[meteorite-list] Meteorite fall rate

From: Matson, Robert <ROBERT.D.MATSON_at_meteoritecentral.com>
Date: Thu Apr 22 09:41:13 2004
Message-ID: <AF564D2B9D91D411B9FE00508BF1C8692C5F93_at_US-Torrance.mail.saic.com>

Hi All,

Thank you Kelly, Phil, Jeff, Sara and Michel for keeping this
thread alive with valuable contributions. I'm impressed with
the variety of thoughtful approaches to deriving the annual
fall rate, and with placing error estimates on those values.
Each approach has its strengths and weaknesses. Kelly's
approach using automobiles and people (with side considerations
to animals, mailboxes, houses and so forth) eliminates weathering
and terrestrial lifetime from the equation, but may suffer from
undersampling and underreporting (and requires special attention
to multiple "hits" from single falls).

Jeff's approach using Japan again has the advantage of known
fall dates, but has an unknown fall recovery rate. It would
be interesting to know the time-of-day distribution of these
falls. As we know, certain times of day favor the survival
of a meteroid -- I would think 11am to 3pm local, and 9pm to
1am would be most favorable from the standpoint of minimizing
meteoroid relative velocity and angle of attack through the
atmosphere. (It also just now occurs to me that higher
latitudes enjoy shallower angles of attack for meteoroids in
orbits close to the ecliptic -- perhaps a good topic for
another thread.) My point in mentioning time-of-day with
respect to Japan is that many smaller daytime falls will
go unnoticed, and falls late at night when most are sleeping
will also be missed.

My approach and Phil's attack the problem using a surface area
canvassing approach. This creates a large statistical sample,
but one which covers an uncertain window of time. The desert
sites allow a relatively high fraction of falls which have
not weathered away to be recovered, but then requires accurate
pairing and terrestrial dating, and accounting for weathering,
transport, and burial. Assuming the pairing is done well, and
a maximum possible terrestrial age is determined, this methodology
will place a good *lower* limit on the number of annual falls.
Phil's calculated result of 20,000/year is thus a good lower
limit.

I still contend that the recovery success (finds made vs. falls
that can be found) is a key factor in the equation. One measure
of confidence is to compute the manhours-per-square-km that
have been spent searching each area of study, and to plot
the number of finds against this figure to confirm that a
point of "diminishing returns" has been reached. At some
number of finds for each region the slope of the curve will
asymptotically approach zero (barring any new falls of course).

Best wishes,
Rob
Received on Wed 28 Feb 2001 01:33:09 PM PST