Sun and Climate; cosmic rays & clouds

Skip to first unread message

A Bjarnason

May 31, 1998, 3:00:00 AM5/31/98

The Economist April 11, 1998
had this interesting article:

--- --- ---


Global warming

*** A strange interaction between cosmic rays and clouds may
help to explain long-term changes in climate. ***

For a while now, some scientists have thought that global warming
may be less humanity's fault than the sun's. Since 1990 evidence
has accumulated that the earth's climate fluctuates on a cycle of
10-12 years, in parallel with the rise and fall of the number of
sunspots - transient dark patches on the face of the sun. This has
led to suggestions that the current rise in the earth's temperature owes

more to a bigger, longer term change in the sun's output -
an increase in the average number of sunspots measured over the course
of a cycle-and less to the greenhouse gases released by
burning fossil fuels.

The trouble with this idea is that, although a spottier sun
is indeed a brighter sun (the dark spots are accompanied by
bright patches elsewhere on the surface) it is
not brighter by enough to explain climate change directly
as we reported on February 21st, several theories about stronger,
indirect links are being looked at, but they rely on assumptions that
are hard to test.

There is, however, a newer theory circulating - and it has two
things in its favour. First, the circumstantial evidence for it is
mounting. Second, two physicists have just proposed a way to test its
possible mechanisms. If this is verified, they hope to put a skeleton
of explanation beneath the loose flesh of coincidence.

The theory comes from the same institution that found an
earlier bit of evidence linking sunspots and climate. In 1991 Knud
and Eigil Friis Christensen of the Danish Meteorological Institute (DMI)

in Copenhagen compared a century of air temperature records with
sunspot records.
They found a loose match between temperature and the number of
sunspots - but a much better one between temperature and the
length of the sunspot cycle. When the cycle was shorter (and the sun,
in general, spottier), the planet was warmer. In 1995, using indirect
measurements of the temperature going back four centuries,
they found the same thing.

**** A ray of enlightenment ***

During a short cycle the solar wind (a breeze of electrically
charged gas blowing off the sun's surface) is stronger than normal.
And when the solar wind is stronger, it is better at shielding the
earth from high energy charged particles from outer
space known as cosmic rays. So another researcher at
DMI, Henrik Svensmark, reasoned that if the solar cycle
were linked to the earth's temperature, cosmic rays
could be part of the chain.

In 1996 he and Dr Friis-Christensen found a place for them.
They compared the cosmic-ray flux with the proportion of the
world's skies obscured by cloud, a figure
that has been available from satellite photographs
since 1979. Again, the two measures kept closely in step with
each other, with the cloud cover swinging from 65%
when cosmic rays were weakest, to 68% when they peaked.

Why more cosmic rays should mean more clouds is not clear,
but there are at least two theories. One depends on the idea that
the water vapour which forms clouds condenses initially around
tiny particles known as aerosols (many of which are produced by algae).
These particles are attracted to electric charges. As cosmic rays
batter their way through the atmosphere, they knock electrons off gas
molecules, creating electrically charged ions. The ions in turn transfer

their charges to any water droplets that might be present, causing them
attract aerosols, which makes them more effective centers of

A second, more complex, idea has been developed by Brian Tinsley of the
University of Texas at Dallas. His theory, known as "electrofreezing",
observes that the ions created by cosmic rays make the atmosphere
more electrically conductive.
This should increase the current which normally flows from the top to
the bottom of the atmosphere, charging up any clouds that lie in
between. That charge helps small ice crystals to form. It is these
not aerosols, that act as condensation centres for further cloud

Regardless of which theory is correct (and both might be wrong), the
effect of fewer cosmic rays would be fewer clouds and - since clouds
to cool the earth by reflecting incoming solar radiation away
from it - a warmer planet.

This is certainly a tempting hypothesis. Although there were no
machines for detecting them systematically before 1935, cosmic rays
have left their footprints in the geological record by creating
elements such as the carbon-14 used in carbon-dating. This record
shows that large, long-term peaks and dips in the
cosmic-ray flux coincide with historical blips in the climate, such as
"little ice age" of the 17th and 18th centuries, when the Thames in
London often froze over, and the "medieval warm period" when Greenland
became mild enough for the Vikings to colonise it, temporarily.

Dr Svensmark has since been trying to strengthen the connection.
His (as yet unpublished) research confirms, he says, that the variations

in cloud cover seen by satellites match those in cosmic-ray flux
better than anything else, such as the sun's total brightness. He also
that wobbles in the earth's temperature have followed those in the
flux for the past 6o - odd years.

All this is highly suggestive, but not conclusive. However, Jasper
and Frank Close, two Britons working at the European particle-physics
centre, CERN, in Geneva, have in mind an experiment to find out if
either theory is correct.

Their idea (not yet a formal proposal) has a certain poetry. The
first particle physicists detected their subatomic subjects, including
cosmic rays, using devices called cloud chambers. These are boxes
containing air that is super-saturated with water vapour. This vapour
condenses into a trail of droplets as a charged particle zooms
through it. Dr Close and Dr Kirkby want to build a modified cloud

In this they would replicate the conditions at various levels
in the atmosphere the humidity, aerosol content and so on.
By firing particle beams through it, they would
then recreate the rain of cosmic rays and see whether clouds formed.
That should let them test at least the first idea, and possibly the
one as well. If either is right, Drs Close and Kirkby hope to be able to

tell just how big an effect cosmic rays should have on cloud formation.

By including such data in their computer simulations, climate modellers
may be able to tell whether solar inconstancy deserves some
of the blame heaped on greenhouse gases. Not that it would be
much comfort. Getting America to cut its emissions may seem hard.
Telling the sun to behave could be even trickier.

The Economist April 11, 1998

--- --- ---

Are our Suns's 11-year sunspot cycle and 80-year Gleissberg
cycle taking the earth's climate along for a ride?

Is CO2 not affecting climate as much as IPCC thinks?


Robert James Kunkle

May 31, 1998, 3:00:00 AM5/31/98

In article <>, A Bjarnason <> wrote:
>The Economist April 11, 1998
>had this interesting article:
...article cut, since you already read it...

Another question related to cosmic ray effect
on cloudiness; How is the creation of O3 in
the stratosphere affected by cosmic ray density?


Joshua Halpern

Jun 1, 1998, 3:00:00 AM6/1/98

In sci.environment A Bjarnason <> wrote:
: The Economist April 11, 1998
: had this interesting article:

: --- --- ---

Do try to keep up....

I'll repost Paul Farrar's latest comment on this numerology
since it was fairly complete, and fairly recent. I get the
feeling that it may reappear frequently, but there is no
point trying to improve on a good and witty post.

josh halpern

Subject: Sunspots and Climate? Latest from Denmark
From: (Paul D. Farrar)
Date: 1998/04/28
Message-ID: <>
Newsgroups: sci.environment
[More Headers]
[Subscribe to sci.environment]

In the Jan. issue of _Journal of Atmospheric and Solar-Terrestrial
Physics_, Peter Laut and Jesper Gundermann of Danish Tech. U. and
Danish Energy Agency, respectively, have published:

Laut, P. and Gundermann, J., 1998. Does the correlation between solar
cycle lengths and Northern Hemisphere land temperatures rule out any
significant global warming from greenhouse gases? _J Atm. & Sol.-Terr.
Phys., 60_, #1, 1-3.

This paper deals with the much-ballyhooed coincidence between
suitably-filtered sunspot cycle length and surface temperatures noted
by Friis-Christensen and Lassen (originally in _Science 254_, p698).
Their finding has been the basis for various far-reaching claims in
the popular and lobbying industry presses, in some cases by
astronomers and such who should know better. Laut and Gundermann show
that the curve fitting technique used by F-C&L also works for the
observed temperature plus additional strong warming, OR cooling. The
flexibility of the technique, which is unconstrained by any known
physical limit, allows the "proof" of many things, reducing its value
for proving anything in particular. L&G show that F-C&L cannot be used
as the basis for any claims as to the relative roles of various
forcings of global climate, such as solar variations or atmospheric
composition.This does still leave us with the much more cautious and
well-founded work of researchers such as Stuiver, Lean, or Wigley.

L&G do not consider in their paper some of the other problems of the
F-C&L paper, such as unusual and inconsistent time series analysis, or
the big problem that immediately sets off alarm bells for so many
readers: the winnowing of sets of time series to find matches,
followed by implications that any matches found are too good to be by
chance. Also, another sunspot cycle has ended, and I have heard that
it does not have a favorable value for the F-C&L conjecture.

I hear that worse may be in store for the F-C&L's idea in an upcoming
publication. (But you never know until things are in print.) The same
is true for the more recent paper by Svensmark and F-C, in which they
may have mistakenly attributed cloudiness variations during a recent
El Nino to cosmic rays. Personally, I think it likely that something
that warms the sea surface and pumps moisture into the atmosphere may
have affected the cloudiness more than cosmic rays. S&F-C should have,
at least, mentioned that an El Nino was occurring during their study
period (if they knew). It was certainly the first thing I thought of
when I saw their curve. Preferably, they should eliminate a major
confounding factor, such as this, before claiming causality. Sea
surface temperature trends during the period seem to be inconsistent
with their cosmic ray conjecture. I have heard that their cloudiness
hypothesis was featured, before its publication in _JA&S-TP_, in a
popular science book by Nigel Calder (of _The Weather Machine_ fame)
called _The Manic Sun_, but I have never seen a copy (of Calder).

Paul D. Farrar

R. T. Pierrehumbert

Jun 1, 1998, 3:00:00 AM6/1/98

Another problem with the FC&L idea on clouds and cosmic rays is that
can't even really say what is the DIRECTION of the effect on surface
As is well know, clouds can have a warming or cooling effect, depending
height and particle size. Low clouds are not limited by condensation
so it seems likely to me that the cosmic ray effect (if any) would give
to an effect mainly on high clouds. High clouds might be expected to
a net warming effect, but if the extra nucleation led to smaller
it could have a cooling effect. One could argue either way. It will be
a long time before the effect can be turned into a radiative forcing
in W/m**2 (even if the correlation indeed holds up). For example,
a 3% increase in cloud cover (e.g. from 30% to 33%) in the tropics would

have essentially no effect on the radiation budget if the IR/solar
balance for
clouds remained the same as at present. If one put the new clouds at
4 km higher than the present clouds, one could get a 4-6 W/m**2 warming
effect. If one put the 3% increase into low clouds (unlikely) with 50%
then depending on the latitude one could get a cooling effect of 4.5
or perhaps more.

R. T. Pierrehumbert

Jun 1, 1998, 3:00:00 AM6/1/98

>Are our Suns's 11-year sunspot cycle and 80-year Gleissberg
>cycle taking the earth's climate along for a ride?
To some extent, perhaps yes. It is within the realm of physical
that some of the temperatures wiggles of the past 100 years are due to
solar variation. A case for this could be made, if one could make a case

that sulfate aerosol effects are smaller than currently thought.

>Is CO2 not affecting climate as much as IPCC thinks?

In a word, no. The issue is how to explain the pattern of warming
of the past 100 years. Greenhouse gases alone give a monotonic
rise in temperature, whereas the observed record goes in fits and

I you take a look at the few studies that try to do a consistent
of solar effects in terms of a climate model, on an even footing with
CO2 effects -- notably the study by Soon, Posmontier and Baliunas
in Astrophysical Journal -- you find that the same model used to
account for past temperature fluctuations partly in terms of solar
variability in fact yields a warming of about 2C when subjected to
doubling of CO2. The solar variability issue has almost no implications

for the assessment of climate sensitivity. Take a careful
look at Soon et al, and you'll see what I mean.

Steinn Sigurdsson

Jun 2, 1998, 3:00:00 AM6/2/98

Joshua Halpern <j...@IDT.NET> writes:

> In sci.environment A Bjarnason <> wrote:

> : The Economist April 11, 1998
> : had this interesting article:

> : --- --- ---

> Do try to keep up....

> I'll repost Paul Farrar's latest comment on this numerology
> since it was fairly complete, and fairly recent. I get the
> feeling that it may reappear frequently, but there is no
> point trying to improve on a good and witty post.

Do you think Laut et al are the final word on this?
You are also being somewhat unreasonable in your critique,
I personally don't think Friis-Christensen and Lassen
are right, but they did propose a physical mechanism
that was testable, had some plausibility and provided
the right order of magnitude effects - that is not "numerology"
last I checked, and comments like that further polarise
what could be a scientific debate.

Reply all
Reply to author
0 new messages