[meteorite-list] Phil Bland Article

From: Kelly Webb <kelly_at_meteoritecentral.com>
Date: Thu Apr 22 09:41:13 2004
Message-ID: <3A9CAACA.A4E97746_at_bhil.com>

Hi, Sara, List, and Phil Bland,

    I believe I also said, "it is essentially unfair to judge any scientific
argument by its paraphase in a popular journal." I always give myself good advice
but I don't always follow it.

    My irritation (no actual hostility) was triggered solely by the often
repeated "about 500 meteorites fall to earth each year" statement which pops up
over and over again, not only in popular journals and literature, but in texts
and primary sources as well.

    It's curious that the journalist's misquote created exactly that value of
500, apparently by random error. Had it been some other number (like 1373), I
would have been puzzled, but not annoyed. Rob Matson had earlier asked, "I'd love
to know where that 500/year figure originally came from."

    I have the answer to that, by the way.

    As it turns out, its origin can be traced to the sainted Nininger who
estimated in 1940 that 500 meteorites of 100 GRAMS OR MORE fall OVER LAND per
year. Of course, for comparison this figure needs to be adjusted to the scale of
the MORP and Bland studies (10 to 10,000 grams or 10 grams and up) and for the
land to water area ratio of the planet, which would raise it to 5000 to 6000
meteorites per year (depending on the exponent in the power law you use to scale
from 100 grams down to 10 grams).

    Considering the data and resources available to Nininger, it constitutes a
remarkably judicious evaluation. And I'm sure it was considered wild and
improbable at that time! Like so many oft-repeated "factoids," its use as a stock
quotation today is both dated and out of context. But the phrase "500 meteorites
fall to earth each year" seems to have acquired a life of its own.

    Rob wrote to me because I had previously (December 12, 2000) posted to this
List an evaluation of fall rate based on a new method of determining a value.

    Its genesis lay in this List having had a long and sometimes entertaining
wrangle about the reported death of the Nakhla dog. There were believers and
non-believers. Those who could not accept a meteorite killing a dog thought the
event so improbable that they would accept only an absolute proof, while those
who thought it might well have happened were willing to accept a preponderance of
the evidence proof. It seemed to me that the difference of opinion lay entirely
in one's estimation of the probability of the thing.

    Say "probability" to a physicist and he hears "cross-section." As we know,
the mean geometric cross section of a canine times the number of dogs in the
world times the planetary fall rate divided by the area of the planet equals the
annual probability of meteorite-dog collision. The function can be inverted to
yield a mean-time-to-impact value. And, if collision frequency were known, fall
rate could be calculated. Whimsical, but quite sound.

    While I couldn't solve for dogs (how many dogs are there in the world? Do
they have post mortems for dogs?), I realized that verified reports of persons
struck by meteorites or of cars struck by meteorites constituted a useful data
set. The human population for historic eras is well-known. The number of
registered motor vehicles in the U.S. is also.

    Simple, methodologically sweet, it does not rely on any variables whose
values need to be taken from any previous studies, in other words, a completely
independent means of verification. A collisional cross section study eliminates
selection effects, like recovery percentages. You do not have to find the
meteorite; the meteorite finds you. To avoid even unconscious bias on my part in
assembling or selecting data, I took a list of 141 such incidents previously
published by John S. Lewis (1996) which could be regarded as well-vetted. The
method's only disadvantage derives from the small sample size and consequent
uncertainty (about +/- 20%).

    The results were a) from humans struck by meteorites, a fall rate of 80,000
to 100,000 meteorites per year; and b) from U.S. motor vehicles struck by
meteorites, a fall rate of 50,000 to 70,000 meteorites per year. The data set for
autos is so small (four) that the addition of one more case in the century would
boost the calculated fall rate derived from autos to the magnitude of that
derived from persons. Possibly humans struck by meteorites are more easily noted
than injuries to cars!

    As soon as I read that the Bland study did not number individual stones but
rather uses the term "meteorite" in the grouped sense, I realized that we may
have no disagreement beyond conflicting definitions. The collisional method
counts individual physical objects, of course, without regard to their
immediately prior association.

    The site of measurement matters here. A fireball survey is measuring at or
near the top of the atmosphere. A collection or collision study is measuring at
ground level. Between the two measurements, two factors apply: first, atmospheric
extinction (objects that do not survive to reach the ground), and second, a
fragmentation multiplier. The two factors have an opposite sense, and their
product would logically scale between the two varieties of measurement.

    Perhaps the difference between the MORP figure (23,930) and the Bland figure
(~20,000) is accounted for by a not quite complete recovery ratio at the sites
studied or by a MORP bias for high altitude objects that do not survive.

    I would be fascinated to know what the Bland study's fall rate would be if
the stones were ungrouped to count every discrete physical object (including the
interpolated small stones) as a collisional study has to do. Perhaps we are all
talking about the same numerical values after all!

    Rob, when you calculate a fall rate of 160,000 per year, are you counting
individual stones?

    Is there any possibility of deriving extinction and fragmention factors from
these comparisons?

    I doubt that any more information could be squeezed from the small data set
of a collisional study, but it pleased me to quantify material that heretofore
has been treated in a solely anecdotal manner and that the two good results
obtained were roughly concordant.

    So, Phil, I apologize for my intemperate initial response. It was unjustified
(and bad form, as I believe you say over there, or do you say that any more?).

    Oh, and I did not mean that you had excluded weathering; I was referring to
the mechanisms which might remove stones by transport and burial, even in desert
areas in which water seems a non-issue. If only once in a few centuries or once a
millennium, they experience violent rainstorms. All these deserts contain
numerous evidences of catastrophic flooding and scouring. Even the flat Nullabar
has pans to which innumerable small stones have been washed, buried, even
concretized (that's where you find 700,000 years' worth of transported tektites).

    Human modification of the meteorite field has ancient roots, too. In the
flood of stones pouring out of Morocco (the poor man's Antarctica), why are there
virtually no irons? (Except for the low-grade Nantans the bazaar merchants try to
pass off on the unwary.) Irons represent a high percentage of finds because of
their persistence, so why no irons? I think it's because the
nth-great-grandfathers of these same merchants sold them to the Romans and the
Carthaginians and the Garamonteans and... The Esquimaux had been making tools out
the Cape York set for centuries before they made the mistake of telling the Great
Explorer about them.

    Looking forward to getting and reading the study.


Sterling K. Webb



Sara Russell wrote:

> Dear All,
>
> Phil Bland, the subject of the Discover magazine article, is not on
> Meteorite Central, but I forwarded the postings about it to him. His reply
> is pasted below.
>
> Sara Russell
>
>

(See original post.)
Received on Wed 28 Feb 2001 02:37:46 AM PST