“Cosmic Rays, Solar Forcing, and 20th Century Climate Change”
A talk to the Physics Dept, UW, March 15th, 2015, by Dr. Nir J.
Shaviv, of the Racah Institute of Physics, Hebrew Univ. Jerusalem.
This review by hh
Dr. Shaviv delivered a spirited talk to the Physics/Astronomy/Atm.Sci.
community at UW last Monday, on the controversial subject of ‘Cosmic
Ray influence on Climate’.
He and others postulate that variations in the solar magnetic field
modulate the incoming flux of galactic cosmic rays [mostly protons]
that impinge upon the earth’s atmosphere. This produces varying
ionization in the atmosphere, which in turn vary the concentrations
of cloud condensation nucleii and the extent and reflectivity of low
level stratus clouds, which in turn modulate the net heating at the
surface. I suspect we’ll be hearing more of this in the popular
press.
This is a fairly old idea, put forward again ‘recently’ [~2000] by
Svensmark and others, in papers that I have reviewed before in this
series. Shaviv’s contribution is some new data, [I think], and lively
support.
Those data include, but were not restricted to:
1. Updated plots of atmospheric and tree-ring Carbon14 [a surrogate
of cosmic ray flux] and of Oxygen18 [a surrogate of planetary temper-
atures], over the last several hundred years. The bumps in these
plots appear to correlate with the ‘Medieval Warm Period’, an inter-
vening ‘little ice age’, and our present warming.
He also presented plots with much longer time series, derived from
meteors.
Also similar tide-gauge plots that appear to show 11-year cycles
that he attributes to thermal expansions and contractions.
All these plots obligingly show synchronies of interesting bumps.
Shaviv correctly points out that the measured variations of both the
solar insolation above the atmosphere [which is well measured by
satellites over the last 40+ years] and the net heating at the surface
[which is less well measured over a longer epoch of about 100 years]
are orders of magnitude too small to account for the ‘observed’
modulation of the ocean volume. If the oceans are truly expanding and
contracting synchronously with the solar cycle, there must be some
amplifying mechanism, on the order of 10X. Shaviv and others
attribute this to the variation in cloud nucleation described above.
2. Shaviv is a co-author of a 2012 paper that describes a ‘2-box’
climate model with least-squares adjustable parameters to account for
the cosmic ray effect. They find a best fit with a climate sensi-
tivity parameter, ‘S’, attributable to CO2 about half that of the most
recent IPCC-AR5 assessment. The difference is attributed to recent
trends in galactic cosmic rays. ‘S’, by the way, refers to the
expected change in global surface temperatures associated with a
doubling of CO2.
Some opinions from hh:
1. The plots shown by Shaviv are intriguing. He may be right.
And we shall certainly be hearing more about cosmic rays. Good Oh!
That’s the way science works.
2. I doubt that either the surface-measured variations in the solar
insolation, or the tide-gauge measurements of ocean volume, are
sufficiently accurate to sustain Shaviv’s attribution to solar
modulation at 11 years via cosmic rays, cloud formation, and so forth.
3. A ‘2-box’ climate model is MUCH too naive. Similar studies with
state-of-art ocean-coupled models should be attempted. My guess is
that the data will not support a sufficiently accurate partition of
‘S’ into effects from CO2 and from cosmic rays. We shall see.
3. Shaviv appears to have been unaware of oceanographic data
suggesting a recent and perhaps periodic acceleration of the ‘AMOC’
[Atlantic Meridional Over-turning Circulation]. An effect of this on
the last hundred years of global ATMOSPHERIC temperatures, was
discussed by K.K. Tung here at UW, and was reviewed by me in earlier
‘precis’ of this series.
Note that the near surface ocean [down to ~700 meters] contains
enormously more heat than the atmosphere. Treating heat transfer into
this layer as diffusive, [as Shaviv does], or studies that neglect
advective deep water formation over time scales on the order of a
decade or longer, are naive and irrelevant.
Or so say I.
hh
> o__
> _,>/’_ o__,
> (*) \(*) /\/\
*********************************************************