
The second feature is to receive a band of infrared rather than accept
monochromatic operation. The passage of 17 years has seen advances in
instrumentation that make a big difference.  Not only will spectral absorption
features of carbon dioxide and water be detectable, but, in the happy event of
more than one planet being present, it will be possible to locate them in
radial distance and position angle. An apparent ambiguity in position angle
can be resolved by a pointing adjustment.

The spinning infrared interferometer is on the way to becoming a major NASA
initiative of the coming millennium; this aspect will be discussed by
Roger Angel on February 28 in Durand 450 at 9:15 a.m.