Added 12 June 2010: There is now a thoughtful reply from Eimear Dunne.
Climate Change – News and Comments, and an echo of a Falsification Test
Wake up! Models don’t trump observations
The most important article since I launched this blog on 1 May may be Do clouds disappear when cosmic rays get weaker? — see here https://calderup.wordpress.com/2010/05/03/do-clouds-disappear/
It tells of unremitting attempts to falsify the Svensmark hypothesis by claiming that there’s no important effect on global cloud cover when eruptions from the Sun briefly cut the influx of cosmic rays, in “Forbush decreases”. The centrepiece is a summary of results published last year by Svensmark, Bondo and Svensmark. They show very plainly, in observations of the real world, that Forbush decreases have big impacts both on aerosols (chemical specks that grow into cloud condensation nuclei) and on low-level clouds.
That earlier post continues with the efforts in 2009-10 by Wolfendale and Arnold and their collaborators, who try to deny the Svensmark group’s result, by using relatively weak Forbush decreases. Svensmark can explain exactly how the impacts in those cases are masked by quasi-random meteorological noise, like tigers hidden in a jungle’s undergrowth.
Real-world results by Svensmark, Bondo & Svensmark (2009) for the remarkable loss of fine aerosols from the atmosphere (black curve) following five strong Forbush decreases in cosmic rays (red curve). Each aerosol datum point is the daily mean from about 40 AERONET stations world-wide, using stations with more than 20 measurements a day.
New nonsense comes in an abstract posted on the CERN website. It’s for a paper by researchers at Leeds, to be presented at a meeting about aerosols in Helsinki in three months’ time.
At issue are the Svensmark team’s results on aerosols (see right). These show fine aerosols disappearing from the sky, because the shortage of cosmic rays lessens the chemical production of the clusters of sulphuric acid and water molecules that seed the aerosols.
According to the people in Leeds, that can’t be right because they have a computer model that contradicts it.
The GLOMAP model was developed by Ken Carslaw, and the unlucky person named as lead author is a graduate student, Eimear Dunne.
An open letter to the lead author
Dear Eimear,
In any other branch of physics, if a model and observations are at odds, there’s almost certainly something wrong with the model. But you’ve evidently been encouraged to think that doesn’t apply in climate research.
I must admit that you have a dreadful role model in the Intergovernmental Panel on Climate Change. It keeps shrugging off glaring mismatches between real-world data and the models used to predict man-made global warming.
As for your new association with the CLOUD experiment at CERN, you may not know that the project was conceived as a direct result of a lecture that I gave at CERN in December 1997, reporting Henrik Svensmark’s discovery of the influence of cosmic rays on cloud cover.
But you should also be aware that, although Henrik inspired the project, some people in CLOUD team now try to disparage his research whenever they can.
Why? Because Henrik keeps insisting, in a politically incorrect manner, that the Svensmark effect is important. Crucial, in fact, for understanding past, present and future climate change.
Fainter hearts would like the link between cosmic rays and clouds to be just a technical footnote to the climate debate. Not so trivial, mind you, as to undermine the case for spending public money on CLOUD. But not so significant, either, as to alarm the politically correct funding agencies.
It grieves me, Eimear, that your mentors have launched you into such a difficult balancing act. It’s bound to produce wobbly results.
But in any case I know that you, together with Prof. Carslaw, signed a declaration in December saying, ‘As professional scientists, from students to senior professors, we uphold the findings of the IPCC Fourth Assessment Report, which concludes that “Warming of the climate system is unequivocal” and that “Most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations’.
So you’re not exactly open-minded about the Svensmark hypothesis. You really don’t want to find an important effect of cosmic rays, do you?
Best wishes
Nigel Calder
I’m notifying Eimear Dunne of this post and will welcome any comment that she may care to send.
Her abstract is shown in full after the references.
References
H. Svensmark, T. Bondo and J. Svensmark, Geophys. Res. Lett., Vol. 36, L15101, 2009
E. Dunne et al. Abstract submitted to IAC2010, Helsinki, 29 Aug.– 3 Sep. 2010. See in full below or on p.16 in http://cdsweb.cern.ch/record/1257940/files/SPSC-SR-061.pdf
GLOMAP http://researchpages.net/GLOMAP/
Declaration signed by Dunne & Carslaw http://www.timesonline.co.uk/tol/news/environment/article6950783.ece
Abstract in full
The global impact of a simulated change in the nucleation rate on atmospheric aerosol
Dunne, Eimear1, Carslaw, Kenneth1 and the CLOUD collaboration
1School of Earth and Environment, University of Leeds, LS2 9JT, UK. Keywords: ion-induced nucleation, Forbush decreases, galactic cosmic rays, modelling.
A link between cosmic rays and the climate is supported by several empirical datasets (Kirkby, 2007). The CLOUD consortium was formed to investigate the mechanisms by which the climate might be altered by a modulation in the intensity of cosmic rays. A likely candidate for this climate mechanism is ion-induced nucleation, the process by which atmospheric ions facilitate the formation of sulphate aerosol from sulphuric acid vapour and water vapour (Carslaw et al., 2002).
The Global Model of Aerosol Processes (GLOMAP) developed at the University of Leeds is an ideal tool for assessing the impact of a change in nucleation on the climate (Spracklen et al., 2007). GLOMAP features online microphysics, including nucleation, condensation and coagulation, and simulates the behaviour of black and organic carbon, sea salt and sulphate aerosol. In the longer term, we will use GLOMAP to model the interactions of ions and aerosols explicitly.
For this study we have used an existing nucleation mechanism included in GLOMAP (Vehkamäki et al., 2003) to place some bounds on the expected response to changes in the particle formation rate due to galactic cosmic rays (GCRs). Increased solar magnetic activity causes a decrease in GCRs on three time scales. The best known is the same 11-year solar cycle which modulates sunspots. Other longer cycles provide an amplitude variation of the 11-year cycle. On shorter time scales, however, powerful ejections of charged particles from the sun can cause reductions in terrestrial GCR levels over a period of days to weeks.This is known as a Forbush decrease.
There has been great interest in the recent Svensmark et al. (2009) paper, which shows a correlation between Forbush decreases and observations of Angström exponent, cloud water content, liquid water cloud fraction (LCF) and low infra-red-detected clouds. However, other authors have since examined the proposed trends in LCF and global cloud cover, and found no statistically significant correlation in the data.
To simulate the effect of a Forbush decrease on atmospheric aerosol, GLOMAP was run for ten days with daily output. The nucleation rate was then decreased globally by a factor of {0.85, 0.5, 0.15, 0} for two days. A control run with no decrease in nucleation was also simulated. The daily model output was analysed for a further 15 days in each scenario to assess the impact of a two-day Forbush decrease with a range of magnitudes on the aerosol size distribution.
In the free troposphere, concentrations of condensation nuclei responded rapidly and substantially to the change in the nucleation rate, but soon returned to normal levels. In the boundary layer, condensation nuclei took longer to recover because the particles are partly supplied from coagulated particles transported down from the free troposphere (Merikanto et al., 2009). Only a minor reduction in cloud condensation nuclei was observed. A longer event at 85% nucleation will be simulated, as well as a Forbush decrease in a pre-industrial atmosphere where it may have a larger impact.
Metzger et al. (2010) found that a nucleation rate of the form J = k [H2SO4]m ×[NucOrg]n produced better agreement with observations. Combining this kinetic mechanism with the findings of the CLOUD 2009 experiment, we will suggest upper bounds for the impact of the enhancement in nucleation caused by ions on condensation nuclei, cloud condensation nuclei and cloud droplet number concentration.
We would like to thank CERN for supporting CLOUD with important technical resources, and for providing a particle beam from the CERN Proton Synchrotron. This research has received funding from the EC’s Seventh Framework Programme under grant agreement no. 215072 (Marie Curie Initial Training Network “CLOUD-ITN”), from the German Federal Ministry of Education and Research (project no. 01LK0902A), from the Swiss National Science Foundation, and from the Academy of Finland Center of Excellence program (project no. 1118615).
Carslaw, K.S., Harrison, R.G., and Kirkby, J. (2002). Cosmic Rays, Clouds, and Climate. Science, 298, 1732-1737.
Kirkby, J. (2007). Cosmic rays and climate. Surveys in Geophysics. 28, 333-375.
Merikanto, J. et al. (2009). Impact of nucleation on global CCN. ACP. 9, 8601-8616.
Metzger, A. et al. (2010). Evidence for the role of organics in aerosol particle formation under atmospheric conditions. PNAS.
Spracklen, D.V. et al. (2007). Evaluation of a global aerosol microphysics model against sizeresolved particle statistics in the marine atmosphere. ACP. 7, 2073-2090.
Svensmark, H., Bondo, T., and Svensmark, J. (2009). Cosmic ray decreases affect atmospheric aerosols and clouds. GRL. 36, L15101.
Vehkamäki, H. et al. (2003). An improved parameterization for sulfuric acid-water nucleation rates for tropospheric and stratospheric conditions. JGR (Atm). 107, 4622-4631.
Calder's Updates 
If the authors already signed a pact with the High Priests of the IPCC their work is biased towards the agenda of the IPCC and not independent research. Those who sign the pact only find results that are predetermined by tweaking their computer games. Sort of like rewriting their solitaire programs to win all the time!
“is VERY LIKELY due to the observed increase”
My old science master (bless his dear departed soul) would have kept me behind after class if that had ever been in a submitted experiment report!
A fine new site Mr. Calder. Please keep up the interesting posts.
I’m holding my breath while waiting for a response from Eimear.
Dear Nigel,
Models may not trump observations, but they do have an important role to play as a scientific tool by which we can test our hypotheses about systems which are too large or too complex for a laboratory experiment. They are an especially important tool for atmospheric scientists, who often make point observation of inhomogeneous systems and then need a means to extrapolate to regional or even global scales, and make predictions based on past observations. Models are particularly vital when attempting to determine the extent of the cosmic ray-cloud-climate effect, given our limited understanding of its mechanism.
My abstract refers to an attempt to place an upper bound on the impact on aerosols of a reduction in the nucleation rate – the mechanism by which Svensmark et al. suggest their observations of cosmic ray and climate data are linked. This is not a test of what they have measured, but rather of how they have interpreted it. Svensmark et al. can hardly object to a model being used as a comparison for their results, given that one of their number, Torsten Bondo, has recently published a similar (but more favourable) comparison (http://www.atmos-chem-phys.net/10/2765/2010/acp-10-2765-2010.pdf). I leave the task of questioning the observations made in the Svensmark et al. paper to Calogovic et al. (http://www.agu.org/pubs/crossref/2010/2009GL041327.shtml) and Laken et al. (http://www.agu.org/pubs/crossref/2009/2009GL040961.shtml), and instead focus on testing the theory behind the paper.
On both a personal and professional level, I would be delighted to find an important effect of cosmic rays. I would be overjoyed to discover that the many observations of an increasing global temperature are a natural fluctuation of the Earth’s climate. But unless we can put forward a quantifiable mechanism, test it using every available method, and determine its extent, as we are doing to the best of our ability in relation to anthropogenic global warming, this theory falls short of being classed as true science. We at the CLOUD project hope to change that, by conducting experiments to test the theory and increase our understanding.
Please don’t dismiss me as an unlucky person. I am a scientist participating in a ground-breaking experiment on an international level, with peers whom I like and respect. I accept that as a student I have a great deal left to learn, but we are trying to better understand a physical phenomenon which we believe to be important using all available means, and whatever the outcome, that is science in its purest form.
One final comment: I am glad to see that you have taken down my picture. Regardless of whether said image is available elsewhere online, it is rather poor form to include a person’s picture in your blog without their permission. It also detracts somewhat from the discussion at hand. This should be about facts and ideas, not faces.
I look forward to further discussion with you or your colleagues at IAC.
Kind regards,
Eimear Dunne
Thank you for your thoughtful comments, Eimear, and I’ll draw attention to them at the start of the item.
I’m happy to let others consider what you say without any further remarks from me.
And abject apologies for including your picture initially, which was as you say, bad form. It was Henrik Svensmark who told me I ought to remove it.
Nigel
Were you able to identify the uncertainty of the results from the simulations? (and indeed the experiments)
I still look towards an explanation why Dr.Svensmark isn’t a member of the CLOUD team.
Hugo. naturally as the discoverer of the cosmic-ray/clouds connection, Henrik Svensmark was a member of the original CLOUD team. His institute provided much of the equipment and manpower for the “CLOUD prototype” experiment. Why he and his Danish colleagues are no longer members of the team is a question that should properly be addressed to the rest of the CLOUD team and to its spokeman Jasper Kirkby. A hint of the explanation is in my remark to Eimear Dunne about tightrope walking. But I really don’t want needlesly to open any old wounds about this matter, which from a scientific point of view is inconsequential. The Danish team already has the key results expected from CLOUD, from its own experiments.
Nigel
Despite her thoughtful response, it still leaves a significant question mark over Eimear’s involvement as lead author.
How can she possibly maintain the necessary enthusiasm and energy to fully investigate the science if she believes that it is a waste of her time?
True science is more than just doing paid research.
Truly great scientists (like Svensmark) evolve and develop ideas and hypotheses which are essential elements of their personality.
They will pursue these in spite of everything that’s thrown at them, and sometimes they will fail.
But that failure would only come after all avenues have been fully explored and all possible options exercised.
Now tell me that someone who has strong beliefs is motivated to investigate openly, thoroughly and honestly something they know in their hearts to be wrong.
A little bit disappointed in Eimear’s response. She appeared to be more interested defending a position, than digesting and responding to Nigel’s open letter.
With regard to modelling complex systems vs empirical data. Modelling can only include what is known, and what is assumed to be known. That is why it is a model and not an equation. So unles it is limited to demonstrated laws of physics, it will exclude necessary parameters and include assumed, but not neccessarily correct parameters. It will be a scientific “best guess”
As far as signing of the declaration is concerned, the thought sends shivers down my spine. Eimear if you read this I would sincerely hope that you’re allegiance to scientific integrity far outweighs your allegiance to the IPCC and AR4.
In my experience with the scientific community, modellers are scientists who know things, empericists are scientists who understand them.
It seems as if Ms. Dunne is following a testing plan that is sequential and evaluates pieces of the puzzle before expanding the test to fit the actual world. We all want to jump on the Big Question and get an answer. But perhaps a step-by-step process is better, so that later interpretations won’t be “clouded” by arguments about internal, bottom-level issues.
A model is as good as the plugs that keep it from leaking to death.