Guest post by David Archibald
When I started out in climate science in 2005, the climate people ignored the solar physics community. A casual perusal of the literature though indicated that the difference in climate outcome from Dikpati’s (NASA) estimate for Solar Cycle 24 amplitude of 190 and Clilverd’s (British Antarctic Survey) estimate of 42 amounted to 2.0°C for the mid-latitudes.
Since then, the prognostications of astute scientists with respect to Solar Cycle 24 amplitude have come to pass. Some commentators though are over-reaching and predicting a recurrence of the Maunder Minimum. We now have the tools to predict climate out to the mid-21st Century with a fair degree of confidence, and a repeat of the Maunder Minimum is unlikely. A de Vries Cycle repeat of the Dalton Minimum is what is in prospect up to the early 2030s and then a return to normal conditions of solar activity, and normal climate.
The three tools we have to predict climate on a multi-decadal basis are the solar cycle length – temperature relationship, the logarithmic heating effect of carbon dioxide and Ed Fix’s solar cycle prediction. Let’s start with the solar cycle length – temperature relationship, first proposed by Friis-Christensen and Lassen in 1991. This is the relationship for Hanover, New Hampshire:
The relationship established for Hanover is a 0.7°C change in temperature for each year of solar cycle length. Solar Cycle 23 was three years longer than Solar Cycle 22, and thus the average annual temperature for Hanover, New Hampshire will be 2.1°C lower over Solar Cycle 24 than it had been over Solar Cycle 23. Why did I pick Hanover? Governor Lynch recently vetoed New Hampshire leaving the Regional Greenhouse Gas Initiative.
Professor Jan-Erik Solheim of Oslo University replicated this methodology for ten Norwegian temperature records, and thus this methodology is confirmed as valid:
These ten Norwegian temperature records all confirm a solar cycle length – temperature relationship, and predict that temperatures of these stations will be about 1.5°C colder over the next ten years than they have been over the last ten years.
The second tool to use is the logarithmic heating effect of carbon dioxide. The pre-industrial level of carbon dioxide in the atmosphere was approximately 290 ppm. It is currently 390 ppm. The first 20 ppm of carbon dioxide in the atmosphere provides half the heating effect to date. By the time we get to the current concentration, each additional 100 ppm provides a further 0.1°C of heating. We are currently adding 2 ppm to the atmosphere each year so carbon dioxide will provide further heating of 0.1°C every 50 years. That said, the temperature fall over the next 22 years should result in a higher rate of carbon dioxide uptake by the oceans. The logarithmic heating effect of carbon dioxide is shown by this graph, using data derived from the Modtran site at the University of Chicago:
Lastly, to put a multi-decadal climate forecast together, we need a prediction of solar cycle length that comes with a very good hindcast match. This is provided by Ed Fix’s long ephemeris simulation. This simulation is described in Ed Fix’s paper which is included in an Elsevier volume edited by Don Easterbrook, “Evidence-Based Climate Science”, due out in September. You can put advance orders in for it now:
This is a window of Ed Fix’s simulation:
The green line is the solar cycle record from 1914 to 2010, with alternate cycles reversed. Solar Cycles 19 to 23 are annotated. The red lines is the model output, from which the lengths of individual solar cycles in the mid-21st Century can be calculated.
Combining all the above tools, this is the climate forecast for Hanover, New Hampshire, which is a good proxy for what is going to happen along the US-Canadian border:
Solar Cycles 24 to 27 are annotated. For the next thirty years odd, temperatures will be at mid-19th Century levels. With the two year decrease in the length of Solar Cycle 26 from 25, temperatures will rise by 1.4°C by mid-century to late 20th Century levels.
By then, anthropogenic carbon dioxide will be providing a very welcome 0.2°C to the temperature.
The graph shows that quantified solar effects dwarf the quantified anthropogenic carbon dioxide effect.
David Archibald
12th July 2011
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Bruce of Newcastle – I didn’t have to poke holes in Friss-Christensen and Lassen (1991), the holes had already been poked, one of them by Lassen himself. The fault here is with Archibald for covering over the holes, which as I said is deeply unscientific, as it suggests that his analysis has the backing of solid peer-reviewed science – it doesn’t.
Also J took the time to try out the correllation on the HADCRUTv3 as you suggested and found that the correllation did indeed break down, just as Thejll and Lassen suggested. It is telling that you do not admit that you were wrong on the HADCRUTv3 correllation.
An analysis that only shows correllations if you select the right dataset is non-robust, especially if an argument about global climate can only be made if you concentrate on e.g. Central England, Norway and Hannover. Station and sub-regional datasets tend to be very noisy, which means that you’ll generally be able to cherry pick one with a good correllation with more or less anything. However that doesn’t mean the correlation is meaningful.
Richard S Courtney:
Thank you for your enlightening and helpfull comment at 12:24 pm above.
I have read Hans Jelbring in the DragonSlayer series of essays. I try to use logic and personal experience to help me understand what happens in the atmosphere. I once had the priviledge of visiting one of South Africa’s deepest gold mines on the Vaal Reef and commented on the high temperature. It was explained to me that this was due to the increased air pressure at some 2,000 metres below sea level. As it was not possible to pump cool air down to that depth, water was employed to control the working environment.
I understand the difficulty of attempting to explain the effect of gravity on the adiabatic lapse rate. For example; take a well sealed, but not well insulated container, and increase the air pressure to say 2 bar. The temperature of the air in the container increases. After a while the temperature of the air in the container will reduce to the local ambient temperature but the pressure in the container is still at 2 bar. So the way I explain it is that the kinetic energy of the air molecules in the container are less than the kinetic energy of the air molecules outside the container. Thus we have two different levels of pressure at the same temperature. Now, release the air from the sealed container and it will feel quite cold! 🙂
Ooopsss… I said below “below sea level”… I should have said below ground level… 🙁
The Sunspecs
“Would Galileo Have Counted These As Sunspots”
All of this is is just pure observational data and have no, I repeat NO, actual basis in… reality… I mean the models of course.
Maybe it got missed along the way, but … is anybody able to explain what looks like a major calculation error in Archibald’s forecast?
Archibald’s fig. 5 (predicted solar cycle length from Ed Fix) shows the next four solar cycles all being just 8-9 years long — shorter than any in the entire history of the Hanover met station. By Archibald’s model, then, temperatures in Hanover ought to be plummeting to Little Ice Age levels between now and 2017, then turning on a dime and soaring to unprecedented heights from 2017 to 2053.
But that’s not at all what Archibald’s fig. 6 shows.
I’m surprised that no one has picked up on this — it’s not just that there’s been no correction of the post, but it hasn’t even been remarked on in the comments.
Am I missing something completely obvious? Or is Archibald’s forecast wildly incorrect, based on his own model?
I hope someone else will chime in on this, because I’m starting to doubt my own grasp of basic algebra here.
Bruce (of Newcastle), thanks for the reply. It sounds like you’re suggesting using multiple regression to predict temperature as a function of both log2(CO2) and pSCL … am I understanding that right? In other words, take the CO2 data for a given solar cycle and convert it to a number of doublings above 280 ppm, as the first independent variable, and the length of the previous solar cycle as the second independent variable.
I like that suggestion, since it would be a great way to compare the relative effects of the two factors.
But I don’t quite get how you used the Mauna Loa CO2 data with CET temperatures, since Mauna Loa only goes back to ~1959.
As a quick test, I just took the Law Dome CO2 data (with a 20-year smooth) for the period up to 1960, and the Mauna Loa data from 1960 onwards, and calculated the mean CO2 during each solar cycle from the 1850s onward. I then calculated the number of doublings relative to the preindustrial 280 ppm (this was 0.03 doublings for solar cycle 10, up to 0.48 for the current cycle 24). I used this in a multiple regression model along with the length of the preceding solar cycle, to predict HADCRUT temperature as you suggested earlier.
For cycles 10-24 (the latter not being complete yet), the model has an adjusted R2 of 0.87, and a standard error of 0.087C.
The coefficient for CO2 is 1.9C/doubling (95% CI 1.5 – 2.2).
The coefficient for pSCL is -0.03/year of solar cycle length (95% CI -0.09 to +0.02, so not significantly different from 0).
Again, thanks for suggesting this. I don’t really “believe” in the results, since it’s still pretty simplistic analysis and it doesn’t account for stuff like aerosols, ENSO, volcanoes, cloud feedbacks, non-CO2 greenhouse gases, or the rest of the forcings. But it does seem to indicate that the solar cycle length might be a bit of a red herring.
Magnus says: “All of this is is just pure observational data and have no, I repeat NO, actual basis in… reality… I mean the models of course.”
Good point. We’re now about a third of the way through solar cycle 24. Does anyone know what the observational data from the Hanover met station look like, over the past couple of years? Is New Hampshire experiencing a “Little Ice Age” style cold snap, as Archibald’s model predicts? I think there’s a way to get station data from Weather Underground, but I don’t remember exactly how. Just curious….
J asked (July 15, 2011 at 3:25 am) “Am I missing something completely obvious? Or is Archibald’s forecast wildly incorrect, based on his own model?”
Not mutually exclusive.
“”””” Richard M says:
July 14, 2011 at 6:42 am
First I would like to thank George Smith and Richard S Courtney for getting to the meat of the issue.
Richard states: And you conclude that statement saying;
“as well as additional water vapor, which directly blocks additional solar spectrum energy from reaching the surface.”
Perhaps, but if it does “block” that solar energy then the “blocked” energy is added to the atmosphere and contributes to back radiation because molecules can be collisionally excited as well as collisionaly de-excited. So, this does not affect my explanation (which I admitted is simplistic).
Yes, it does contribute to back radiation and that radiation eventually is radiated back to the atmosphere . IWO, back radiation is a non-issue. “””””
First let me say that Richard (S. Courtney) is much closer to this subject than I am, and further more, I don’t have any access to some teracomputer or any modtran calculations of what CO2 or any other so-called GHG such as H2O or O3 is supposed to absorb in the way of electro-magnetic radiation energy.
The news that such absorption of energy does not alter the atmospheric Temperature, is indeed a surprising revelation to me. But then I had always had this simplistic view, that “Temperature” was simply a measure of the purely mechanical energy stored in atoms or molecules (materials), by reason of the various modes of atomic or molecular “vibrations/rotations/translations/whatever”, so that Temperature has no meaning whatsoever in the absence of physical materials, and in particular is not a property of electro-magnetic radiation, as described by James Clark Maxwell, and his famous equations.
But every credible plot that shows the spectrum of incoming solar energy, outside earth’s atmosphere (air mass zero TSI of about 1362 W/m^2 (according to a recent NASA report), shows a nearly black body like curve corresponding (roughly) to about a 6,000 K black body Planck radiation curve.
But at the earth surface under normal incidence (air mass one), the spectrum is very different, although still somewhat BB like. The shorter wavelength regions (blue-green) are considerably attenuated evidently as a result of first, absorption by Ozone (O3); and please note, when I use the term “blocked”, I do not mean to the point of total extinction; only some photons are blocked, and in their case, it most certainly is to the point of extinction; those are permanently removed from the incoming solar spectrum stream. And of course there is the very important primarily Raleigh scattering of the blue-green region of the spectrum, which results in the day sky being blue, rather than black with stars.
This blue sky is itself instructive, because it converts a nearly collimated near point source of blue-green radiation (sunlight) into a nearly isotropic diffuse source of same spectrum radiation that fills the whole sky.
Moreover, as anyone who has ever flown to Hawaii, in daylight can attest, the blue sky looks the same looking down, as it does looking up. When you look up in daylight, the black sky with stars, is completely washed out by the blue scattered light. Well actually you can observe the stars n broad daylight, if you can just point your telescope to the exact direction of the star; they then become quite visible.
And when looking down, you do not see the black ocean (2% reflectance) or the wind blown white cap “stars”. Once again, if you can point your telescope downwards, and focus on one of those ocean stars, they too become quite visible.
The whole point of this explanation, is that any isotropic source of radiant energy looks the same from any direction. Skiers and other snow bunnies know full well what a white out looks like.
This is exactly why the isotropic thermal radiation (LWIR) from the atmosphere is a roughly 50-50 split of upwards, versus downwards. It matters not a jot, whereabouts in the altitude range some sample atmospheric layer is, the isotropic emissions from THAT layer split about 50-50, upwards versus downwards. There is nothing magic about that particular sample layer, when it comes to absorption and emission of EM radiations. The layer above, and the layer below, will do exactly the same thing; doesn’t matter whether we talk about a 100 metre thick layer of a 100 micron layer; its nearest neighbor layers, will absorb and emit pretty much the same.
But, as I pointed out, the upper layer (100 microns) is both colder, and lower density, as compared to the adjacent lower layer (100 microns), so the molecular or atomic absorption spectral lines or bands will be narrower for the upper layer, than they are for the lower layer. A given radiated spectrum, will be less gutted by a narrower absorption band above, than a wider one below. This near 50-50 split applies to every single layer, of whatever thickness, but the escape path upwards, is slightly favored, over the downward path, where recapture is more likely.
But the bottom line end result, is that about half of the total isotropic emissions from the atmosphere must escape to space, and about half will eventually reach the surface; with the caveat, that we are talking about radiation effects only. Other energy bled off as a result of thermal (heat) processes is a different isue.
So when the dust settles, ANY solar energy, that is extracted from the incoming nearly collimated sunbeam, by H2O,O3, or CO2, and is subsequently re-incarnated in some form as atmospheric LWIR emission, will have half escape to space, and only half reach the ground; which in my system of mathematics is ALWAYS a net loss of solar energy to the earth.
Water in any form anywhere in the atmosphere MUST always result, in a net loss of solar energy to planet earth; and you don’t need any modtran or teracomputer to see that. Clouds also, whatever they may do to LWIR radiation ALWAYS result in LESS solar energy reaching the surface, where it can be stored as part of earth’s energy budget. Nobody ever observed it to warm up in the shadow zone of a cloud. The Temperature always drops; it never increases; despite what LWIR from the surface may get returned from that cloud; ANY cloud, ANYWHERE.
And for the legal disclaimer; the intrusion of some other warm AIR mass from some other region is a convection process, and unrelated to solar energy blockage.
How would volcanic activity at various rates (e.g. small eruption, large eruption, increased activity, decreased activity) affect this forecast?
Has David Archibald or anyone else for that matter got a link to the Hanover data used in the analysis.
I predict that it will turn out that each branch of science focuses on their own speciality and we get either global warming or global cooling. The real world does not recognise these man made boundaries so we will end up with the random events like Icelandic volcanoes chucking up muck ruining our summers and quiet years when we get good ones but overall the same tiny insignificant rise that has happened for thousands of years.
@J says:
July 15, 2011 at 3:25 am
“..is Archibald’s forecast wildly incorrect, based on his own model? ”
Yes, 8yrs long for SC25 (2017 to 2025) would imply by this theory, a very hot period indeed.
http://wattsupwiththat.com/2011/07/13/archibald-climate-forecast-to-2050/#comment-699216
J says:
July 15, 2011 at 3:25 am
Maybe it got missed along the way, but … is anybody able to explain what looks like a major calculation error in Archibald’s forecast?
I’ve only just read this post and I think Anthony (Watts) should consider asking David Archibald to clarify the issue. J is right. The ‘Ed Fix’ Solar Cycle length sequence for SC24-SC27 is 9, 8, 9, 9 years respectively. Based on David’s calculations that means temperatures will drop by 2.1 deg over SC24 but then rise by ~2.5 deg over SC25 and a further ~0.7 deg over SC26.
I notice one or two other readers have noticed the same thing (Ulric for one). I admit I missed it initially but there’s obviously a problem when you look at it closely. It’s complete nonsense and, lefit as it is, it doesn’t do much for the reputation of WUWT
Apropos sunspots…
“15 Jul 11 – “NOAA is inflating their Sunspot count by counting specks,” says reader Al Morris. To get a more scientific count that compares to the old Wolf Sunspot number please review the Landscheidt website.”
http://www.landscheidt.info/?q=node/50
The purported logarithmic relationship between the atmospheric abundance of CO2 and the global mean surface Temperature; which cannot be confirmed using the available credibly measured atmospheric CO2 abundance that has been going on at Mauna Loa, since the International Geophysical Year in 1957/58, is alleged to have its theoretical basis in Beer’s Law; or the Lambert-Beer Law as it is sometimes called; which is a law relating to chemical solutions of light absorbing atoms/moleculaes, in non-light absorbing base liquids.
In essence the law says that the the fraction of an incident light beam that is absorbed in a fixed thickness of the solution, is proportional to the molar concentration of the absorbing species; it also says that each constant increment of thickness (optical path length) absorbs the same fraction of the beam incident on that tickness element.
This leads to an absorption equation of the form:
(a) = 1-exp (alpha. t), where alpha is the absorption coefficient, and (t) is the thickness of the optical path.
Careful texts do point out that Beer’s Law is an approximation, and is valid only for small concentrations of the absorbing species.
Some other things the reader is urged to note. Beer’s law relates only to the direct absorption, of the incident light (EM radiation).. In particular it does NOT say anything about the absorption of energy, or the transmission of EM energy. Only absorption of the incoming incident radiation is affected by Beer’s law, and the assumption is that some of that incident radiation is absorbed, and is NOT subsequently re-emitted as some other EM radiation at some other wavelength or frequency (wve number).
Beer’s Law, does NOT have anything to say about the net EM energy transmission of any photo-luminescent material or atomic/molecular species. It assumes that the absorbed EM radiation that was input, does not result in any subsequent radiative transition, but that only phonon transactions take place, and the energy is dissipated as heat, and ultimately conducted (or convected) out of the solution. Given enough thickness of the absorbing material, the net energy transmission ultijmately approaches zero, in the usual exponential fashion.
And there is where the problem arises in atmospheric absorption of EM radiation, either directly of solar spectrum radiations, as H2O, O3, and CO2 do to different extents, or as LWIR radiation absorbed from the earth’s surface; or simply from some other part of the atmosphere.
At ordinary atmospheric Temperatures, the LWIR emissions from the various atmospheric layers, is quite similar in wavelength to that emitted from the surface; and moreover, the atmosphere is continuously absorbing, and re-emitting LWIR radiant energy at pretty much the same range of wavelengths.
Beer’s law simply does not cover such processes. The energy can be largely transmitted, with relatively little loss, in a continuous cascade of emission and absorption. In particular, a specific absorption band such as the CO2 15 micron band, can hardly become “saturated”. More CO2 simply means a thinner layer of atmosphere, is needed to achieve a given percentage of absorption, and the cascade of emissions and absorptions, simply becomes longer.
LWIR emitted energy from the surface, is not simply blocked by the first few cm or metres of atmosphere; it undergoes a continuous string of absorption and re-emissions, that ultimately propagates some of that surface energy to outer space, as does the energy emitted from any layer of the atmosphere. Beer’s Law does not describe such processes.; it applies only to the exponential absorption of the original incident radiation.
It certainly doesn’t establish that the logarithm of the atmospheric abundance of CO2 or any other so-called GHG is linearly related to the global mean surface Temperature.
Global Orbit Decay (Updated Version)
By: Willie McDonald
cdnld30@gmail.com
The events below were discovered by scientific organizations such as NASA, not by me! These events were discovered in 20th century, and are occurring simultaneously, and is slowly worsening. Many of of these events have been occurring for less, than a millennium (LTM). I believe they are now beginning to affect the earth’s climate. I’m a climate change expert with 28 years of experience. Green house gases has nothing to do with global warming. Many scientists believe the sun plays a larger role in climate change, than first thought. Global warming can be reversed go to http://www.orbital-decay1.blogspot.com, but if its not reversed in time all life on this planet will perish. Below are the reasons I believe the earth’s orbit around the sun is destabilizing, and is responsible for global warming.
1. The earth is moving away from the moon.
A. http://www.uni.edu/morgans/astro/course/Notes/section4/new17.html
2. The earth’s rotation is slowing down. (LTM)
A. http://bowie.gsfc.nasa.gov/ggfc/tides/intro.html.
3. The earth is shifting on its axis. (LTM)
A. http://divulgence.net/axis%20shift%202.html.
4. The earth is wobbling on its axis
A. http://www.world-weather.com/world-weather/our-wobbling-earth-wobbled-by-the-worlds-weather/
5. The earth is developing a breach in its magnetic field. (LTM)
A. http://www.science.nasa.gov/science-news/science-at-nasa/2008/16dec_giantbreach
6. Both polar ice caps are beginning to be melted by the sun, during each ice caps summer season, and the oceans are rising. (LTM)
7. The sun is getting hotter, and brighter. It’s possible earth is moving closer to the sun. (LTM)
A. http://www.rs2theory.org/node/106
B. http://telegraph.co.uk/science/science-news/3325679/the-truth-about-global-warming-its-the-sun-thats-to-blame.html
C. http://www.nasa.gov/topics/solarsystem/features/sun-brightness.html
D. http://www.nasa.gov/centers/goddard/news/topstory/2003/0313irradiance.html
8. Many scientists are how saying that the IPCC report, and the green house gas theory is false.
A. http://www.jimball.com.au/features/IPCC-False_evidence.htm
B. http://www.mclean.ch/climate/IPCC.htm
C. http://www.petitionproject.org
References: NASA, NOAA, USGS, The American Astronomical Society, etc.
If the links don’t open when you click on them, please enter them by hand.
People shade themselves from the sun, not from the greenhouse effect. You will never get sun burn, sun stroke, or skin cancer from the green house gas effect, beware of the sun.
“”””” Taylor says:
July 14, 2011 at 3:00 am
david…thanks for the link to RealClimate…but i was hoping you had more than that! Their discussion goes back to Idso’s work…a long way back, and though that was empirical, it is not clear how it relates to the issue as it dealt with surface flux and temperature response.
The modtrans data is tricky to interpret “””””
I would hardly call Modtran “data”. It’s a computer model.
But then I notice that even guys like Peter Humbug talk about Modtran simulations as if they are real data; he even claims that such simulations “confirm” the various hypotheses about what CO2 does or does not do. So laboratory experiments confirm the logarithmic relationship between atmospheric CO2 abundance, and global mean surface Temperature.
I would like to personally observe such an experiment in progress; of course using a REAL source of 288 K black body like LWIR radiation, such as is (on average) emitted by the earth surface, and having it peak pectral radiance at about 10.1 microns, and a Planckian spectrum covering say 5.0 to about 80.0 microns, which would contain about 98% of the total energy contained in a full BB spectrum at that 288 K Temperature.
If your source of LWIR EM emissions does not have an equivalent BB Temperature of around 288 K, then please don’t claim to be conducting a lab experiment to confirm CO2 “greenhouse” phenomena..
Geoff Sharp says:
July 13, 2011 at 5:28 pm
I appreciate your concern but solar cycle length explains up to 90% of climate variation in some locations. It is just about all we need to know. Everything else is noise.
F. Ross says:
July 13, 2011 at 8:33 pm
See comment re Geoff Sharp above – up to 90% of climate variation (paper in press). What we don’t know is the mechanism by which a longer solar cycle means less solar magnetic activity in the following solar cycle. Understanding the mechanism is something for the solar physicists to get on to.
Geoff Sharp says:
July 13, 2011 at 11:00 pm
I have a paper in the same volume and the peer review was rigorous. Elsevier need to maintain their reputation well after this climate thing has blown over.
Geoff Sharp says:
July 13, 2011 at 11:40 pm
Sorry Geoff, I have played with Ed’s model live. Its fine detail is a lot better than yours or anyone else. The need to reset is a feature, not a bug, and tells us something about the solar process.
Peter Taylor says:
July 14, 2011 at 3:00 am
Real Climate couldn’t break my graph which was inspired by a line graph by Willis Eschenbach. It’s simple. CO2 is 10% of the 30 degrees of greenhouse effect. That is 3 degrees. Then apportion the 3 degrees by 20 ppm increment. If the first 20 ppm provides half the heating effect, then it gets 1.5 degrees and so on. Once you have figured out something that works, no need to overwork the problem.
J says:
July 14, 2011 at 4:52 am
Sorry J, what was unsaid was that interpretation of Ed’s graph is subjective. My take is that we will have two long, weak cycles instead of the three short cycles as predicted by the graph. This is all to do with the reset required by Ed’s graph. What I am trying to say in this post is that we are not headed for a Maunder Minimum.