[meteorite-list] Cyclic mass extinction?

From: Matson, Robert <ROBERT.D.MATSON_at_meteoritecentral.com>
Date: Thu Apr 22 10:00:10 2004
Message-ID: <AF564D2B9D91D411B9FE00508BF1C86901B4E261_at_US-Torrance.mail.saic.com>

Hi Al, Tracy, Edward and others interested in the mass extinction
periodicity
problem...

For newcomers, we had a brief exchange on this subject about a year and a
half
ago. Here's an excerpt from one of my posts in reply to Sam Kimpton:

Matson 1/22/2001:
-----------------

"I've heard various figures, with several million year error bars,
but there does seem to be some evidence of periodicity. Naturally,
one would like to identify the 'trigger'.

"... the proposed Nemesis star would have to be a red dwarf
or a brown dwarf. It would not be that difficult with
current technology to see such a star at aphelion (which is
about where the star would be for the current cycle). But
simply observing the star is only part of the battle; the
hard part is observing it for a long enough baseline to
determine that it is co-orbiting with the sun.

"Phil Bagnall once e-mailed me that the proposed orbit of this
star would have to be so elongated that it would not remain
stable (and that the orbit could not have survived for the
duration of the Solar System). He may be correct -- I cannot
say without seeing the model assumptions and the resulting
data. A 26-million-year period requires an orbital semi-major
axis of 87,760 a.u. Assuming the orbit is extremely elliptical,
and the star is currently at aphelion, its distance would be
twice this, or ~175,500 a.u. This is about 2.8 light-years --
more than half the distance to the nearest known star system.

"Depending on the plane of the orbit relative to Alpha-, Beta-
and Proxima Centauri (and/or possibly Barnard's star, Wolf 359,
and the other 6 stars within 10 light-years of the solar system),
the orbit might be unstable. At best it is chaotic long-term
because of the gravitational perturbations of the Centaurus
trinary system. But suppose perihelion lies between our solar
system and Centaurus. If so, it would take quite some time
(tens of orbits, more?) before the line of apsides would precess
to bring aphelion in the direction of Centaurus. If the Nemesis
period has remained at roughly 26 million years, then there hasn't
been time for even 200 orbits."

- - - -

Now, in 2002, I'd say it's highly unlikely that a dwarf companion
to the sun -- even in such a highly elliptical orbit -- would
have evaded detection by Hubble or other telescopes.

Others have theorized that the sun's orbit around the galactic
center causes it to periodically pass through the galactic plane
(presumably twice per orbit -- one ascending node, one descending
node), and that this plane passage somehow perturbs comets in
the Oort cloud leading to increased chaos and impacts in the
inner solar system. But I see two problems with this theory,
which perhaps someone out there can help me with. One is that
this "plane passage" is really a non-event; the local stellar
density really doesn't oscillate that much since the galactic
plane is not really a "plane" in the geometric sense. It has
considerable "thickness" -- on the order of 2000 light-years.
The sun is currently only ~50 light-years north of the galactic
"plane", so the local stellar density is about as thick as it
gets. Unless the sun's orbit takes it far above and below
the plane (say > +/- 500 light-years), it doesn't seem like
the density is going to oscillate much.

I considered that it might not be the plane crossings of a
slightly inclined solar orbit that produces the disruptions,
but the periapsis of a slightly eccentric orbit that takes the
sun closer to the denser galactic center once per orbit. The
problem with this is that the sun's orbit around the galactic
center is almost perfectly circular.

But the second (and larger) problem is that even if there is a
galactic "carousel" effect, the period is wrong. The sun orbits
the galactic plane about once every 225-250 million years, so
plane crossings would occur every 112-125 million years (and
periapsis once every 225-250 million years). Not a good match
to the much shorter extinction cycle.

What about the spiral arms? Passage through these, I think,
would have a much more disruptive effect. But from what I've
read, this occurs very rarely since the sun is at a distance
from the galactic center that has an orbital period very close
to that of the spiral arms.

So the question remains: what sort of natural events can occur
that have a periodicity of ~26 million years? Gravitational
laws place limits on the range of a sun-orbiting "trigger" --
something in the 1.5 to 3 light-year range. Gravimetric events
associated with motion through the galactic plane appear to have
periods much longer than the observed mass-extinction period.
It's an interesting problem... --Rob
Received on Thu 25 Jul 2002 05:30:34 PM PDT