Determining the Hubble constant very accurately

In New Scientist 10 March 1990 page 17 refers to an article in Nature
vol 343 p726 by Tom Broadhurst, Richard Ellis, David Koo and Alex
Szalay, a joint Anglo-Australian deep sky survey of galaxies. It has a
graph (see http://homepages.ihug.co.nz/~ray.tomes/galwall1.gif)
which shows very clearly a periodicity in the distances of galaxies
which they refer to as 128 MPc (although there should be a h^-1 in there
that New Scientist omitted), and also some other periodicities which
appear to be 1/2 and 1/4 of that.

The quoted 128 MPc periodicity is to multiplied by 100 and divided by
the correct hubble constant (referred to by astronomers as h^-1). As
recent hubble constant estimates are of the order of 71 km/s/MPc that
would imply a periodicity of 180 MPc or about 590 MLY.

This distance of 590 million light years matches a period that I have
seen mentioned as around 590 million years. This matching of figures in
years and light years is not the commonly made dimension error, but
results from the fact a wave of length 590 MLY will oscillate in a
period of 590 MY.

G Williams and other geologists in "Megacycles", the proceedings of a
geological conference reports a series of geological cycles of ~600,
~300, 147 and 37 million years indicates that the largest cycle is about
590 MY assuming ratios of 2 throughout. However geology Prof S Afanasiev
of Moscow University in his book "Nanocycles Method" has determined this
cycle very accurately as 586.238 million years. This accurate figure
allows a very precise Hubble constant to be determined.

It seems highly probable that this very powerful geological periodicity
is related to the periodicity of galaxies. Working on that assumption,
and combining Prof Afansiev's period with the measured periodicity gives:

Hubble constant = 128 MPc * 100 km/s/MPc * 3.2616 MLY/MPc / 586.238 MY =
71.2 km/s/MPc (provisional value)

I am attempting to get the data from the authors to do a more accurate
determination and also to very that the 586 Million year cycle is the
observed periodiodicity. This can be done by checking that the other
known cycles, 293.1, 146.56, 73.28, 36.64 and 26.65 MY are also
present. The 26.65 million year is a cycle that has been extensively
reported in mass extinction events. My initial analysis from reading
figures off the graph and doing a spectral analysys indicates that it is
indeed so. From the original data it should be possible to determine the
Hubble constant to an accuracy of about 0.1% which is far more accurate
than any existing method.

Regards
Ray Tomes

[[Mod. note -- My (non-expert) understanding is that the Broadhurst
et al result is now thought to be a statistical fluctuation, not
indicative of an overall periodicity of the universe. A quick search
on gr-qc for titles containing the words "pencil beam" turned up
astro-ph/9402059, which looks rather relevant:

Non-Gaussian statistics of pencil beam surveys

Authors: Luca Amendola
Comments: 9 pages,2 figs available on request,Latex, revised version with significant changes, preprint Fermilab-Pub-94-043-A
Journal-ref: Astrophys.J. 430 (1994) L9

We study the effect of the non-Gaussian clustering of galaxies on
the statistics of pencil beam surveys. We find that the higher order
moments of the galaxy distribution play an important role in the
probability distribution for the power spectrum peaks. Taking into
account the observed values for the kurtosis of galaxy distribution
we derive the general probability distribution for the power spectrum
modes in non-Gaussian models and show that the probability to obtain
the 128$\hm$ periodicity found in pencil beam surveys is raised by
roughly one order of magnitude. The non-Gaussianity of the galaxy
distribution is however still insufficient to explain the reported
peak-to-noise ratio of the periodicity, so that extra power on large
scales seems required.

-- jt]]

[[Mod. note -- My (non-expert) understanding is that the Broadhurst
et al result is now thought to be a statistical fluctuation, not
indicative of an overall periodicity of the universe. A quick search
on gr-qc for titles containing the words "pencil beam" turned up
astro-ph/9402059, which looks rather relevant:

Non-Gaussian statistics of pencil beam surveys

Authors: Luca Amendola
Comments: 9 pages,2 figs available on request,Latex, revised version with significant changes, preprint Fermilab-Pub-94-043-A
Journal-ref: Astrophys.J. 430 (1994) L9

We study the effect of the non-Gaussian clustering of galaxies on
the statistics of pencil beam surveys. We find that the higher order
moments of the galaxy distribution play an important role in the
probability distribution for the power spectrum peaks. Taking into
account the observed values for the kurtosis of galaxy distribution
we derive the general probability distribution for the power spectrum
modes in non-Gaussian models and show that the probability to obtain
the 128$\hm$ periodicity found in pencil beam surveys is raised by
roughly one order of magnitude. The non-Gaussianity of the galaxy
distribution is however still insufficient to explain the reported
peak-to-noise ratio of the periodicity, so that extra power on large
scales seems required.

Dear JT

Reading this it seems to me to be saying that although a pencil beam
combined with non-Gaussian distribution enhances the probability of
false periodicities in this range, the actual case is still significant.
So your note above is too strongly worded I think.

Anyone looking at the graphic should be able to see that it is a very
real periodicity: http://homepages.ihug.co.nz/~ray.tomes/galwall1.gif

Ray Tomes