between 1 and 5 mHz, i.e., within one order of magnitude
of the LISA goal.
(Noteworthy: two ``graybeards'' at the Symposium gave strong warnings
about the technology demonstrator. Rai Weiss warned repeatedly
that it should not be made too ambitious, since 1) you can't risk it
failing and 2) you don't want it to absorb all your time/energy.
Ben Lange, a pioneer of drag-free flight and
a veteran of many, many of successful space missions,
advised that he'd ``avoid a demonstrator
mission like the plague.'')
The Symposium included about 50 talks, which is too many to summarize.
I'll confine myself to listing what were, to me, a few highlights,
and apologize in advance for the many excellent presentations I won't
even mention here, but which you'll be able to read in the Proceedings.
Sterl Phinney gave a beautiful and very upbeat summary of
LISA sources.
New to me were the quite optimistic estimates
for the merger rate for massive black hole binaries (MBH's), based
on a hierarchical clustering picture of structure formation,
where small galaxies form first and merge to form bigger galaxies.
This picture leads to estimates of event rates of
/yr
for
BH's out to z=2. But, importantly, LISA can
see far beyond z=2; Phinney argued that the merger rate
for
BH's out to 
might be 
day.
John Armstrong and Massimo Tinto discussed their very important work (done
with Frank Estabrook), showing how one can (with some changes in hardware)
linearly combine the LISA data
streams, with time delays, to form three linearly independent combinations
for which the laser phase noise exactly cancels. Two combinations contain
information on the two gw polarizations, and the third describes a
``breathing mode'' that doesn't couple to GR. This third mode
can be used to help calibrate and
eliminate noise sources, and to discriminate between
non-Gaussian noise bursts and real gw bursts.
There were several very interesting talks on
solar-mass compact objects spiraling
into MBH's. Scott Hughes showed that, when the
MBH is near-extreme Kerr, the inspiral is strongly
affected by
superradiant scattering of gw's from the BH horizon.
Gw's that scatter off the horizon tend to ``hold up''
the test-body and increase the inspiral time by
(which is a lot of cycles).
Wolfgang Tichy discussed work-in-progress with E. Flanagan,
claiming that, because the background Kerr metric is stationary,
it actually is
possible to determine how the Carter constant evolves from fluxes at infinity
(and at the horizon).
And Bernard Schutz discussed his worry
(aroused, I assume, by recent work by Janna Levin) that because
the orbits of spinning bodies in Kerr are chaotic, the number of matched
filters will grow exponentially with the integration time, and
may be vastly greater than previously anticipated-effectively obliterating
the usual gains from matched filtering.
This was a warning, not result, and somebody needs to look more
carefully at this issue.
Large extra dimensions (perhaps as large as 0.3 mm)
are now much-discussed in string theory, and Craig Hogan
showed how these might lead to a gw background observable by LISA.
He argued that the early universe should produce
copious gw's with wavelength comparable to the size
of the large extra dimension, which would be
redshifted into the LISA band today.
Since the gw spectrum so-produced would be highly
peaked rather than flat, it is not constrained
by bounds at much lower frequencies coming
from millisecond pulsar timing and COBE.
Ben Lange, who attended the whole meeting and then
gave us his outsider's perspective, made several recommendations,
especially emphasizing advantages of spherical test masses, instead
of cubes as in the current plan.
And he gave a delightful, short
summary of how in practice one can use a felt pen and
the classical mechanics of precession to
find the principal axes of an almost-perfect sphere.
Something new for a LISA Symposium: there were several
talks describing laboratory prototypes for LISA systems. (Oliver
Jennrich: ``LISA is now more than just ink on paper.'') Harry Ward
discussed his work on developing an interferometric read-out
system. He also reported on tests of how well hydrogen catalysis
bonding of optical elements would survive the rigors of launch and
space-and found the bonding held quite well under shaking and thermal
cycling. Oliver Jennrich described his experiment showing the
feasibility of the LISA phase measurement scheme, using the same
amount of light as will be available for LISA. Manuel Rodrigues
described a laboratory prototype for the inertial sensor, and Stefano
Vitale described a torsion pendulum test bench he is building to
testing the performance of the inertial sensor on the ground to some
.
And Michael
Petersheim discussed a a prototype for the highly stable laser
required by LISA (
).
Lastly, there was very interesting discussion both of possible
variations in the LISA mission and possible follow-on missions (the
latter to be flown around 2020-2025, so fancy was free). Bernard
Schutz pointed out the possible advantages of LISA starting out as
short-arm interferometer, before moving the satellites to the current
baseline separation of
km. He also suggested the
addition of a 4th spacecraft, to fly at the midpoint of one of the
three arms. NASA has strongly encouraged LISA scientists to think
about possible follow-on missions to LISA, and is especially
interested in missions that might detect a primordial background of
gw's. There is little chance that LISA itself can detect a primordial
background, since in the LISA band it will be swamped by the
background from galactic and extra-galactic binaries. The binary
background falls off at high frequencies, which leads to a
next-generation LISA concept featuring 3 constellations of
mini-LISA's, with the constellations forming a equilateral triangle
around the Sun at 1 AU, and each mini-LISA having short (
-km) arms to push the sensitivity band up to 1-10 Hz. All of
this seems technologically feasible even in the near term-it's
``just'' a matter of money.