[Prev][Next][Index][Thread]

Re: (meteorobs) Doomsday Icarus



Thanks Wayne, for the reply to Kevin's messages.  I would refer back to my
earlier message.  If you "keep it private", you are blamed by the media for
"covering it up" and you don't get the observations made usually, as there
are so few people involved at the actual observational end (I was employed
in NEO discovery and observation from 1990-1996 until the Australian govt.
cut our funds; not important enough although we discovered 16% of the 108
potentially hazardous asteroids).  By keeping quiet, eventually someone will
come up with the story and you have no control over how it is used.  Even
in this case, despite the measure words of Marsden and others, the comments
of the news anchors introducing the story irreversibly color it  "Astronomers
disagree about the danger posed..."  "Some astronomers predict a collision
of an asteroid with the Earth...".

I was a little annoyed in seeing various NASA representatives stating that
the danger was "zero" being interposed with Marsden and others saying the
threat was small.  The NASA statements were made after the 1990 precovery
by the Helin's JPL team, and the statements were either chosen by the 
media to create a false tension, or the NASA folks involved really were
trying to put one over.

Wayne's point on the errors from the first estimate, not including the
more accurate estimate of miss distance can be explained by two factors.
Firstly, an error estimate in such a circumstance is probabilistic and
I guess Marsden took a suitably large figure to include perhaps a 99%
probability.  However the wings of the remaining 1% can be very large.
The second point is that the error estimates assume that the observational
errors are random.  This will not be true for two reasons.  Each
observatory that contributed observations (and only a handfull did with this
object) can introduce small systematic errors due to their observation or
reduction techniques, or small timing errors (a few seconds can count).
Perhaps more importantly, the star catalogue used, especially if not
the latest, will have systematic position errors in different parts of the
sky.  I have had experience with this myself.  10 days after the discovery of
comet Hale-Bopp, I would a tiny faint smudge of the comet on a plate taken
a year and a half-earlier (in the same sort of precovery process that
Ken Lawrence used for 1997 XL11).  It was initially about 7 arcmins from
the predicted position based on the small arc.  Marsden trusted me and
published this.  As time went on, the discrepancy dropped, but eventually
that last 30 arcsecs or so would not disappear.  JPL made an announcement,
without discussing the issue with me, that there was a problem with my
precovery observation, and they had excluded it from their orbital
analysis to have the highest accuracy.  However, the comet was moving so
slowly during that time that the observed arc since discovery was in one
small portion of the sky, and most astrometrists were using the low-
accuracy Hubble Star Catalog.  The huge number (thousands) of observations
which included systematic errors due to the star catalog, were the cause
of the problem and Marsden was the first to discover this by suitably
weighting the observations.  It has not been shown (with astrometry from
discovery to present day) that my precovery measure is accurate to around
1 arcsecond.

Anyway, I'll also get off my soapbox (and put my dirk back in my sock,
yes "dirk").

Cheers, Rob
(rmn@aaocbn.aaodot gov.au)