On Wed, 26 Aug 2015 11:22:27 -0700, "Glenn" <g...@invalid.invalid>
wrote:
Try actually reading the paper you cite.
Specifically the authors looked at 112 known minerals containing
Beryllium. Some were very common, beryl for example, but many were
quite rare. The abundance distribution followed a Zipf-Mandelbrot
model just as, as cited in the paper, do species in a forest ecosystem
or words in a book. The authors predicted from the distribution
pattern that there are likely some 90 as yet undiscovered but very
rare Beryllium species here on Earth.
Of the total of 204 presumptive Earth minerals, the authors calculated
that 34 would likely occur on all earth-like planets with another 26
having > 95% chance of occurring. So more than half of the Beryllium
minerals found on Earth would likely be found on any Earth-like
planet. The differences are in the very rare minerals found on Earth
at only one or a few locations. The very specific set of these
rarities is extremely unlikely to be reproduced exactly on another
planet.
The authors predict "that every mineral-forming chemical element has
the potential to form at least 1000 species; thus, we suggest that as
many as 800 additional plausible Be minerals may occur in the cosmos,
significantly decreasing the likelihood of finding the identical suite
of 112 species on another planet." Repeat that for all the 4831 known
Earth minerals and you get that ridiculously small probability quoted.
However the authors also write "The most common rock-forming minerals
are likely to dominate on any Earth-like planet; thus, major
lithological units such as basalt and granite are deterministic.
However, the major-ity of mineral species are rare, with 1062 species
known from only 1 locality..." It is this set of rare minerals with a
chance of less than 50% for any one to appear on another planet that,
in exact combination, yields that number.
I don't think this result comes as any surprise. The basic minerology
of most of the mass of a different Earth-like planet will be identical
but the rare outliers are certain to be unique to each planet.
I also doubt that an Earth-like planet with tectonic plates will
include a Pacific plate rubbing against the North American plate to
produce a San Andreas Fault. I would expect, though, that any
Earth-like planet with tectonic plates will have subduction and
upwelling or rift zones as well as transform faults like the San
Andreas just as any Earth-like planet will have a wide variety of
Beryllium containing minerals with perhaps 50 of the most common types
agreeing completely with the same minerals we see on Earth.