Natural Science Forum / Physics / Research / July 2008

...

Al, are there astrophysicists or astronomers of decent repute that are
predicting that Betelgeuse is going supernova on a particular date so
close to us in the future?  that would be really neat to be able to
witness a supernova that is visible to the naked eye. i know that
Betelgeuse is a red giant of a size bigger than our orbit around our
own sun, but i never heard of a prediction like that, by credible
science, being made in modern times.

that's even better than the time Mars was so close a few years back.

r b-j

To return to the original poster's question, there have not (yet)
been any confirmed direct detections of gravitational waves (GWs).
[The qualifier "direct" is important; see below.]

As Gerry Quinn has pointed out, there are strong theoretical reasons
to expect that many astronomical systems (among others) emit _some_
sort of GWs.  However, the details of those GWs turn out to be highly
sensitive to aspects of strong-field (nonlinear) gravitation which have
not (yet) been experimentally verified.  That is, there are otherwise-
-plausible theories of gravitation (eg, Rosen's bimetric theory) whose
weak-field limits, and hence whose predictions in almost all experiments
done to date, are almost identical to those of general relativity (GR),
yet whose strong-field limits (an predictions for GWs) are profoundly
different from GR's.

This was the situation in 1974, when Hulse & Taylor discovered the
binary pulsar PSR B1913+16.  This system contains a pulsar (a rapidly
spinning neutron star) whose periodic radio signals are observed, in
a tight eccentric orbit around another neutron star.  GR predicts that
such a system emits (mainly) quadrupole GWs, which carry away (positive)
energy, resulting in the binary-pulsar system loosing energy, which
shows up as the binary-pulsar orbit graually shrinking.  In contrast,
Rosen's theory predicts the emission of *dipole* GWs [which don't
exist in GR, because they'd violate conservation of momentum], which
are typically many orders of magnitude stronger than the quadrupole
GWs preicted by GR.  Moreover, Rosen's theory predicts that dipole
GWs generally carry *negative* energy, so the binary-pulsar system
would *gain* energy in emitting the GWs.  Thus a measurement of
binary-pulsar orbital evolution can help discriminate between GR and
Rosen's theory.

Binary-pulsar orbital evolution can be measured by precisely timing
the arrival times of the pulsar's radio pulses.  The results for this
for PSR B1913+16's were announced in 1979, and they agree very precisely
with the GR prediction (and strongly *disagree* with the pedictions
from Rosen's theory; because of this, Rosen's theory is now dead).
The binary-pulsar observations constitute a clear, although indirect,
experimental detection of GWs.  Hulse and Taylor shared the 1993
Nobel prize in physics for their work.

The original poster asked if any of the current GW detectors have
seen anything.  These detectors are designed to *directly* detect GWs;
if this can be done, it would provide much more information about GWs
(for example, their detailed amplitude/phase/polarization structures)
than the indirect binary-pulsar detection (which essentially only
measured the time-integrated energy carried by GWs).  Direct detection
would thus let us make lots of interesting inferences about the
astrophysics of the GW sources, as well as improve our knowledge
of study strong-field gravitation.

Contrary to what Uncle Al suggests, supernova (SN) explosions are
actually not a very likely GW source for current detectors.  It's not
that SN don't emit GWs (they surely do).  Rather, astrophysically-
-plausible SN models probably produce only relatively weak GWs,
probably too weak to detect with any current detector.

The best known candidate sources for current or near-future GW detectors
are coalescing neutron-star and black hole binaries.  The detectors
have a certain noise level, which sets the minimum GW strength arriving
at the Earth for which a reliable (e.g., 5 sigma) detection can be made.
For a given intrinsic GW source strength, this (by the inverse-square
law) then sets the maximum distance at which such a source can be
detected.  Actually, current GW detectors detect GW *amplitude*, not
power, so the falloff with distance is only $1/r$, not $1/r^2$.
(This helps a lot, see below.)

For the currently-running LIGO, Virgo, and GEO detectors, this maximum
detection distance is typically around 10 megaparsecs or so for binary
neutron star coalescence.  Unfortunately, binary neutron stars are not
that common (amongst all binary star systems), and the current best
estimates of how frequently such a coalescence will occur within 10
megaparsecs of us are very roughly from 1 per year to 1 per 1000 years.
[The wide error bars here are a reflection of our uncertain astronomical
knowledge of massive-stars lifecycles and populations, and thus of the
size of the parent population of binary neutron star systems.]  So,
it's not particularly surprising that LIGO, Virgo, and GEO haven't
seen anything (yet).

Fortunately, help is on the way:  LIGO and Virgo both plan major
upgrades (using technology partly pioneered by GEO) which should cut
their noise levels by roughly a factor of 10 by 2014 or so.  Since
GW amplitude only decays 1/r, this means that the maximum detection
distance for a given source will go up by roughly a factor of 10.
This in turns means 10^3 = 1000 times the volume of space, and thus
about 1000 times the event rate, so after these upgrades we expect
strong-enough-to-be-reliably-detected binary neutron-star coalescences
somewhere between 1000 times a year (= roughly 3 events/day) and
once per year.

Thus, allowing for a few years worth of schedule slips, funding
crises, and debugging of the (very complicated) new hardware, it's
very likely that GWs will be directly detected by roughly 2015-2018.
[2018 is also the *earliest* planned launch date for LISA, a
proposed space-based GW detector which (if built) will have very
high sensitivity to low frequency GWs.]

Finally, there is always the possibility that there are new types
of strong GW sources that (human) astrophysicists haven't thought
of yet.  If we're lucky, these sources might turn out to be readily
detectable sooner than the known sources I've discussed above.

ciao,

Maybe, but isn't it true that the only ones we are likely to detect come
from very high field events?