Date: 05 Apr 94  04:10:13
   From: Donald Lindsay
     To: All
Subject: Review: Origins 20(1)

From: lindsay+@cs.cmu.edu (Donald Lindsay)
Organization: School of Computer Science, Carnegie Mellon


	Origins 20(1)  Winter 1993 pp. 6-21
	"An Alternative Explanation of
	Oceanic Magnetic Anomaly Patterns"
	by Norm Smith and Jane Smith
	Review by Don Lindsay

This article begins by explaining that there is a heck of a lot of
evidence that the ocean floor is being slowly created at long
volcanic fissures, such as the one down the Mid-Atlantic Ridge. They
pass briefly over the fact that this also ties cleanly in to a lot of
other evidence.

They claim that there is a shortage of direct measurements: that is,
magnetic measurements (or radioactive datings) of actual samples, or
even in-situ magnetic measurements made up-close to the actual ocean
floor. I'm not enough of a geologist to claim otherwise without
searching the literature, but I certainly had the impression that we
possess many direct measurements. I know we've dated samples.

In any case, the article eventually gets to its actual suggestion,
which is that a single linear magnetometer flight yields a graph that
looks a lot like band-limited noise. [These are signal artifacts from
trying to measure a too-high-frequency signal.] This idea is defended
by pointing out that an airborne magnetometer is kilometers above
the bedrock, and is hence measuring an average taken across a large
area.

At this point, I should explain my qualifications. I did some of the
digital filtering for the 1969 Canadian magnetometer flights across
the Atlantic. I have since had professional involvement in areas that
are mathematically related, such as acoustic signal processing.

The article is correct in saying that such noise can occur. In fact,
we used a digital band filter: this to eliminate noise sources such
as the ocean waves beneath the aircraft. (We could equally have
eliminated everything but the waves, and produced a graph of wave
height along our flight path.) There are indeed conceivable geologic
circumstances which would have introduced noise into our filtered
results. I will return to this point.

The conventional explanation is that the sea floor is being produced
along such ridges, and is pushed away from the ridge as more rock
arrives. As molten rock cools to below its Curie point, it acquires
a magnetic direction, under the influence of the earth's magnetic
field. [This has been directly observed on land.] The banding of the
graphs is interpreted as evidence that the earth's magnetic field
has reversed direction at intervals, thus causing parallel bands
on the sea floor.

The article then suggests that a ridge might have produced physical
features which parallel the ridge, but by means other than sea floor
spreading. They admit that the features may be "tens or even hundreds
of kilometers long parallel to a ridge yet separated by only a few
kilometers perpendicular to a ridge". Assuming that these features
exist, the authors go on to do a computer simulation. They assume
that the rock has magnetic variations on a much finer grain than the
averaging window, but that the each physical feature is distinct
magnetically (by virtue of amplitude values).

This simulation yields results that resemble the magnetic
measurements in the literature. From this, they conclude that their
simulated geology is as reasonable as the conventional one, and that
the jury should remain out on the conventional explanation.

From the signal-processing perspective, this is semi-reasonable.  The
existence of fine-grained variations seems to be no more of an
assumption than any other. However, from the physical perspective, it
replaces something which is known to occur on land, with an unknown
mechanism. Possible, no doubt: but unknown. Furthermore, this unknown
mechanism must operate reliably and uniformly, because if it
sometimes does not, then this non-operation would show up in our
measured data.

If we assume that this mechanism does exist, what is in doubt?

- Not sea floor spreading. The article suggests that stresses and
strains near the ridge are their alternate explanation for the
parallel features. This does not begin to explain why the features
should retain symmetry about the Mid-Atlantic Ridge, a thousand
kilometers from the Ridge. It also does not explain why the rock
nearer the Ridge should be younger than the rock nearer the
continents.

- Not continental drift. That is proven conclusively, not by
knowledge of its mechanism, but by the existence of matching
geological formations on the now-separate continents. The fact that
the continents are *now* in motion (at exactly the right speed) is
clear evidence that some mechanism does exist.

- Not magnetic reversals. Their existence is known from land data.

In short, this article proposes an unknown mechanism, in place of a
known one, and claims that field data does not yet preclude it.  That
seems to be true, but no reason is given why the unknown mechanism
is to be preferred as an explanation.

Furthermore, should field evidence prove this idea correct, nothing
important will be different: North America will continue towards
China: the sea floor will spread (what else, stretch?): and magnetic
reversals will still have actually occurred.

It would indeed be nice to get better field data, and I would not
wish to discourage Creationists from assisting in the funding of
oceanographic research.
--
Don		D.C.Lindsay	Carnegie Mellon Computer Science