Spot the science error

Guest post by Dr. Leif Svalgaard

The following abstract of a poster to be presented next month at the Fall Meeting of the American Geophysical Union caught my eye:

Session Title: GC11A. Diverse Views From Galileo’s Window: Solar Forcing of Climate Change Posters Chair: Willie Soon, Nicola Scafetta, Richard C Willson

ID# GC11A-0685: Dec 14 8:00 AM – 12:20 PM

Revised Assumptions and a Multidiscipline Approach to a Solar/Climate Connection

C. A. Perry (US Geological Survey, Lawrence, KS, USA).

Click to enlarge

Abstract:

The effect of solar variability on regional climate is examined using a sequence of physical connections between solar variability , Earth albedo, ocean temperatures, ocean currents (Ocean Conveyor Belt), and atmospheric patterns that affect precipitation and streamflow. The amount of solar energy reaching the Earth’s surface and its oceans is thought to be controlled through an interaction between Galactic

Cosmic Rays (GCRs), which are theorized to ionize the atmosphere and increase cloud formation. High (low) GCR flux may promote cloudiness (clear skies) and higher (lower) albedo at the same time that Total Solar Irradiance (TSI) is lowest (highest) in the solar cycle which in combination creates cooler (warmer) ocean temperature anomalies. These anomalies have been shown to affect atmospheric flow patterns and ultimately precipitation over the Midwestern United States. A study has identified a relation between geomagnetic index aa (GI-AA), and streamflow in the Mississippi River Basin for the period 1878-2004. The GI-AA was used as a proxy for GCRs. There appears to be a solar “fingerprint” that can be seen in hydroclimatic time series in other regions of the world, with each series having a unique lag time between the solar signal and the hydroclimatic response. A progression of increasing lag times can be spatially linked to the ocean conveyor belt, which could transport the solar signal over a time span of several decades. The lag times for any one region vary slightly and may be linked to the fluctuations in the velocity of the ocean conveyor belt.

A graph is attached to the abstract (as seen above):

http://www.leif.org/research/MissGeomagGraphBW.jpg

The poster seems to report on earlier work presented here:

http://ks.water.usgs.gov/waterdata/climate/

Where the same figure appears.

Now, what is wrong about this graph [and the conclusion, of course] ?

I’ll let you all find out what.

It is an example of three things:

  1. The desperate need for establishing a Sun-Climate [or is it weather, when on a decadal basis?] causing this kind of sloppy work (the graph contradicts the mechanism given for it)
  2. The lack of internal quality control by USGS
  3. The lack of quality control by the conveners of the AGU session.

UPDATE:

Thanks to all the readers who so generously [some gleefully] have pointed out my misinterpretation of the figure. This, of course, makes my initial assessment of the quality control moot and void, with an apology to those involved. Perhaps this shows how important a graph can be [cf. the impact of the Hockey Stick] and how important is clear labeling of what is shown.

UPDATE2:

click to enlarge

Since GCRs follow the the sunspot numbers and not the aa-index, the proper parameter to compare with would be the sunspot number. This also allows use of the streamflow data back to the beginning of the series in 1861. The following Figure shows the correlation with this parameter, providing a prediction of the flow to beyond 2040, should the flow indeed be correlated with the sunspot number 34 years earlier.

Advertisements

215 thoughts on “Spot the science error

  1. Easy. Except for the 1930 and 1998 peaks (in which no directional cause may be inferred) it would appear (assuming a causal relationship) that Mississippi flow forces Geomagnetic flux with a 1-3 year lag.
    That’s why they call it “The Mighty Mississippi”, I guess.

  2. Something seems to be, in the vernacular, bassackwards.
    Why would a higher flux result in greater flow? Wouldn’t that result in fewer cosmic rays, clearer skies, and potentially less flow?

  3. Looking at my guest comment, it struck me that an interpretation of the graph could be that the streamflow could be controlled by solar activity three solar cycles earlier. This would make my complaints moot, but that conclusion was so far from my mind that it did not register as a possible one. What say all of you?

    • Leif, I don’t think the timescales for water transport work:

      The term thermohaline circulation (THC) refers to the part of the large-scale ocean circulation that is driven by global density gradients created by surface heat and freshwater fluxes. The adjective thermohaline derives from thermo- referring to temperature and -haline referring to salt content, factors which together determine the density of sea water. Wind-driven surface currents (such as the Gulf Stream) head polewards from the equatorial Atlantic Ocean, cooling all the while and eventually sinking at high latitudes (forming North Atlantic Deep Water). This dense water then flows into the ocean basins. While the bulk of it upwells in the Southern Ocean, the oldest waters (with a transit time of around 1600 years) upwell in the North Pacific (Primeau, 2005). Extensive mixing therefore takes place between the ocean basins, reducing differences between them and making the Earth’s ocean a global system. On their journey, the water masses transport both energy (in the form of heat) and matter (solids, dissolved substances and gases) around the globe. As such, the state of the circulation has a large impact on the climate of the Earth.

      1600 years? Even if it were a factor of ten shorter for water trasnport it still wouldn’t work. Besides the other issues, I just don’t see how they can suggest a ocean heat transport>evaporation>precipitation mechanism at the timescale they propose (3x solar cycle) – Anthony

  4. Well, Leif, I’ve heard it said that the Mississippi is an old man, thus he has a LOONNGG memory. You never know …
    ‘course, one might ask if the paper itself posits a 33-year time lag. I’d bet against it.

  5. While the numbers for Geomagnetic flux for 2011, and perhaps 2012, appear to be in already, it appears that the Mississippi stopped flowing somewhere around 2004 or 2005…

  6. I’d say it’s not connected because the Mississippi river flow is controlled by a continuous set of locks and dams along its entire length, not so much by nature.

  7. “Leif Svalgaard (21:33:34) :
    Looking at my guest comment, it struck me that an interpretation of the graph could be that the streamflow could be controlled by solar activity three solar cycles earlier. This would make my complaints moot, but that conclusion was so far from my mind that it did not register as a possible one. What say all of you?”
    I think this was what was intended. Following the second link you provided, I found the following quote below the graph:
    “Comparison of the 36-month moving average of the Mississippi River streamflow and the Geomagnetic Index AA. Streamflow has been lagged 34 years after the Geomagnetic data.”

  8. #1 – the Mississippi River flow shows little overall trend
    #2 – the aa index mirrors, falls below and rises above in agreement and contrarian to the river flow. There is no consistency to the relationship, linear or non-linear.
    What it says is that climactic zones of the Mississippi River basin are too varied to utilize flow rates as an indicators of continental climate.
    There is no way to distinguish competing zones having disparate precipitation from a truly basin-wide average year. GCR levels would be too far ahead in the chain of command even if they were unquestionably first cause.

  9. Kurt (21:45:09) :
    >i>“Comparison of the 36-month moving average of the Mississippi River streamflow and the Geomagnetic Index AA. Streamflow has been lagged 34 years after the Geomagnetic data.”
    The graph makes it look like the aa is lagging. OK, so I was wrong [given that interpretation – goes to show how careful a graph should be made in order not to make a wrong conclusion easily]. If indeed a 34-yr lag was meant, then the last 34 years of aa would provide a prediction of the streamflow for the next 34 years. THAT would have been an interesting plot.

  10. If you have several thousand years worth of data, like the Nile or the Yangtzee Rivers, and a good proxy for aa, you might be able to find some sort of repeatable sequences. Failing that, compare as many varying river basins for the 1860-present timeframe and see what emerges.

  11. Mark Twain commented on the Mississippi flow years ago. Here is his quote
    “In the space of one hundred and seventy-six years the Lower Mississippi has shortened itself two hundred and forty-two miles. That is an average of a trifle over one mile and a third per year. Therefore, any calm person, who is not blind or idiotic, can see that in the Old Oolitic Silurian Period, just a million years ago next November, the Lower Mississippi River was upwards of one million three hundred thousand miles long, and stuck out over the Gulf of Mexico like a fishing-rod. And by the same token any person can see that seven hundred and forty-two years from now the Lower Mississippi will be only a mile and three-quarters long, and Cairo and New Orleans will have joined their streets together, and be plodding comfortably along under a single mayor and a mutual board of aldermen. There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact.”

  12. I was wrong in calling the work ‘sloppy’. Tricky to interpret, perhaps, and a bit far out: the 3 cycle delay is hard to swallow. Perhaps, that is what makes the paper a breakthrough, of sorts…

  13. I showed it to the cat. As far as I can tell he is laughing.
    Of course I cannot be sure of that since I have not done a statistical analysis as to whether he is laughing.
    Must get government grant to investigate this further.
    Kindest Regards

  14. While the hypothesized causal relationship sounds plausible, I question a 33 or 34 year time scale. Geomagnetic phenomena affect incoming cosmic rays, which affect atmospheric ionization (instantly — there is no time lag). The ionization affects cloud formation, presumably on a time scale of hours to .. perhaps … weeks (that’s probably too long). Cloud formation affects precipitation on a scale of weeks or days (here I’m completely ignoring the chaotic dynamical system, just looking at the direct path of water from evaporate to precipitate). From the event of precipitation one might imagine perhaps a 2 year residency within the Mississippi river system (not including deep lake water that may be considered, for purposes here, fixed and outside consideration). There is very little multi-year snow/ice in the Mississipi basin, so there is no further storage beyond the system of locks and dams mentioned by Mike McMillan above.
    At the outside I can imagine sources for up to 3 years of lag between geomagnetic flux changes and affect on Mississippi flow.
    I think the hypothesis remains plausible, but that Mississipi flow is too problematic an indicator. Is there not a record of worldwide precipitation that can be obtained or reliably proxied? That would be the right place to look to support or falsify the hypothesis. Too many things can go wrong in a geographically localized proxy such as this.

  15. This, like much here, is way over my head. But here’s a nit I can pick, in Leif’s intro: “at the Fall Meeting of …”
    The seasons aren’t capitalized.

  16. I agree with Mark Twain.
    Measured at St Louis, you’re just downstream of where the Missouri and Illinois rivers join in, and lest we forget the #%@& up in Chicago reversed the Chicago river in 1900 so their sewage dumps into the Mississippi, rather than polluting their sacred Lake Michigan. Did I leave anything out?
    The connection with the oceanic conveyor is laughable. Increasing ‘lag’ times just mean there’s no correlation, and therefore no causation. That doesn’t mean I intend to take off my tinfoil hat, though.

  17. The attempt to correlate a single factor to another in a complex system where any number of factors might dominate at different times?

  18. The graph seems to roughly correlate but it AIN’T LAGGED! OOOPs it kind of undermines their thesis of a multi-decadal delay.

  19. I’ve been working on these interannual-timescale aa index phase-relations with precipitation and other climate indices on-the-side. I am 100% certain there is something to them [that is possibly going to look dead-simple once someone works out some key detail(s)]. I can devise dozens of aa-based variables that show clearly-nonrandom phase-relations and I can get the eras of phase-match to shift around (by working with different timescale-contrasts), but I’ve put the investigations aside for awhile because:
    a) it is complex (…so I have to wait for revelation — complex discovery is not easily forced onto a schedule).
    b) the research community is not terribly supportive of pioneering that does not [yet] have all the phase-reversals worked-out.
    c) funding & opportunity direct focus.
    Btw: I’m not convinced that the signals of interest are (entirely) of solar origin. (See the works of Yu.V. Barkin and run comparisons with SOI.) Also, I don’t assume this is about GCRs.
    Very interesting. I plan to look into this again in future (hopefully following a revelation).

  20. Roger Knights (22:18:50) :
    This, like much here, is way over my head. But here’s a nit I can pick, in Leif’s intro: “at the Fall Meeting of …”
    The seasons aren’t capitalized.

    Nit pickers are my specialty 🙂
    From AGU’s website:
    The Fall Meeting is expected to draw a crowd of over 16,000 geophysicists from around the world. …

  21. Could this become a peer-reviewed paper and an additional source for their funding? I’m dying of curiosity.

  22. Steve Schaper (22:41:52) :
    I do believe that this is a case for Dr. Boli.
    1st question in Dr. Boli’s IQ test:
    1. What is the next number in this series?
    827, 827, 827, 827, 827, ___

  23. Mike McMillan (22:22:48) “Increasing ‘lag’ times just mean there’s no correlation, and therefore no causation.”
    Acoustic info can manifest itself in interesting ways in time-integrated cross-correlation best-lag matrices. It’s not meaningless, but rather challenging to decipher (particularly when there are nearly-overlapping harmonic scales).

  24. “There appears to be a solar “fingerprint” that can be seen in hydroclimatic time series in other regions of the world, with each series having a unique lag time between the solar signal and the hydroclimatic response. A progression of increasing lag times can be spatially linked to the ocean conveyor belt, which could transport the solar signal over a time span of several decades.”
    If I understand this correctly, the abstract is claiming that they have matched streamflow to solar activity in many regions besides the Mississipi basin, albeit by using different lag times. I’d ordinarily not be impressed by this because if you get to pick the lag time for each region, it could just be coincidence that variations in streamflow have ups and downs at some point that match an arbitrary patten. But wouldn’t you expect the lag time to vary randomly if there is no correlation here? The abstract also seems to be claiming a pattern in the lag times, i.e. the lag time runs in one direction (increasing) along the ocean conveyor belt. That would seem a little too much to be coincidence.

  25. Leif wrote:
    Nit pickers are my specialty 🙂
    From AGU’s website:
    The Fall Meeting is expected to draw a crowd of over 16,000 geophysicists from around the world. …

    You beat me to the punch! I just realized that the phrase “Fall Meeting” was a title (indicated by the capitalization of “Meeting”), not just a descriptive phrase, and was rushing here to post a correction first.
    (Well, anyway, at least readers here now know, if they hadn’t before, that in plain text the seasons aren’t capitalized.)

  26. It’s not a one-on-one correlation at all. It’s a mixture of a whole bunch of signals and events against a single one. It would be much more interesting to see separate signals compared against aa to see what goes along with it and what doesn’t.
    What’s the significance of 34 years? A PDO tenure perhaps?

  27. A number of commenters see the decrease in river flow with increase in cosmic rays and presumably cloud formation (i.e. stemming from lower geomagnetic flux) as problematic. Why not an anti correlation with cosmic rays? Might cosmic rays primarily cause extra moisture depletion before air masses move from ocean regions to the interior of a continent? It seems that cosmic rays would best influence precipitation nearest the source of evaporation where the air is closest to saturation with water vapor. I would wager C. A. Perry has thought of this.
    As an aside, I take issue with Leif’s charge of poor quality control by conference organizers by allowing such a poster. It is not uncommon to find factually incorrect ideas displayed in poster sessions. Posters do not receive the same peer review as journal articles. Their public display does not imply sanctioning by the conference committee. Posters can be admitted for the sake of generating interesting discussion. A presenter may use the forum as a way of testing the waters on a novel idea. Look what the poster is doing on this thread.

  28. I don’t know, but my first impression is that this is strange.
    From this link, that was given above, we read in the graph legend that “Streamflow has been lagged 34 years after the Geomagnetic data.” That and some of the other numbers given there indicate there may have been a lot of not-so-transparent parameter adjustment going on?
    Also, shouldn’t it say “correlation” and not “connection?” in the graph title we see ‘above the fold’?
    I do believe that cosmic rays have a lot to do with earth’s climate, but this looks like it’s pushing the envelop, to say the least. If this were the only evidence, I wouldn’t rush to buy.

  29. Note in response to comments about lags:
    Say there is some underlying shared-(quasi)-periodicity, but there is one major event that seriously affects one-but-not-the-other of a pair of somewhat-related variables — this will introduce an (apparent) lag. The investigator has to discover the conditioning variable (the major event) to realize that the lag does not indicate a lack of relationship, but rather a conditional relationship.
    I don’t have time for elaborate explanations, but here’s an image that was, for some time, a mystery:
    http://www.sfu.ca/~plv/CCLR1LPPT1.png
    It was part of a pattern of best-lags for precipitation at a site I was studying:
    http://www.sfu.ca/~plv/BestLagMatrix.PNG
    A clean relationship looks more like this:
    http://www.sfu.ca/~plv/CCPxXTR.png
    This^ clean image led me to discover the importance of this event:
    http://www.sfu.ca/~plv/PolarMotionPeriodMorlet2piPower.PNG
    I’m sure people around here have heard of the Dirty 30s. An event like that can “mess up” precipitation-linear-correlations with other variables.
    Having worked on this sort of stuff for 2 years now (after changing fields once again), I’m not surprised to see how “stuck” climate science appears to be (in general). I’ve taught stats – and there weren’t many of my students that would figure this stuff out – and I don’t mean because of lack of intelligence, training, &/or resilience to plow through a series of obstacles – more importantly it’s a matter of having the intellectual, social, & economic freedom to pursue things that require prolonged, undivided attention.

  30. Marginally on topic, a question to Dr. Svalgaard:
    The sunspot number is again down to zero these days.
    What is its prospect for coming weeks or months?

  31. @Leif Svalgaard
    1) Shouldn’t your error be corrected in your guest-post?
    2) They say: “There appears to be a solar “fingerprint” that can be seen in hydroclimatic time series in other regions of the world, with each series having a unique lag time between the solar signal and the hydroclimatic response.”
    If there is a similar correlation for many rivers, and the patterns of the variations of the streamflows are rather complex, is this not a proof for a link between the geomagnetic index aa (GI-AA), and the streamflow?
    I have the same question for the exact correlation between the MWP, little ice age minima etc.. and the glacier advances in Europe and in the tropical Andes.
    http://ff.org/centers/csspp/library/co2weekly/20060725/20060725_08.html
    “Another important aspect of the Polissar et al. study is that it clearly implicates solar variability as the cause of the climatic variations they observed. They note, for example, that “four glacial advances occurred between AD 1250 and 1810, coincident with solar-activity minima,” and they state that the data they present “suggest that solar activity has exerted a strong influence on century-scale tropical climate variability during the late Holocene, modulating both precipitation and temperature” and demonstrating the “considerable sensitivity of tropical climate to small changes in radiative forcing from solar irradiance variability.””
    or here http://www.co2science.org/subject/s/summaries/solartempsamer.php
    My question is:
    If all events occur at the same time, and if these changes have very complex pattern (LIA-minima) and last over a long time period (1000 ad until now), isn’t this a 100% proof that the solar fluxes did influence the climate?

  32. OT, I suppose, but the geomagnetic data appears to correlate just as well with war. There is a deep dip around the start of WWI, WWII, the Korean war and the Vietnam war. Surely a coincidence.

  33. Paul Vaughan (23:29:09) :
    Just one last thing. Am I supposed to believe that from solar activity in (or for some years around) 1958, the great Mississippi flood of 1993 could have been predicted?
    http://www.noaanews.noaa.gov/stories/s1125.htm
    Sorry, but I’m not buying it. It’s way too Rube Goldberg.
    Also, I’m not sure if this is what L.S is referring to, but increased solar activity results in a decrease in cosmic ray flux, which should result in less cloud cover, and hence less rain, not more. As this link clearly shows, there is an inverse correlation between solar flux and neutron detection.
    http://www.leif.org/research/CosmicRayFlux.png

  34. A terrible lack of quality control here, worthy of RealClimate. For goodness sake, why couldn’t Svalgaard find out what Charles Perry is really saying before wading in with comments on a graph he didn’t understand, and inviting all these other snide comments.
    A place to start reading Perry’s story is http://ks.water.usgs.gov/waterdata/climate/

  35. Early on the troughs of the geomagnetic index are lower than the troughs of river flow. More recently they are higher. So I’d posit no discernable connection.

  36. ann riley (23:56:15) :
    “OT, I suppose, but the geomagnetic data appears to correlate just as well with war. There is a deep dip around the start of WWI, WWII, the Korean war and the Vietnam war. Surely a coincidence.”
    Perhaps not entirely.
    “War has historic links to global climate change”
    http://www.newscientist.com/article/dn12936

  37. Leif Svalgaard (21:33:34) : “…This would make my complaints moot, but that conclusion was so far from my mind that it did not register as a possible one. What say all of you?”
    Peak matching is always problematic, in my opinion. Over a short term, someone could probably correlate Mississippi flow with, say, the positions of Jupiter and Saturn…
    Just kidding, Leif. Nobody would do that.

  38. “The Word “connection” on the graph should be correlation.”
    No, no. They have connected two points of the curves and show that they aren’t correlated.
    BTW, thanks for the reference to Dr. Boli.

  39. Paul Vaughan (22:52:09) :
    Acoustic info can manifest itself in interesting ways in time-integrated cross-correlation best-lag matrices. It’s not meaningless, but rather challenging to decipher (particularly when there are nearly-overlapping harmonic scales).

    Acoustic from cosmic rays? You must mean that ping I hear when they bounce off my tinfoil hat.
    .
    I compared the Illinois corn crop failure years with the shifted river curve, and I get some bad years lining up, and other bad years totally opposite. Nothing that would be worth pasting together a chart. The good years, closer to 9 out of 10, seem impervious to the river or the cosmic rays. Yields sure do track CO2 nicely, though.
    I think the GCR/river link is pretty thin, thinner than the Mississippi water, anyway, which is too thin to plow.

  40. In broad terms I see no problem with a periodicity in river flows worldwide from the effect of the 30 year or so phase shifts of the oceans and in particular PDO.
    Those phase shifts move all the air circulation systems latitudinally and change regional rainfall distributions globally.
    The next step, in linking that with solar variability is the real issue. I am torn between Leif’s certainty that the solar variations are too small and the historical correlations (not perfect admittedly) which I feel I have to give some weight to (but no idea how much).
    Given that Leif himself has spotted the three cycle lag and acknowledged it’s potential significance and given that the narrative specifically refers to it I am prepared to attribute some weight to this study and others that I have seen, including a similar finding relating to South African waterflows.
    I think tallbloke here has pointed out that the oceanic phase shifts seem to occur at every third solar cycle.
    The isue is worthy of further investigation.

  41. One should do a 13-years running average analysis of the aa index and river flow data to eliminate the sunspot cycle oscillation and see how these results correlate with the corresponding analyses of global temperature and length-of-day (see Fig. 2.2 of the fao-report, below).
    World climate is a regular quasi-periodic phenomenon (see Fig. 2.1 of http://www.fao.org/DOCREP/005/Y2787E/y2787e03.htm ) that is driven by solar activity with a period of 75 – 85 years (Gleissberg cycle). Because of this regularity, it can be stated with absolute certainty that the mean Earth temperature will continue to decrease until 2040.
    1. There exists an extremely close correlation between the changes in the mean global temperature and the small changes in the rotational velocity of the Earth – two physically unrelated geophysical quantities – (see Fig. 2.2 of http://www.fao.org/DOCREP/005/Y2787E/y2787e03.htm), which has been ignored by the mainstream climatologists, and leaves little room for a human influence on climate. Note that temperature lags rotation by 6 years. This close correlation results from the action of an hitherto unknown form of gravitational waves, galactic vacuum density waves, on the Sun and on the Earth (see http://www.icecap.us/images/uploads/Lobert_on_CO2.pdf ).
    2. The orbital periods of all Solar System planets are very close to integer multiples and integer fractions of the periods of the Hale (22.14 years) and the Gleissberg (84 years) solar cycles. (See the posts of September 26, 2009 and October 6, 2009 in http://www.pakteahouse.wordpress.com/2008/09/15/a-new-book-elucidates-the-life-and-work-of-Dr.-Abdus-Salam ). This provides further evidence for the existence of super-Einsteinian gravitational waves and of their action on all celestial bodies of the Solar System.
    Progress in climatological science can only be achieved if the above physical facts are looked into in full depth.

  42. hmmmm,
    Should be interesting to see how “the team” responds, after all the 30iesh years lag is almost the prime-über-golden number for climate trends that Gavin/Rahmstorf/ ++ occationally subscibes to 🙂
    (Perhaps because) I am not an anglophone, I would think there could be another reason for employing the term connection rather than correlation. It carries the core of teleconnection, a term quite often used in this particular context that may include heat transportation distantly over time.
    Cassanders
    In Cod we trust

  43. Further to my previous post (since North America hasn’t woken up yet) let me paraphrase Svalgaard’s initial comment on Perry’s work to say that this is an example of Svalgaard’s own desperate need to refute Sun-Climate connections.
    On WUWT in September, Svalgaard likened Henrik Svensmark to Al Gore. It amazes me that on a skeptical website this man is treated as some kind of guru, when he keeps dismissing the main scientific challenge to the CO2 story.

  44. I think the correlation is interesting. The major mistake on the graph is a lack of clear annotation regarding the time lag applied to the Geomagnetic Index.
    It also seems that the magnetic lows should correlate to the high flow volumes if the correlation is indicative of a GCR-cloud connection.
    If the time-lag is dynamic, the concept will be kind of difficult to depict on a simple graph.

  45. At least the Perry’s proposed relationship can be easily tested during the next years – if the Ol’ Man River will indeed follow the aa values 34 years ago. Plain and simple.
    We don’t have to wait till the end of the century or insert corrections for this and that, like with global climate models.

  46. Nigel Calder (04:42:11) :
    I can only say that I agree with you. It seems clear that the correlation between temperature and CO2 during the last few centuries is almost zero. All you can say is that they both went up in the previous century, but then lots of things went up. In contrast, there is an overwhelming amount of evidence showing links between solar activity and all kinds of historical measurements such as temperature, drought, river flows and even corn prices.
    Of course you have to be very careful about the manipulation of data, but as dramatic correlations with solar activity pop up in so many different areas of research, the case for a solar impact on climate is almost unanswerable.
    I assume you are the ex-editor of New Scientist. Just as I’m sure you are, I’m very sorry to see the depths it has plumbed. I used to buy it every week, but no longer. I just hope that in my lifetime this absurd delusion is swept away and that the honesty of science is restored. And that I will be able to return to buying New Scientist every week.
    Chris

  47. This debate reflects well on WUWT – instead of just piling on, some readers took the trouble to check Leif’s finding, faulted it, and he acknowledged his error.
    It would be good for Leif to complete the process by updating his post.

  48. May I assume that the river flow as charted has not been influenced by human development since 1900? Or has the data been hansenized by some method to remove any human footprint?

  49. Nigel Calder (04:42:11) :
    On WUWT in September, Svalgaard likened Henrik Svensmark to Al Gore. It amazes me that on a skeptical website this man is treated as some kind of guru, when he keeps dismissing the main scientific challenge to the CO2 story.
    At least some of us are switched on.
    Dr. Scafetta is switched on also, and recently showed the correlation of the PDO and Solar distance from SSB, I wouldnt discount anything he has to do with so quickly, especially if you have little knowledge in this area.
    The upcoming AGU conference should be interesting, Dr. Scafetta invited me to speak but unfortunately I am unable to attend…maybe 1 more year might make a big difference.

  50. I have seen a couple of papers of Charles Perry before on this theme:
    ‘Midwestern streamflow, precipitation and atmospheric vorticity influenced by Pacific sea surface temperatures and TSI Int. Journal of Climate 26 (2) 207-218 (2005)
    ‘Evidence for a physical linkage between galactic cosmic rays and regional climate time series’ Advances in Space Research 40, 353-364 (2007)
    Perry & Hsu (2000) ‘Geophysical, archeological and historical evidence support a solar output model fro climate change’ Proc US National Academy of Sciences 10.1073/pnas.236423297
    So, Leif before you jump the gun on peer-review, check out these papers in respectable climate journals. I have been impressed by Charles Perry’s willingness to take on a difficult issue that crosses many disciplines and at a time when the generally established view is set against solar drivers of climate change. What astonishes me is that the science community fails to cite work like this and gets away with it. All three citations are in my book ‘Chill’, together with a discussion of Perry’s thinking – of inrterest to me because it fits with some of my observations and conclusions that:
    a) there are pulses of short-wave energy to the earth’s surface picked up by satellite surveys and these correlate to the highs of the solar cycle – such pulses are larger than would be the case from TSI variability and obviously involve less cloud cover (as Svensmark readily found for cycle 22 – but with much less clarity for cycle 23) – it is for this reason that Perry cites the GCR connection – from the peer reviewed solar science literature;
    b) whatever the ultimate cause (and I am not entirely convinced of the GCR factor), the pulses are real. These will create pulses of warm water in the ocean surface (70% of the planet) – and this heat is stored (unlike in the atmosphere or on land) and the signal – according to Warren White’s work at Scripps, e.g.:
    ‘Sources of global warming of the upper ocean on decadal period scales’ J.of Geophysical Research 108 (c8) doi:10.1029/2002JC001396
    and
    ‘Responses of global upper ocean temperatures to changing solar irradiance’ J Geophysical Research 102, 3255-3266 (1997)
    c) as even the IPCC admit – 80% of the ‘global warming’ signal is held in the upper ocean, and as model analysis has shown
    (Compo & Sardeshmukh (2008) ‘Oceanic influences on recent continental warming’ (Climate Diagnostics Centre, U of Colorado)
    this signal is transferred to land.
    d) so then we come top Perry’s hypothesis which relates not so much to the ultimate cause – or the mechanism – as a geolgist/hydrologist he is simply referrring to the peer-reviewed solar-terrestrial science literature, but to the nature of the time-lagged correlation – which does indeed show a similar time period to the PDO.
    The whole purpose of an hypothesis is to stimulate new thinking and new ways of gathering data in the development of theory – and often this requires stepping out, being willing to make mistakes, and being opne to feedback – that is the very lifeblood of science. The reason so few go down this road is well illustrated by the rush to judgement and comment bordering on ridicule – which I would have hoped not to see on WUWT – its the kind of response one gets from RealClimate if you try to outline anything contrary to their model-based faith in warming (as I recently did!).
    Consider therefore that we do not yet know what drives the 30-40 year oscilllation in the northern Pacific, or the amplitude of ENSO to which it is connected:
    (Biondi et al, 2001 ‘North Pacific Decadal Climate Variability since AD 1661’ J of Climate 14 (1) 5-10
    My suggestion (rehearsed in a chapter devoted to ocean cycles in my book) that the PDO was important in global temperature patterns, I was ridiculed by Gavin S. Yet the peer-reviewed literature leads to this conclusion.
    Perry’s work should have long ago stimulated research into time-lags – yet it has been systematically ignored. I could readily see teleconnections between the phase of the PDO and Arctic Ocean temperatures. Those temperature patterns follow a pattern in sea level pressure (time lag about 5 years according to Polyarkov at IARC).
    What happens in the Arctic and north Pacific connects to what happens in the mid-West.
    The solar cycle has long been recognised as a signal in sea surface temperatures
    as also in the Asian Monsoon. But if you are predisposed to think there can be no such link, then you don’t look for the literature – it is abundant.
    I am not so convinced the shorter term pattern of pulses is entrained in the thermo-haline circulation because that turnover is on centennial time scales (though longer frequency oscillations may be at work – such as the MWP/LIA type variability). Rather, i would look to the nature of surface currents – which are not as simple as most maps indicate, but rather zigzag and take much longer to move warm water from the equatorial regions to the massive accumulation in gyres of the Pacific and Atlantic’s northern waters. i detail this in my book – drawing from analyses of the spatial distribution of the upper ocean heat content – perhaps it takes 30 years to accumulate and 30 years to release and the pulses feed into this pattern – the release phase of the PDO (dumping heat on Alaska) ended late in 2006 and coincided with the current solar minimum. These phases of accumulation and release also link to the shifting jetstream (which Perry also looks at). The Atlantic is now on the turn after 25 years of build up in the northern gyre – the jetstream also shifting, with huge dumps of rainfall over Britain and Europe in the last three years leadign to serious flooding – my guess is that once that heat reservoir is exhausted, the accumulation phase will begin again and northern Europe will really freeze up.
    Although GCR may have some impact, I am inclined more to finger the UV signal and its effect on the Polar Vortex – also very under-researched.

  51. tokyoboy (23:43:58) :
    The sunspot number is again down to zero these days.
    What is its prospect for coming weeks or months?

    Two weeks of low, then higher again, judging from the 27-day recurrence.
    Edouard (23:53:26) :
    1) Shouldn’t your error be corrected in your guest-post?
    Yes, but it is perhaps good to let it stand to show that to err is human [helped along by an unclear graph].
    Nigel Calder (04:42:11) :
    when he keeps dismissing the main scientific challenge to the CO2 story.
    If it were so, it would be nice that we had a solid challenge. But a ‘challenge’ should not be judged on how well it debunks CO2 because we want CO2 to be debunked. A three-cycle delay between GCRs and mesoweather does not seem very credible. If there were a delay through the ocean, I would expect that the oceans would also dampen out the 11-year cycle.
    John A (03:28:26) :
    Why does the GEOmagnetic flux have anything to do with the SOLAR magnetic flux?
    Small variations [less than 1%] in the geomagnetic field have an external cause which is the solar magnetic field brought to us by the solar wind.

  52. Dave Middleton (04:42:17) :
    I think the correlation is interesting. The major mistake on the graph is a lack of clear annotation regarding the time lag applied to the Geomagnetic Index.
    As I have already pointed out, the correlation [with the 34-yr lag] provides a prediction of the flow 34 years into the future, as the aa-index for the past 34 years is known already. Since the strength of a theory must be judged by its predictions rather than by its fit with past data, there is the prospect of validation in the coming years.
    I could imagine the same poster with a different title:
    “A Prediction of the Flow of the Mississippi for the Next 34 Years” that would certainly generate a lot more interest and be less susceptible to the mistake I [admittedly] made, and perhaps also more skepticism as this is an extraordinary claim [imagine that the government should base maintenance of the levees upon it].

  53. Mike, (02:41:05)
    Sounds like the Roda (Rodah / Rowda / Rawda etc) nilometer dataset – the one Hurst originally used for his analysis of nile river flow for the design of the Aswan dam back in 1952.
    Just like the Nile, the Mississippi is likely to be subject to the Hurst phenomenon, which will result in low-frequency components having greater amplitude (due to constant power/octave). This, in turn, makes “accidental” correlation highly likely, especially if you give yourself a tuning parameter or two (e.g. lag).
    I suspect that once the Hurst phenomenon is accounted for the supporting evidence for these correlations being significant would disappear (in the same way that evidence for CO2 based AGW disappears when you correctly account for it).
    I did try explaining this to Leif over at CA some time ago but the spam filter stopped me from posting part way through explaining it 🙁
    Nigel Calder – respectfully, I disagree. We should not be like the Lysenkoists at RealClimate and refuse to listen to alternative viewpoints. We should, of course, provide scientific criticism and response where appropriate. And also, if our best hope of rejecting AGW is to replace one blinkered “dominant climate driver” (CO2) with another (solar) then I really do pity the future of science.

  54. My initial thought whenever I see these kind of correlations is:
    1. There are lots of physical processes in nature that exhibit cyclic behaviour.
    2. A subset of 1. will have cycle lengths of approx 9-12 years (or 22 years or whatever).
    If you add in some wiggle room from lags (and allow the lags to vary in length) then you’re always going to find some correlations if you look hard enough. I’ll keep an open mind though.

  55. Dr. Gerhard Loebert (02:31:10) :
    1. There exists an extremely close correlation between the changes in the mean global temperature and the small changes in the rotational velocity of the Earth – two physically unrelated geophysical quantities –
    This really SETTLES SCIENCE on this issue. FAO study fits and works OK predicting sea temperatures and related fish catches.(for example, as happends now, the emergence of the Humboldt´s current along the west coast of SA and consequently the increase in anchovy catches) .What is lacking now is the overall mechanism. I think Dr.Svalgaard can explain it in detail.

  56. Peter Taylor (05:51:23) :
    So, Leif before you jump the gun on peer-review, check out these papers in respectable climate journals.
    I have admittedly my mistake. However, there are still some issues with the poster apart from the unclear labeling. If the GCRs are responsible, then a more direct indicator of them would be the sunspot number rather than the aa-index and the more direct parameter should have been used. This would also make use of the three cycles before 1868 [not 1878 as the abstract say] as the sunspot number in 1834 would predict the flow in 1868 [first year shown on the graph]. Perhaps they were not used because the correlation was not so good then [just by eyeballing]. If someone has the patience, it might be interesting to read off the yearly values of the flow and correlate against the sunspot number.

  57. A quick glance look appears that the solar aa index lags the river flow, thus one might conclude the river controls the sun …. hmmmm , that doesnt sound right.
    For this sort of analysis, a comparison of spectral content – both power spectra & phase spectra – would be interesting to see.

  58. Mike McMillan (22:22:48) :
    I agree with Mark Twain.
    Measured at St Louis, you’re just downstream of where the Missouri and Illinois rivers join in, and lest we forget the #%@& up in Chicago reversed the Chicago river in 1900 so their sewage dumps into the Mississippi, rather than polluting their sacred Lake Michigan. Did I leave anything out?
    Only spent 5 min. with this, but the above was “my first thought,” on the graph. Don’t remember when that sewage canal from Chicago was linked up with the Mississip but should be considered. It’s no longer done this way which might also be a contributing factor.

  59. Geoff Sharp (06:46:21) :
    I hope your update includes an apology to the conveners of the AGU session.
    And what is that to you?
    This is what I submitted:
    Perhaps it is time to update my post:
    “Thanks to all the readers who so generously [some gleefully] have pointed out my misinterpretation of the figure. This, of course, makes my initial assessment of the quality control moot and void, with an apology to those involved. Perhaps this shows how important a graph can be [cf. the impact of the Hockey Stick] and how important is clear labeling of what is shown.”

  60. Leif Svalgaard (06:05:00) :
    Dave Middleton (04:42:17) :
    I think the correlation is interesting. The major mistake on the graph is a lack of clear annotation regarding the time lag applied to the Geomagnetic Index.

    As I have already pointed out, the correlation [with the 34-yr lag] provides a prediction of the flow 34 years into the future, as the aa-index for the past 34 years is known already. Since the strength of a theory must be judged by its predictions rather than by its fit with past data, there is the prospect of validation in the coming years.
    I could imagine the same poster with a different title:
    “A Prediction of the Flow of the Mississippi for the Next 34 Years” that would certainly generate a lot more interest and be less susceptible to the mistake I [admittedly] made, and perhaps also more skepticism as this is an extraordinary claim [imagine that the government should base maintenance of the levees upon it].

    I don’t think you necessarily made a mistake. The paper’s abstract was not terribly clear and the attached graph wasn’t annotated very well.
    One of the problems we geologists often encounter is that we never run out of “past data” to correlate. Personally, I wouldn’t try to “fine tune” levee maintenance based on any geological predictions… I’d use Murphy’s Law.

  61. Leif Svalgaard (06:51:11) :
    And what is that to you?
    This is not a time for you to be bullish, I was offended by your remarks and regard Dr. Scafetta as a hard working scholar with honorable credentials and sound integrity. Try and be humble for once in your life.

  62. Geoff Sharp (07:04:19) :
    This is not a time for you to be bullish, I was offended by your remarks
    It is offending that you suggested the possibility that I would not apologize.
    REPLY: And it is offending to me that I might have to moderate yet another war of words between Leif and Geoff. Both of you, stand down. – Anthony

  63. Nigel Calder (04:42:11) :

    Further to my previous post (since North America hasn’t woken up yet) let me paraphrase Svalgaard’s initial comment on Perry’s work to say that this is an example of Svalgaard’s own desperate need to refute Sun-Climate connections.
    On WUWT in September, Svalgaard likened Henrik Svensmark to Al Gore. It amazes me that on a skeptical website this man is treated as some kind of guru, when he keeps dismissing the main scientific challenge to the CO2 story.

    I profoundly disagree. What is his latest paper really all about? From what I’ve seen Leif sticks to objective, sound science and is equally sceptical of all hypotheses (including his own) as any scientist should be. If we had more like him and the politics, celebrity and media sensationalism that’s polluted modern science had been kept in check, this WEB site probably wouldn’t even exist!
    We could get into who or what might have prevented that pollution or GG3 versus ‘creative’ thinking but that’s not appropriate in this thread. Science is either science or it isn’t. It’s fine (up to a point) for others to stray into pseudo-science for five minutes but expect any good scientist to play whack-a-mole with it to the nth degree each and every time.
    The “main scientific challenge” is not “challenging the CO2 story” but establishing the truth whether we like it or not. Politics and social engineering is another matter.
    IMHO there is an alarming gulf between the accessibility of real science and the exponential growth of sensationalism and pseudo-science all over the WEB. This is unhealthy; for a casual observer distinguishing one from the other is getting increasingly difficult and ‘science’ needs to do something about it. Sorting out the NSF and NASA media presence would be a good start. I have no idea how you deal with it across science as a whole but Leif’s presence here is refreshing and much appreciated.

  64. When God created the swallow, a migrating bird that winters in Africa, He started to show him how to build his nest. He showed how to make small mud ballswith his tongue and how to gradually build up the nest; but He was interrupted just before reaching the point where He started to show the swallow how the nest should get covered and have a roof. The swallow, flexing its wings and ready to fly off swiftly after juicy flies and mosquitoes said “OK, OK, I know, I know” and flew away.
    That is why swallows’ nests are only half built and they have to be under a roof or an outcropping. The swallow never had the patience to listen to the end of the demonstration.
    Ever since a child I have often found myself in the swallow’s position, which is why my father would repeat the tale to me equally often.
    Glimpsing the TRUTH is so exciting that a scientist has the tendency to say ” I know, I know” and jump to conclusions in order to rush to the next exciting manifestation of nature/fellow-scientists.
    Ah well, it would all be the lovely game it should be, some you win, some you lose, if politicians had not high jacked the subject for their own insidious ends.
    And I am saying that science should be a game,because, contrary to what the public thinks of science, it is extremely creative in the same way that writing a poem or a play is creative, and creativity needs room to dance in.

  65. Leif
    Have you see this interview with Hathaway?
    http://www.earthfiles.com/news.php?ID=1635&category=Science
    “I am writing a paper – it’s on my computer as we speak (laughs) – basically saying that I made a big mistake – myself and Bob Wilson – when we wrote a paper in 2006, suggesting Solar Cycle 24 was going to be a huge cycle based on conditions at that time.”
    “But there also were people back at that time saying otherwise. A group of colleagues led by Leif Svalgaard, Ph.D., were looking at the sun’s polar fields and saying even at that point, the sun’s polar fields were significantly weaker than they had been before and those scientists back then predicted it was going to be a small cycle.”
    IS IT FAIR TO SAY THAT THE SUN IS BOTH PECULIAR AND UNPREDICTABLE?
    “Yeah, I’d buy that! (laughs) Most definitely!”

  66. Leif Svalgaard (07:58:46) :
    fred (06:48:46) :
    If anybody wants to look at the data, it is here.
    http://waterdata.usgs.gov/nwis/nwisman/?site_no=07010000
    Thanks, fred.
    I put the data to use, see UPDATE 2 at the top of the thread.

    The authors did not want to use the sunspot cycle, I suppose because they get a better fit with their “The GI-AA was used as a proxy for GCRs.”
    Considering the LP effect maybe there is sense in their choice?

  67. Just The Facts (08:24:34) :
    Have you see this interview with Hathaway?
    “those scientists back then predicted it was going to be a small cycle.”

    Yes, David is a good scientist, and being wrong [as this thread attests to] is ok when you admit it.
    anna v (08:35:42) :
    The authors did not want to use the sunspot cycle, I suppose because they get a better fit with their “The GI-AA was used as a proxy for GCRs.”
    That is, of course, cherry picking, especially since AA is a poor proxy for GCRs.
    Considering the LP effect maybe there is sense in their choice?
    We have other ways of calibrating the Sunspot number [e.g. http://www.leif.org/research/Napa%20Solar%20Cycle%2024.pdf ] and although there are indeed some corrections needed [20-40%], they are small enough that the basic cycle is not affected.

  68. Molon Labe (22:45:16) :
    Isn’t it remarkable, though, that Mississippi flow rates appear to have an 11 year periodicity?

    Interesting, yes, but not necessarily meaningful. I once noted a coincidence in period between air pressure fluctuations and 29 day lunar cycle. There is nothing to it, though, just coincidence. One needs not only the observation of coincident periods, but also a plausible mechanism by which it occurs, and then testable consequences of the model.

  69. If THC is causing the lag:
    THC transit = 1600years (possible)
    Assume transit distance is 40075km (circumference of earth) (Over estimate)
    then in 34 years THC will have moved 851km
    So TSI will have caused peak temp to trough temp in a distance of 141km
    Assuming no mixing (impossible)
    Then climate over the catchment are of Miss river is being affected by 141km of strangely heated ocean! HHHMMM!
    Why do we not see these paths of hot/cold travelling in the ocean from satellites?
    In my electronics oriented brain you can maintain a signal in a delay line (you pump in signal at one end and it eventually plops out the other only slightly distorted)
    The ocean is not a delay line. It reacts more like a low pass filter. (you pump signal in at one end and the average drops out of the end).
    TSI will warm the ocean above average on the peaks and allow it to cool on the troughs.
    This simple simulation shows the effect
    http://img509.imageshack.us/img509/8022/simplelpf.jpg
    When TSI (green) is high Temperature (VOUT) (blue) rises the instant tsi falls temperature begins to fall. There is no delay!

  70. Now this is a question for weather men, but related to the relevance of the magnetic fields.
    I have noticed here in Greece that if the rainclouds do not have a lot of electricity, i.e. not much thunder and lightening, the precipitation is poor. This can happen for a whole year, it is not that clouds do not come and we have a drought, it is the the clouds that come are not “pregnant” 🙂
    Has there been a correlation found between magnetic fields, geo and sun, and the intensity of thunderstorms and precipitation?

  71. So, who is betting on a midwest drought in 2040-44?
    Should be some doozy flooding between here and there. N’awlins better get those levees upgraded, I think the rest of us would look dimly on paying for them twice…

  72. Leif
    See Perry’s web page:

    This mechanism is outlined in two papers (Perry, 2006) and (Perry, 2007), and is shown schematically below.

    The 34 years comes from the correlation peak shown in Perry 2007 Fig. 6.
    Perry 2007 Fig. 11 shows Perry’s physical interpretations of the lags in ocean circulation.
    For a popular article on Perry’s work see:
    Scientist Uncovers Long-term Weather Patterns Linked to Droughts and Floods

    Seems that the path that the WOWs and COWs take is much longer than originally thought. This path may begin in the shallow ocean around Indonesia, travel west through the Indian Ocean to the southern tip of Africa where it bends back east to pass south of Austrailia. Once in the South Pacific Ocean it is thought to move slowly northward to the area of the North Pacific that affects Midwestern weather.)
    He turned to the worldwide web to see if any other scientist had found a similar lag time. Indeed, there was one. A German oceanographer, studying the correlation between sea surface temperature in the Pacific and the salt water anomalies in the Atlantic Ocean came up with the same 34-year lag.

    This correlation is not to “3”: cycles, but to the impact on clouds which affects climate which is delayed by the ocean circulation lag which influences weather and precipitation over the Mississippi river basin.

  73. Mike Lorrey (09:56:14) :
    So, who is betting on a midwest drought in 2040-44?
    “Should be some doozy flooding between here and there. N’awlins better get those levees upgraded, I think the rest of us would look dimly on paying for them twice…”
    Might be closer to here than there, if the flood warnings in the news are any indication.
    NWS map
    http://www.nws.noaa.gov/
    “The National Weather Service said Monday that water levels at the Meramec and Mississippi rivers in Missouri are above flood stage.”
    http://www.upi.com/Top_News/US/2009/11/02/NWS-warns-of-river-flooding-in-Missouri/UPI-84681257180383/
    “A flood warning is in effect for counties along the Illinois River, according to the National Weather Service.
    The flood stage at Peoria is 18 feet. At 9 a.m. Monday, the river stage was 19.9 feet with minor flooding occurring. The river will continue rising to near 21.1 feet by early Wednesday morning then begin falling.”
    http://www.pjstar.com/homepage/x880799953/Flood-warning-in-effect

  74. As I understand it, the Mississippi basin receives it’s moisture from two distinctly different sources. The upper Miss and the western tributaries get most of their water from the Pacific, while the southern and eastern basin get much of their water from the GOM. Each of those sources are subject to different climatic conditions.
    If this ‘solar/water flow’ correlation is inferring an influence in global rainfall, then all river flows should follow a similar pattern. Otherwise you would really need to filter out the Pacific rainfall patterns from the Gulf rainfall to figure out what might be driving such a correlation.

  75. Well if I read my horoscope (leo) I can convince myself that it really describes me. Funny thing is that I get the same feeling when I read the other 11 as well.
    There is a whole filed of mathematics relating to the generation of data sets that seemt o be more correlated than chance would indicate.
    That’s what I see when I look at those graphs, either the original one, and the 34 month delayed one.
    What would be most interesting would be to see a continuous graph of the crosscorrelation of those two functions as a function of the time displacement (forwards and backwards.)
    I can’t say that the causal link is in any way apparent. Cosmic ray events, after all, would seem to be independent of where the land masses are on the planet. So would you get a similar result for say the nile flow, or the Amazon ?
    I’m sure one could capture a string of random noise especially one with a 1/f noise component, that would equally well correlate with either of these two graphs.

  76. Nigel Calder: On WUWT in September, Svalgaard likened Henrik Svensmark to Al Gore. It amazes me that on a skeptical website this man is treated as some kind of guru, when he keeps dismissing the main scientific challenge to the CO2 story.
    Thanks Nigel. I appreciate this color to the oddly aggresive attack by Lief – it explains the motive. I find it very encouraging that Svensmark theories are so aggressively rejected by the existing scientific church – this is often a sure sign that he is on to something BIG ( I have seen this in many areas of Geophysics).
    Since albedo is for sure a HUGE factor in Global Climate I suspect that Svensmark, even if proved wrong eventually, is at least looking in EXACTLY the right place for a new mechanism to explain things on a Global Scale. In layman’s terms he is hunting elephants while those who look at other factors (winds, ocean currents, CO2 and other trace gases etc. etc.) are clearly barking up a tree and most likely have been bitten by ants (all second order or third order effects).
    A great scientist/engineer will always frame the question in the right way …”What could be BIG enough to have a significant effect on Global Climate?” Firstly one would suspect the source of the energy: the Sun. One might even suspect warmth from the planet interior as primary sources of heat. Secondly one would immediately suspect cloud cover, since, in the absence of large variations in teh Sun’s TSI, it is mind numbingly chlidlike obvious that cloud cover will have a primary first order effect (same level as TSI).
    I find it highly amusing that there is so little research in Svensmark’s chosen place to hunt. (Why do scientists chose to examine their navels or their own backyards instead of looking at the BIG picture…no doubt it is a 9 to 5 job and more research funding that drives most people.)
    Svensmark may be immortalized for slaying more than just an elephant – he may be remembered for slaying the CO2 man-made Global Warming Dragon, once and for all. It helps to have a theory that actually works when trying to shoot down a “consensus” and so far the skeptics have no alternative theory to displace the faulty “CO2 runs the show” hypothesis (other then to point out that obvioulsy CO2 has a minor effect).

  77. Leif Svalgaard (10:27:33) :
    Thanks for the reference.
    One does not need galactic cosmic rays in particular if the every day available solar wind can have a strong effect. GCR would be another factor in the brew.
    If one accepts this study, it indicates that precipitation is corelates to the magnetic fields better than to the sunspot cycle,no?

  78. I may be wrong here but I think the Mississippi is the most engineered river in the world. I’m not sure the flow rate is a good proxie for anything with all the dredging going on upstream, downstream, and on the tributaries. And the all the dams on the Mississippi for navagational purposes. No, No, the stream flow data is to human manipulated to mean much to me but I may be wrong.

  79. George E. Smith (11:25:24) :
    That’s what I see when I look at those graphs, either the original one, and the 34 month delayed one.
    34 YEARS, not months
    anna v (11:41:01) :
    GCR would be another factor in the brew.
    If one accepts this study, it indicates that precipitation is correlates to the magnetic fields better than to the sunspot cycle,no?

    Perry himself advocates GCRs which are correlated with sunspot number but much more waekly with aa-index, which in turn is mostly determined by solar wind speed and less by the magnetic field.
    The Inconvenient Truth (11:26:55) :
    I find it highly amusing that there is so little research in Svensmark’s chosen place to hunt.
    Because most scientists don’t find much there to hunt for. Measurements of the albedo shows that it does not track the solar cycle and hence not GCRs.

  80. The Inconvenient Truth (11:26:55) : not only Svensmark but also UN ‘s FAO ORG, in the document cited by Dr. Gerhard Loebert (02:31:10) : .
    Temperatures follow (with a time lag of course) LOD. That FAO curve perhaps in turn oscillates also.
    As you pointed out, there is a tendency to watch fixedly ownself’s navel.
    While everybody, as the fool in the famous Beatles’ song “The fool on the Hill” keep on singing:
    But the fool on the hill,
    Sees the sun going down,
    And the eyes in his head,
    See the world spinning ’round.

  81. The Inconvenient Truth (11:26:55) :
    so far the skeptics have no alternative theory to displace the faulty “CO2 runs the show” hypothesis (other then to point out that obvioulsy CO2 has a minor effect).
    One should not support a theory just because it is needed to counter a theory one doesn’t believe in.
    —–
    There are oceans [no pun] of correlations. One that was particularly popular some time ago was Mitchell’s 22-yr cycles of droughts in the US. This is clearly at variance with an 11-yr cycle as in Perry’s paper. And so on. Many of the correlations are contradictory and there is very little of the normal building on other’s results in the field.

  82. yonason (00:20:37) “Am I supposed to believe that from solar activity […] could have been predicted? […] Sorry, but I’m not buying it.”
    …and I’m not selling it. Until the conditioning is worked out, only fools & gamblers are predicting. Further analysis is warranted, but the job is far from done.

    Mike McMillan (02:03:50) “Acoustic from cosmic rays? You must mean that ping I hear when they bounce off my tinfoil hat.”
    Do you really think I’m talking about sound? [ :
    lol at your tinfoil ping (no sarcasm intended – rhetorical Q – thanks for the exchange of light-hearted humor)

  83. Leif Svalgaard (12:23:14) : One should not support a theory just because it is needed to counter a theory one doesn’t believe in.
    I agree, however, Svensmark’s theory has not been debunked (although it is hardly proven either) and yet it appears extremely promising in explaining historical data as well as current observations…at least to a far better degree than atmospheric CO2 concentrations which seem to be influenced by Global temperatures rather than being implicated as the cause behind Global temperature fluctuations….
    Am I missing something? What new evidence has shot down in flames the Svensmark hypothesis?

  84. Speaking of “Oceans [literally] of Evidence,” Tim Patterson of Carleton University in Ottowa, has a wonderful series that elucidates the 11 and 22 year cycles, and much more. Here’s the first of 3.

    The rest can be found there.

  85. Leif Svalgaard (06:23:56) :
    Geoff Sharp (05:49:32) :
    Dr. Scafetta is switched on also, and recently showed the correlation of the PDO and Solar distance from SSB
    Is there a 34-year lag?

    I found a similar 33 year lag between my correlation of solar equatorial distance from the Solar System centre of mass and the changes in Earth’s length of day:
    http://i630.photobucket.com/albums/uu21/stroller-2009/lod-ssb.gif
    I think this is a very promising line of research.

  86. The earlier graph’s Y axis is labeled as square feet per minute, the later one is labeled “cubic feet per minute”. The data are, however, identical.

  87. One more thing occurred to me about the 34-year lag hypothesis: why an (essentially) exact multiple of the solar cycle? It is being discussed as if this was perfectly natural — assuming the causal connection, plus the existence of some lag. Of course it’s not! Why not 17 years, or 3, or 28?
    What mechanism can be proposed that would make the lag conform to the periodicity of the causal agent? If there is a lag, it pertains to geographical and hydrological artifacts. Am I misunderstanding the nature of the “lag”? I have assumed that the hypothesis contends that the geomagnetic flux affects precipitation or other factors IMMEDIATELY, but that the effect somehow consumes a certain fixed lag time before its fingerprint appears in the Mississippi flow.
    Does the paper suggest that there is a 34-year lag between the solar cycle and any physical signal in the earth’s system, be it PDO or other cyclical phenomena? On the face of it, this is preposterous (I say this as one who tentatively accepts that there may be value in the supposition of a detectable relationship between solar flux and Mississippi flow). What could mediate such a lag, and with such a constant length and — back to the original point — why over an exact multiple of the solar cycle?

  88. My daughter attends KU and I ran into this guy at a football game. Since I’m involved in Agriculture and watershed management, I went to his website. It looked interesting and I forwarded all the information about 6 weeks ago on TIPS AND NOTES TO WUWT. But, as always, Leif steals the spotlight.
    :<[[

  89. The Inconvenient Truth (13:16:09) :
    Am I missing something? What new evidence has shot down in flames the Svensmark hypothesis?
    Nothing especially [except the albedo measurements], but the hypothesis was never well founded to begin with.
    R. Craigen (13:58:25) :
    One more thing occurred to me about the 34-year lag hypothesis: why an (essentially) exact multiple of the solar cycle?
    I think one piece of his evidence is that there are 11-yr cycles in both flow and aa. Would it be convincing if the cycles weren’t there or didn’t line up? So, consider to perfectly cyclical phenomena. If you try to find a ‘lag’ by cross-correlation [as I understood he did], then you’ll find perfect correlation at lag 0, again at lag 1 cycle, and 2 cycles, and 3 etc. The maximum correlation would a at lag of a whole number of cycles, therefore the appropriate lag is 3 cycles, not 34 years. In the real world the correlation is not perfect but will still peak where the cycles line up the best, i.e. the two cycles would be in sync as much as possible.
    Tim Clark (14:54:24) :
    But, as always, Leif steals the spotlight.
    Especially when I screw up so people can delight in dumping on me 🙂

  90. The Inconvenient Truth (13:16:09) :
    Am I missing something? What new evidence has shot down in flames the Svensmark hypothesis?
    Nothing especially [except the albedo measurements], but the hypothesis was never well founded to begin with.
    R. Craigen (13:58:25) :
    One more thing occurred to me about the 34-year lag hypothesis: why an (essentially) exact multiple of the solar cycle?
    I think one piece of his evidence is that there are 11-yr cycles in both flow and aa. Would it be convincing if the cycles weren’t there or didn’t line up? So, consider to perfectly cyclical phenomena. If you try to find a ‘lag’ by cross-correlation [as I understood he did], then you’ll find perfect correlation at lag 0, again at lag 1 cycle, and 2 cycles, and 3 etc. The maximum correlation would a at lag of a whole number of cycles, therefore the appropriate lag is 3 cycles, not 34 years. In the real world the correlation is not perfect but will still peak where the cycles line up the best, i.e. the two cycles would be in sync as much as possible.
    Tim Clark (14:54:24) :
    But, as always, Leif steals the spotlight.
    Especially when I screw up so people can delight in dumping on me 🙂

  91. R. Craigen (13:58:25)
    That’s what we would all like to know.
    The approx 30 year phase change of the oceans could be 3 solar cycles long just from coincidence. The timing could be due to some other mechanism altogether.
    It isn’t an exact correlation anyway. The length between oceanic phase changes varies and the length of the solar cycle varies. I don’t think anyone has yet established a direct causative link between longer or shorter solar cycles and longer or shorter oceanic cycles or vice versa.
    The global water flow effects on a regional basis are well matched with oceanic changes as are air temperature changes but the link to solar variability is a problem.
    From obervations we seem to have at least 3 ocean cycles impacting on global climate. The ENSO phenmenon on an interannual basis, the PDO phase shifts on a multidecadal basis and it seems another longer cycle over 1000 years which supplies the changes from Mediaeval Warm Period to Little Ice Age to Modern Maximum.
    I diagnose the 1000 year cycle as also having been ocean induced because I read somewhere that during the LIA the ITCZ was on or near the equator instead of somewhat north of it as now. In my opinion latitudinal air circulation shifts are a fingerprint of oceanic forcing.
    The interesting thing is that those three oceanic cycles can explain all observed climate shifts without needing much or indeed any solar forcing at all.
    Now there is some historical correlation between low solar activity and colder temperatures but the correlation is not perfect because oceanic and solar effects can combine or offset one another.
    It remains possible (as per Leif) that the solar influence is very tiy compared to oceanic effects and that it is merely a coincidence that for the past 2000 years or so the solar changes have been in phase with the oceanic forcings.
    We need information going back more than 2000 years for both solar and oceanic cycles to see whether oceanic cycles are ever out of phase with solar variations but we just don’t have enough data on the oceans to work that one out. Even using the positions of the ITCZ and the other air circulation systems as a proxy for ocean cycles we cannot get back prior to the LIA.
    I think that what we need to do is just watch what now happens under the current weak solar regime and try to disentangle any solar influence from the undoubted effects that we are already seeing from the oceanic phase change which was due around now whether or not the sun became less active at the same time.
    If oceanic cycles can explain all observed climate changes going back at least 500 years then the sun becomes a minor issue and we need to consider instead whether human produced GHGs might have a cumulative long term effect on the Earth’s energy budget and equilibrium temperature.
    To resolve that issue we need most to know whether the ocean skin effect is real because if those gases cannot warm the oceans then the oceans will not allow them to alter the Earth’s equilibrium temperature because the oceans control the air temperatures.
    The climate system deals well enough with warmer ocean surfaces causing water vapour changes (a greenhouse gas) by altering the speed of the hydrological cycle. The same mechanism would also deal with extra energy from CO2 in the same way. Both warmer ocean surfaces and more CO2 (via more downward re-radiated infra red) cause the same system response i.e. an increased rate of evaporation so it must be the case that the same mechanism deals with both in the same way.

  92. “Since GCRs follow the the sunspot numbers and not the aa-index, the proper parameter to compare with would be the sunspot number.”
    Leif, wouldn’t the earth’s magnetic field modulate the effect of the GCR’s? Could it be that this might mean the aa-index is as important as the sunspot numbers?

  93. Re: R. Craigen (13:58:25)
    The cause of the lag is simple. All one has to do is note the amount of time that passed between the ~1900 aa index low and the 30s drought.
    To be clear: There are a lot of nonsensical notions about lags in WUWT discussions. People here (generally speaking) need to do some learning about what lags convey – seriously.

  94. Actually, perhaps we could reconstruct ocean states prior to the LIA by guessing at the approximate latitudinal positions of the air circulation systems from the records of ancient civilisations. It would probably be a bit too coarse to distinguish all three timescales but perhaps we could get more evidence of a 1000 year cycle.
    Note my deliberate error in my previous post in implying that ocean cycles could affect solar cycles 🙂

  95. tallbloke (15:43:36) :
    Leif, wouldn’t the earth’s magnetic field modulate the effect of the GCR’s? Could it be that this might mean the aa-index is as important as the sunspot numbers?
    The Earth’s magnetic field is the primary modulator of GCRs. Typically ten times as important than the Sun. But the Earth’s magnetic field varies very slowly [centuries – millennia]. The aa-index is a measure of the very small [less than 1%] additional variation of the geomagnetic field that is caused by the solar wind and thus a variable sun. The GCRs are mainly modulated by flares and CMEs which largely follow the sunspot cycle and not the aa-index. The latter being dominated by high-speed solar wind streams which often occur late in the cycle. The correlation between GCRs and sunspot number is larger than that between GCRs and aa-index. So, if GCRs are the mechanism [as the author claims], the sunspot number is the correct parameter to look at.

  96. “”” Leif Svalgaard (11:55:23) :
    George E. Smith (11:25:24) :
    That’s what I see when I look at those graphs, either the original one, and the 34 month delayed one.
    34 YEARS, not months “””
    Slip of the tongue there Leif; but now that you mention it, I don’t see that a 34 month delay looks any better either. Is there a message in that 34 years; or is there no theoretical basis for assuming such a delay ?

  97. George E. Smith (16:15:28) :
    Is there a message in that 34 years; or is there no theoretical basis for assuming such a delay ?
    Due to the way he determined the delay, it has to be a whole number of cycles. I don’t know if there is any basis for 3 cycles.

  98. The Inconvenient Truth (13:16:09) :
    Am I missing something? What new evidence has shot down in flames the Svensmark hypothesis?
    Nothing especially [except the albedo measurements], but the hypothesis was never well founded to begin with.

    In that case, how about a critique of the Svensmark theory on WUWT. I know you are all pretty busy but I would very much appreciate learning how certain albedo measurements shoot holes in this “never well founded” hypothesis. I hope I am not in the minority here – so others please chirp up if you are also interested in a critique of what is badly wrong with Svensmark theories.

  99. yonason (13:30:47) :
    Speaking of “Oceans [literally] of Evidence,” Tim Patterson of Carleton University in Ottowa, has a wonderful series that elucidates the 11 and 22 year cycles, and much more. Here’s the first of 3.

    The rest can be found there.

    Nice video lecture. I’m curious where he gets his cosmic ray data from which he deduces a bigger variation over gleissberg cycles from.

  100. Stephen Wilde (15:50:22) :
    Actually, perhaps we could reconstruct ocean states prior to the LIA by guessing at the approximate latitudinal positions of the air circulation systems from the records of ancient civilisations. It would probably be a bit too coarse to distinguish all three timescales but perhaps we could get more evidence of a 1000 year cycle.

    Stephen, watch the video linked by yonason above. The silt cores show the cycles and he also discusses the movement of the jet streams north and south.

  101. Leif Svalgaard (16:25:08) :
    George E. Smith (16:15:28) :
    Is there a message in that 34 years; or is there no theoretical basis for assuming such a delay ?
    Due to the way he determined the delay, it has to be a whole number of cycles. I don’t know if there is any basis for 3 cycles.

    Perhaps something worth thinking about is that since there is a difference between odd and even solar cycles, successive 33 year periods will alternate between having two odd + one even cycles and two even + one odd.
    I think this may turn out to have a bearing on the 60 year oceanic cycles once the cloud amplification issue becomes clearer.

  102. Leif Svalgaard (16:10:00) :
    tallbloke (15:43:36) :
    Leif, wouldn’t the earth’s magnetic field modulate the effect of the GCR’s? Could it be that this might mean the aa-index is as important as the sunspot numbers?
    The Earth’s magnetic field is the primary modulator of GCRs. Typically ten times as important than the Sun.

    Bonzer!
    So although the GCR variation follows the solar cycle closely at the decadal level, the effects of GCR’s on the Earth’s climate oin the long term is modulated much more by geomagnetism than changes in TSI.
    So if Svensmark is on the money, he has identified an important climate driving mechanism on the multi-centennial scale.
    I like it. 🙂

  103. tallbloke (16:44:00) :
    So if Svensmark is on the money, he has identified an important climate driving mechanism on the multi-centennial scale.
    Except that the data we have don’t support any climate connection. Although a bit dated, the following Figure
    http://www.leif.org/research/CosmicRays-GeoDipole.jpg
    shows how the Earth’s dipole moment and 14C [GCRs] have varied the past 10,000 years. The temperature variations do not fit that pattern.

  104. If there is no reason for a 34 year lag, then someone needs to tell the climate modelers that (there is a roughly 30 year lag built in right now due to the oceans absorbing some of the additional greenhouse forcing).
    Okay that was off-topic, but I am not a big fan of smoothing and long-term lags. By the time one builds in a 36 month moving average into two different series, all kinds of spurious correlations start to show up. Now lag one of these smoothed series over any time period one wants and now you have double spurious correlations showing up.
    There can be lags between indicators but one needs to have a solid physical basis for showing how that lag occurs. Smoothing is okay if there is uncertainty in the data or seasonality in the data or if it is too variable to view/study properly.
    There is a rationale for smoothing the Mississippi flows over a 12 month moving average given there is a seasonal component to it. But after that, one is just begging for a correlation to show up that isn’t there. I don’t think there is any rationale for smoothing the geomagnetic index. It is already smooth enough and the solar cycle signal should be preserved anyway if one is trying to show it impacts something on Earth.
    So, redo the data with no lags and smooth the Mississippi flows over 12 months and let’s see what shows up.

  105. Bill Illis (17:46:43) :
    I don’t think there is any rationale for smoothing the geomagnetic index.
    There is a 20% semiannual variation, so a 12-month running mean would remove that too. The aa-index is the wrong one to work with if the GCRs are part of the equation. Then it must be the sunspot number.
    So, redo the data with no lags and smooth the Mississippi flows over 12 months and let’s see what shows up.
    Then there is hardly anything to write home about. It would seem yo me that the paper claims to be able to predict streamflow 34 years in advance.

  106. Re: R. Craigen (13:58:25)
    The cause of the lag is simple. All one has to do is note the amount of time that passed between the ~1900 aa index low and the 30s drought.
    – P. Vaughan

    To be clear, Paul, I don’t have any difficulty imagining how one might infer a lag from data. When I asked about a cause I was talking about a physical phenomenon, not a relationship between two graphs. What you describe is a criterion for identifying a candidate period, under the assumption that geomagnetic flux influences river flow after a time lag. It is no more a cause of that lag than the speedometer is the cause of a vehicle’s speed.
    I was looking for something more like the following (not a serious proposal): “heat or mechanical energy caused by a spike in geomagnetic flux, stored deep in the ocean (mountain snow, upper-atmosphere moisture, etc…), then undergoes some evolutionary process that takes approximately 34 years before it ultimately produces a sudden influx of H20 to the Mississippi basin, causing higher flow.”
    It’s easy to find putative correlations between roughly cyclical data by shifting until a pretty-good match shows up. But time-shift implies lag, and a specific lag burdens the interpretation of the causation–not only must one explain how flux affects flow, but why it takes exactly this amount of time, and how the signal’s shape is so well insulated from other influences over that period.
    I hope that helps.

  107. To further illustrate my thought on causation of lag, consider the well-known 800 year lag between warming/cooling changes and CO2 concentration changes in the geological/ice records.
    One might say, “that’s preposterous — what could cause an 800 year lag?” Seeing the record might convince you that there is some lag-causing agent, but it certainly doesn’t constitute an “explanation”.
    Fortunately, for CO2/temperature time lag, there is a ready explanation at hand (I haven’t done the calculation so I don’t know how good it is, but it seems in principle to do the trick).
    It has to do with heat and carbon transfer through levels of the ocean, the tendency for the deep ocean to act as a CO2 sink, and the rate at which Henry’s law affects CO2 equalization on the boundary layer of the ocean at different temperatures.
    Maybe someone better versed in fluid dynamics and the physics involved can opine as to the sufficiency of this explanation. I confess to never having seen it written up — I just assume that it’s the obvious cause for the temperature/CO2 lag.
    Sorry for the digression, but I thought discussion of the problem I was considering would possibly benefit from a positive example of what I think is needed.

  108. Ok, picking up on some of the later comments, there appear to be some plausible lag-explanations. If the earth’s magnetic field is the primary modulator, and if it responds in a resonant fashion to solar flux then we have an explanation of response at a whole number of solar cycles. I think of my coffee-cup that can overflow when I’m walking if I take 3 or 4 injudicious steps that give pulses causing constructive interference (to prevent this time one’s steps to avoid resonance and cause destructive interference).
    There remains a smoothing problem: Addition of waveforms in which the signature of a pulse appears after 3 cycles also means that the energy of at least 3 (probably 6) pulses is present in each peak. Thus, there should be a 3-cycle (or more) averaging, and the variation in signatures from pulse to pulse would be more masked than in these graphs. At least SOME of the correlation must be purely coincidental, even if such a mechanism has validity.

  109. R. Craigen (19:00:57) :
    If the earth’s magnetic field is the primary modulator, and if it responds in a resonant fashion to solar flux
    Some misunderstanding here. The Earth’s main field is the primary modulator of GCRs, but on time scales of thousands of years, and it does NOT respond to solar activity. Geomagnetic activity [the aa-index] does NOT modulate GCRs. Sunspot activity does. And there is no coupling between solar activity and Earth’s internal magnetic field. So no resonance.

  110. PERFECT OPPORTUNITY TO TEST THE PREDICTIVE POWER OF HIS THEORY
    Leif Svalgaard (18:07:58) :
    “It would seem yo me that the paper claims to be able to predict streamflow 34 years in advance.”
    Yes, as I also mentioned in my ” yonason (00:20:37) : [to] Paul Vaughan (23:29:09) : ” above.
    Well, thanks to the folks at IceAgeNow, I was tipped off to a not so little scenario to test the theory. Does that theory/hpothesis find that what happened on or about 1975 can predict the massive floods currently underway?
    http://www.marketskeptics.com/2009/10/rains-swamp-crops-and-wash-away-any.html
    My guess is that they don’t. But I’m willing to be surprised.

  111. Sioned L (19:53:09) : [in the Sea Surface Temperature makes a jump” posting] says:
    Interesting article in the Salt Lake Tribune today 11/02/09: “Scientists Find patterns in Utah’s wet-dry cycles”
    http://www.sltrib.com/ci_13681950
    Note the figure comparing SST in the southwestern Pacific to precipitation in Utah.
    They have a good correlation with only a 3 – 4 year lag.
    Compare this to the Charles A. Perry paper
    http://ks.water.usgs.gov/waterdata/climate/c.perry.asr2007.pdf
    which compares SST with Great Basin precipitation lagged 3 years (fig.5)
    (then he goes on to 34 year lags)

  112. Except that the data we have don’t support any climate connection. Although a bit dated, the following Figure
    http://www.leif.org/research/CosmicRays-GeoDipole.jpg
    shows how the Earth’s dipole moment and 14C [GCRs] have varied the past 10,000 years. The temperature variations do not fit that pattern.

    Leif,
    You are out of date. Svensmark has proposed that it is NOT simply overall cosmic ray flux but HIGH ENERGY cosmic ray flux (the muons that reach the lowest levels in the atmosphere). According to Svensmark, the earth’s magnetic field does not affect the highest energy cosmic rays (it only modulates the weaker kind). Svensmark has modified his theory to say that the earth’s magnetic field has little bearing on high energy cosmic rays.
    Can you comment in light of this modification of his theory….certainly he is correct that most cosmic rays DO NOT reach the lower levels of the atmosphere…

  113. This discussion reminds me of sitting on a beach by the sea lake and watching the waves. Something like Pooh sticks.
    Speed boats criss cross the middle of the lake and about 20 minutes later their wake arrives at my feet. Except there are more than one boats, their wake hits the far shore and comes back too, and here I sit, watching my flip flops floating down beach guessing which speed boat driver to grumble at.
    All these correlation chasing and lags come into focus if one seriously considers chaotic behavior, and climate science is a long way off on getting a real handle on it. It needs a change of mentality to be able to start thinking in terms of chaotic models. Tsonis et al have made a start, but it is just a start.
    It is the proverbial elephant and four blind men parable.

  114. I see the discussion continues to be off-the-rails regarding lags (& now smoothing too) —- let’s try this again:
    The cause of the lag is simple. All one has to do is note the amount of time that passed between the ~1900 aa index low and the 30s drought.
    A little more:
    There are different kinds of lags.
    Say I hear a train moving through hilly terrain. If I re-hear the whistle, the lag conveys info about a physical (lag) process.
    Say I note a 6 month lag between NH & SH temperatures — that conveys info about anti-phase.
    Say I note cyclical (time-integrated) cross-correlation, like here…
    http://www.sfu.ca/~plv/ccLaaLR1crf4532.png
    http://www.sfu.ca/~plv/ccLR1CRF.PNG
    …that tells me I’m dealing with cycles (with similar period).
    Say I note the 34 year lag being discussed here — that tells me global minima (in the 2 series) have hijacked my best-lag (…which is what people here are failing to realize).
    I’m resorting (again) to being blunt because I am pressed for time, but nonetheless it is important to get the point across (perhaps less diplomatically than I would if I had more time).
    This is basic Stat 101, part 2: intro regression. The global minima & global maxima have high leverage — THAT is what is going on here …but there are other things going on here too, like 11 year cycles (not just ones of solar origin – there is confounding) & even interannual shared-variance – i.e. it is complex (speaking more broadly) – but as for the “why 34 years?” – well, that is just dead-simple – understanding it is as simple as understanding why a regression line slopes upward if the data-points do so. [I think it probably helps with intuition-development if people have programmed their own cross-correlation analyzers and run dozens of familiar series through them; I realize not everyone has time for that, so I share these brief notes.]

  115. anna v (22:38:18) “Tsonis et al have made a start, but it is just a start.”
    These guys claim this:
    “This is the first time that this mechanism, which appears consistent with the theory of synchronized chaos, is discovered in a physical system of the size and complexity of the climate system.”
    They are not the pioneers of the methodology (but I suppose their claim might help keep the gravy train flowing…)

  116. Leif Svalgaard (16:10:00) :
    The Earth’s magnetic field is the primary modulator of GCRs.
    Leif Svalgaard (19:29:26) :
    Geomagnetic activity [the aa-index] does NOT modulate GCRs.

    [snip ~ I believe he was asking for clarification ~ ctm]

  117. I see what Leif means now, and in the light of the the comment about Svensmark’s theory from Jeremy to Leif above,
    “You are out of date. Svensmark has proposed that it is NOT simply overall cosmic ray flux but HIGH ENERGY cosmic ray flux (the muons that reach the lowest levels in the atmosphere). According to Svensmark, the earth’s magnetic field does not affect the highest energy cosmic rays (it only modulates the weaker kind).”
    I’d be interested in any ideas about the way changes in the earth’s geomagnetic field affect GCR’s and climate. I’ve watched an interesting animation of the changes in the field from 1600AD and it’s highly suggestive that variation in the field affects climate at a regional and global scale. I suspect the effect of the geomagnetic field on GCR’s of various energies is proportional, rather than cutting off at high energy levels.

  118. Paul Vaughan (23:33:05) :
    NICE FIND!
    Glenn (23:43:42) :
    O.K., I see that now. But look at the graph in “Update2” where the sunspot data says exactly the opposite.
    My problem is that I don’t see a mechanism. Maybe I just don’t know enough background material, though it seems few others here don’t seem to either, or they could explain it.
    I don’t know how they came up with 34 years, except that the two phenomena match up pretty well, mostly, when that number is used [though the great flood of 1936 is completely missed, and an event that would have been predicted for 1960 never occurred].
    I feel somewhat less uncomfortable with correlations in real time, but picking what appears to be an arbitrary period of more than 3 decades virtually screams for an explanation, and I don’t see they have given any; other than some general concepts, but they don’t give details on how they all fit together. Did I just not read carefully enough, or is there other material out there? Unless I see it, and it makes sense, I can’t.

  119. Further to Paul Vaughan (22:41:07) …
    Suggested exercises:
    Remove the multi-decadal patterns and then re-run the cross-correlation analysis. See how that affects the best-lag for some insight.
    Based on some of the comments, I can also suggest running multi-scale cross-wavelet phase-contrasts of aa, GCR, & solar variables (for constructive insights that aren’t forthcoming via linear-correlation analysis).
    Best Regards.

  120. @Leif Svalgaard
    I’ve asked you 2 times before, why there is a correlation between climate (glacier stands in the Alps and in the Andes) and the LIA-minima, and you answered me both times, that there is no such correlation for the global climate (as far as I remember).
    And here again you don’t answer my questions.
    Do you think that the correlation found all over Europe, Greenland and now in the Andes is pure coincidence, or “luck”? Or do you think that there is a link between climate and the sun, but not on a global level.
    Pro-Agw-scientists like Stefan Rahmstorf and Georg Hoffmann don’t answer my questions at all. I feel like I’m an actor in a truman show. I really hope that some of you scientists are little nicer to us laymen and try to give us answers. Very many people have the same questions!
    I read these blogs and books now for more than 10 years and I still found no answers to these questions. In my opinion the sun-climate-link is obvious in Europe.
    I would be very pleased to get an answer from you.
    Thank you in adavance and have a nice day
    Eddy

  121. I was astounded to follow through Dr Gerhard Loebert’s comments, rather belatedly unfortunately. I think this man is on to something big. Even if he too has his personal limitations. I’ve long had my suspicions that even back of Svensmark is something deeper in the line of “causation” and that it has to do with quantum physics, Zero Point Field material – as Lynne McTaggart explains beautifully in her book of that name, and as I’ve seen other incidental evidence for elsewhere.
    I’m certain that astrophysics is ultimately going to unravel and elucidate the prime drivers for climate. Not just the Sun but behind that, the SSB. Not just the SSB, but behind that, the galactic centre and the workings of the Zero Point Field. And always, statistically significant correlations necessary first, and likely to precede explanations of causation.
    Paul Vaughan, I’d love to see a post by you here.

  122. Edouard (01:54:06) :
    “@Leif Svalgaard
    I’ve asked you 2 times before, why there is a correlation between climate (glacier stands in the Alps and in the Andes) and the LIA-minima, and you answered me both times, that there is no such correlation for the global climate (as far as I remember).”
    I too feel frustrated that answers are not forthcoming. I don’t think it’s because scientists don’t want to give us the answers, the real reason is that big systems like climate and solar activity display much turbulent chaotic behaviour and science is poor at dealing with these sorts of systems.
    Worthwhile reading the post from anna v above, as this is a good illustration of why chaos is such a big problem when it comes to trying to understand what is happing now and trying to make predictions about the future.

  123. Jeremy (22:01:54) :
    Svensmark has proposed that it is NOT simply overall cosmic ray flux but HIGH ENERGY cosmic ray flux (the muons that reach the lowest levels in the atmosphere).
    The muons are generated by the cosmic rays as they hit the atmosphere.
    According to Svensmark, the earth’s magnetic field does not affect the highest energy cosmic rays (it only modulates the weaker kind).
    By the same token, the solar wind does not affect the highest energy cosmic rays either.
    Edouard (01:54:06) :
    I’ve asked you 2 times before, why there is a correlation between climate (glacier stands in the Alps and in the Andes) and the LIA-minima
    The global temperatures were low during the LIA [by definition – since the LIA is defined as a time when temperatures were low], but the LIA lasted much longer than the low solar activity that by coincidence was also found at times during the LIA.

  124. According to Svensmark, the earth’s magnetic field does not affect the highest energy cosmic rays (it only modulates the weaker kind).
    By the same token, the solar wind does not affect the highest energy cosmic rays either.

    Ok let me get this straight, the earth’s magnetic field sits tightly around the earth. The heliosphere goes out around 9 times the distance to Neptune. The heliosphere ,”by the same token”, does not affect high energy cosmic rays from space any more than the localized Earth’s magnetic field does. I find this odd fact quite surprising – are there any papers showing this unusual fact to be the case – or are you simply making an assumption that two independent and unrelated fields behave in exactly the same way when it comes to cosmic rays?

  125. Leif Svalgaard (04:54:21)
    According to Svensmark, the earth’s magnetic field does not affect the highest energy cosmic rays (it only modulates the weaker kind).
    By the same token, the solar wind does not affect the highest energy cosmic rays either.”
    Doesn’t a stronger solar wind push the heliopause further out into space, thus makes it harder for GCR to reach the earth?

  126. The majority of the water that flows out of the Mississippi River is sourced from the Gulf of Mexico. The clouds that come from the prevailing westerlies, loose most of their rain in Rockies, which sets up the rain shadow band of dry land that ends about 250 miles west of the Mississippi River.
    The weather pattern that creates enough rain, that will saturate the ground, and then flow into the rivers that form the river valley, is a slow moving low, that moves across the South, drawing moisture north were it clashes with colder air decending out of Canada.
    The larger precip patterns, come from strong sub-tropical highs that set up in the pacific and create a jet stream that buckles over the middle of the lower 48, and create this pattern of rain, that will end up flowing out the Mississippi River.
    I can think of several reasons why this would occur on a 34 year pattern, but I don’t yet understand, why there is a 34 year lag. My guess would be the way the Pacific and the Atlantic change phases, and how this area in the Gulf of Mexico can be pushed and pulled by both oceans.

  127. Jeremy (07:08:19) :
    Ok let me get this straight, the earth’s magnetic field sits tightly around the earth. The heliosphere goes out around 9 times the distance to Neptune.
    Bengt A (07:10:23) :
    Doesn’t a stronger solar wind push the heliopause further out into space, thus makes it harder for GCR to reach the earth?
    It is not the size of the heliosphere as such that determines the modulation of the GCRs. Within the heliosphere these are knots and bundles and sheets of tangled magnetic fields. Those magnetic inhomogeneities are what scatter GCRs. Charged particles ‘gyrate’ around magnetic fields. The radius of the gyration determine if the field will scatter the rays. If the radius is large enough, the GCR will not ‘see’ the magnetic field irregularity, much as a big truck will not be bothered by a small pothole; the radius depends on the energy of the GCRs and inversely on the magnetic field strength. The Earth’s magnetic field is 100,000 times stronger than the field in the distant heliosphere, so will mean a similar reduction of gyroradius. The highest energy GCRs have very large gyroradii by virtue of their high energy [they are the big trucks]. So, at low energy, both Earth and Sun will modulate GCRs, at the highest energies, neither will. For medium energies, there will a ‘sweet spot’ where the Earth will not, but the sun will. Svensmark argues that ‘his’ GCRs sit right at that sweet spot. This is a weakness of his theory. Furthermore, at higher energies the GCRs become progressively rarer, so another ‘sweet spot’ must be there, namely that just enough GCRs to have effect are present at the first sweet spot. Too many special pleadings like this make the argument weaker.

  128. Jeremy (07:08:19) :
    Bengt A (07:10:23) :
    Without going into too much technical discussion, a graph often helps. Look at Figure 1 of
    http://www.atic.umd.edu/pub/Shikaze.pdf
    that shows the change in GCR intensity for minimum year 1997 to maximum year 2000. Note that for low energies there is a clear difference [‘modulation’], but for higher energies the curves come together and there is in the end [at the highest energies, the right hand edge of the graph] no more modulation.

  129. The nonsensical musing about the 34 year lag continues.
    A little more to help people see clearly:
    It’s not bivariate. [Thank goodness we don’t live in such a simple world.] Think about conditioning variables.

  130. Paul Vaughan (11:36:22) :
    The nonsensical musing about the 34 year lag continues.
    A little more to help people see clearly:
    It’s not bivariate. [Thank goodness we don’t live in such a simple world.] Think about conditioning variables.

    The reason people continue is that you give them no help in clearing this up, just general waffle that they can’t relate to.

  131. Leif Svalgaard (09:08:14)
    For medium energies, there will be a ’sweet spot’ where the Earth will not, but the sun will. Svensmark argues that ‘his’ GCRs sit right at that sweet spot. This is a weakness of his theory.
    I believe that this thing with medium energies is derived from calculations with the CORSICA-program as explained in Svensmarks latest papers. There are convincing arguments for that if GCRs affect low clouds it ought to be mid and high energy GCRs. The highest energies will not be modulated by either geomagnetic or inter planetary fields (as you state) so if there is modulation that leaves us with the mid energy GCRs. We can call this a ‘sweet spot’ but I still don’t see the weakness in this line of argumentation?
    Furthermore, at higher energies the GCRs become progressively rarer, so another ’sweet spot’ must be there, namely that just enough GCRs to have effect are present at the first sweet spot. Too many special pleadings like this make the argument weaker.
    I don’t understand this part about another sweet spot? To affect low clouds, according to Svensmark, you need CGR in the GeV-range. That is all there is to it. There sure are loads of CGRs in the GeV-range, though not as many as in the lower energies, so that should not be a problem for his theory.

  132. Bengt A (11:50:50) :
    There are convincing arguments for that if GCRs affect low clouds it ought to be mid and high energy GCRs.
    You have now moved from ‘highest’ to ‘mid’. These also hit the atmosphere at the level of the high clouds.
    There sure are loads of CGRs in the GeV-range, though not as many as in the lower energies, so that should not be a problem for his theory.
    the issue is whether there are enough. People have modeled this and find that there are not enough to have any effect.

  133. Lief: Too many special pleadings like this make the argument weaker.
    Thanks for the reply. I see your argument makes sense… a “special case” tends to weaken the foundation for a hypothesis. So I agree with you. I also appreciate your kindness in educating me with how you came to your conclusions.
    However, the BESS paper shows a difference in solar effect on cosmic rays of an order of magnitude (factor of 10) at low energies to a factor of 2 at medium energies and down to a small percentage at higher energies (between what looks like 1% at the very highest and to around 10% at 10 GeV/n).
    Far from undermining Svensmark’s theory the data seems to explain very well why the modulations in cloud cover might be very small (a few percent). If all cosmic rays in general were influencing cloud cover then it would indeed be a surprising special case that an order of magnitude (factor of 10) change in GCR’s would have such a small effect (a few percent in cloud cover over humid oceans). However, if only a small proportion of cosmic rays (those with enough energy to reach the appropriate lower altitude levels) are involved then it might be construed as actually encouraging (for Svensmark hypothesis) to see that modulations of a few percent can be expected at higher energies (in line with the necessary small magnitude of the albedo changes from cloud cover that are purported to influence global temperatures).
    Of course, although we disagree on the “weakness” of the Svensmark theory, this is all speculative conjecture about a mechanism. Only a proper lab experiment will determine if the relationship between GCR’s and cloud formation at various altitudes (pressure, temp & humidity) is of the precise type needed to make Svensmark theory plausible. This kind of experiment would either strengthen or help throw out Svensmark’s hypothesis. With so much money spent on Climate research (30 billion or more by some estimates), I sincerely hope Svensmark gets the necessary funding to make the requisite experiments.

  134. Jeremy (13:09:02) :
    >i>However, the BESS paper shows a difference in solar effect on cosmic rays of an order of magnitude (factor of 10) at low energies to a factor of 2 at medium energies
    People have calculated from present knowledge how large the effect might be and fins that it is generally too low:
    http://www.leif.org/EOS/2009GL037946.pdf
    My point is that the evidence is weak, the jury is out, and nothing is settled here. This is, of course, far from the rabid followers of the hypothesis that assume it true because it gives them an argument against AGW.

  135. Leif Svalgaard (13:51:50) :
    My point is that the evidence is weak, the jury is out, and nothing is settled here. This is, of course, far from the rabid followers of the hypothesis that assume it true because it gives them an argument against AGW.

    And we wouldn’t want that now would we?
    It seems Leif is out to bash Nicola Scafetta, Henrik Svensmark and Mike Lockwood among others, and to promote the PMOD TSI of Frolich and the revised line toeing of Judith Lean.
    I see battle lines being drawn in the physicists sandpit.

  136. tallbloke (15:58:14) :
    It seems Leif is out to bash Nicola Scafetta, Henrik Svensmark and Mike Lockwood among others,
    Not Lockwood anymore. He has seen the light and agrees closely with us now, e.g. http://www.leif.org/research/HMF-Convergence.png And we don’t ‘bash’. Scientific disagreement is different from ‘bashing’.
    and to promote the PMOD TSI of Frolich and the revised line toeing of Judith Lean.
    Froehlich’s PMOD TSI is wrongly calibrated [and he knows it and is in the process of fixing it]. e.g. http://www.leif.org/research/TSI%20Difference%20PMOD-SORCE.pdf
    No battle lines. The way it works is that the wrongs are quietly forgotten [Lockwood is a good example] by everybody, except [at times] by lone holdout by the authors [and they eventually die].

  137. yonason (15:15:51) :
    “Seems the clouds never got the memo?”
    You can find ANY viewpoint whatsoever on the Internet, even that the Goracle is 110% correct. Reading the Pierce and Adams’s paper I find no flaws with it.

  138. Leif Svalgaard (04:54:21) :
    The global temperatures were low during the LIA [by definition – since the LIA is defined as a time when temperatures were low], but the LIA lasted much longer than the low solar activity that by coincidence was also found at times during the LIA.
    Not by coincidence, we had 3 of the strongest grand minima in a row all separated by 172 years average. This is the lowest and longest period of solar activity during the 11,000 year Holocene record.
    http://www.landscheidt.info/images/c14nujs1.jpg

  139. Leif Svalgaard (16:55:40) :
    “Reading the Pierce and Adams’s paper I find no flaws with it.”
    Nor does the author of the comment I linked to, as his comments show.
    http://theresilientearth.com/?q=content/attempt-discredit-cosmic-ray-climate-link-using-computer-model#comment-105
    But their paper is a model, and the fact is that clouds do seem to empirically respond to cosmic rays.
    Also, here’s 50 years of ocean data that correlte amazingly well with CRF
    http://www.sciencebits.com/calorimeter
    – more empirical evidence that says the models are missing something important.
    So, when it comes to a well written theoretical paper and empirical observations that contradict it, I have to side with the data over the calculations, at least until (reliable) contradictory data becomes available.

  140. yonason (17:41:41) :
    the fact is that clouds do seem to empirically respond to cosmic rays.
    Have a link to that? That clouds correlate with the solar cycle does not show that they correlate with cosmic rays.

  141. Leif Svalgaard (18:04:14) :
    Not at all:
    http://www.leif.org/EOS/2009GL039439.pdf

    Are you suggesting the work of Solanki, Usokin and Steinhilber et al is now in question?
    To deny the importance and strength of the Wolf, Sporer and Maunder minima is going against most of what recognized science suggests in this field. Perhaps it fits into Leif’s law:
    If it doesn’t fit your agenda its pseudo-science.

  142. Geoff Sharp (19:04:16) :
    Are you suggesting the work of Solanki, Usokin and Steinhilber et al is now in question?
    But there is a big question mark about the last 150 years of the series. Much too high. But that is another matter. The major flaw is with your statement not their work.

  143. Leif Svalgaard (19:30:51) :
    Geoff Sharp (19:04:16) :
    Are you suggesting the work of Solanki, Usokin and Steinhilber et al is now in question?
    —————
    But there is a big question mark about the last 150 years of the series. Much too high. But that is another matter. The major flaw is with your statement not their work.

    Your rambling….the last 150 years are not important, I was stating the LIA had some of deepest and most prolonged grand minima of the Holocene (which cant be argued against).
    Of course if was cold.

  144. Leif Svalgaard (21:56:38) :
    The 14C (INTCAL98) records certainly suggest the LIA was the longest and deepest. You can split hairs like you always end up doing, but it is not strengthening your argument.
    the LIA had some of deepest and most prolonged grand minima of the Holocene (which cant be argued against).
    Of course if was cold.

  145. Leif Svalgaard (22:08:58) :
    So you ARE arguing with Solanki and Usoskin….lets just leave it at that. Conversation is pointless, but I will continue to point out where you are leading us astray.

  146. Leif Svalgaard (12:39:43)
    There sure are loads of CGRs in the GeV-range, though not as many as in the lower energies, so that should not be a problem for his theory.
    the issue is whether there are enough. People have modeled this and find that there are not enough to have any effect.

    The knowledge about how aerosols are put together and aggregated into bigger aerosols is limited. Pierce & Adams make assumptions about the microphysics of aerosols in accordance with present scientific findings, but due to the uncertainty I think there are two alternative conclusions to draw from their results: GCRs have no effect on clouds or The reason that we see no effect of GCRs on clouds in Pierce & Adams model could be a result of their initial assumptions being wrong.
    A couple of months ago Svensmark published a paper on Forbush decreases. His conclusion is that during these events GCRs have a considerable effect on low clouds and this seems to contradict Pierce & Adams findings. So which one should we go with? The model of Pierce & Adams or the measurements of Svensmark?

  147. Pamela Gray (23:22:55) :
    Geoff, if you mean a collective “us” I will have to disagree about being led astray.
    I am sure I am not the only one here that think the Sun plays an important role in our climate.

  148. @Leif Svalgaard
    In the german wikipedia they say this:
    “Das Spörerminimum (nach dem deutschen Astronom Gustav Spörer) ist eine Periode besonders geringer solarer Aktivität im Zeitraum zwischen 1420 und 1570 mit einem Schwerpunkt zwischen 1420 und 1550. Da der Zeitraum vor der Beobachtung von Sonnenflecken liegt, lässt sich das Minimum nur indirekt durch den C13 Gehalt in Baumringen dieser Periode nachweisen. Die Periode ging mit einem besonders kühlen Erdklima einher, wie beim Maunderminimum.”
    That means, that the temperature minima of the LIA correlate with the minima of the activity of the sun.
    A similar paper can be found here for the alps for the whole holocene :
    http://alpen.sac-cas.ch/html_d/archiv/2004/200406/ad_2004_06_12.pdf
    „Noch spannender wird die Geschichte, wenn die von den Proben bestimmten Alter mit der 14C-Produktionsrate verglichen werden: Die Proben folgen auf Zeitabschnitte mit reduzierter 14C-Produktion in der Atmosphäre.Da die 14C-Produktion zu einem grossen Teil von der Sonnenaktivität abhängig ist – starke Sonnenaktivität bedeutet starkes Erdmagnetfeld und daraus folgernd schwache 14C-Produktion –, entsprechen offenbar die bestimmten «Holz- und Torfphasen » der Alpengletscher Zeiten mit einer bestimmten Sonnenaktivität.“
    All the glacier retreats during the holocene correlate with low C14-production.
    and we must not forget this:
    http://ff.org/centers/csspp/library/co2weekly/20060725/20060725_08.html
    “four glacial advances occurred between AD 1250 and 1810, coincident with solar-activity minima,” (for the tropical Andes).
    All these scientists speak about a sun-climate link. Isn’t that to much a coincidence?
    Best regards and have a noce day
    Eddy

  149. Bengt A (23:21:11) :
    So which one should we go with? The model of Pierce & Adams or the measurements of Svensmark?
    Other people looking at Forbush Decreases find no such effect, e.g. Sloan and Wolfendale.
    David Ball (01:35:11) :
    It is interesting that Svensmark rates high enough to be tested by the LHC.
    Not at all. Svensmark is only using some vacant facility there.
    Edouard (01:00:59) :
    All these scientists speak about a sun-climate link. Isn’t that too much a coincidence?
    Good for funding purposes.
    http://www.leif.org/research/Loehle-Temps-and-TSI.png
    compares solar activity [bottom] and global temperatures [top] the past 2000 years. As should be obvious there is no correlation between the two. You can find dips that coincide and dips that do not. An example of the latter is the two hundred year period centered on 660 AD where solar activity was even lower than during the Spoerer minimum, but it was warmer than today.
    Geoff Sharp (23:47:15) :
    I am sure I am not the only one here that think the Sun plays an important role in our climate.
    What a crock.

  150. David Ball (01:35:11) :
    It is interesting that Svensmark rates high enough to be tested by the LHC.

    Indeed, and we won’t resolve the different opinions voiced here until they get some LHC results to help prove disprove Svensmark.
    I would like to say a big than you to Lief for his comments! I feel I continue to learn a huge amount from his contributions and I deeply respect that people can have a discourse here and yet still respect different individual positions on an issue.
    This kind of healthy discussion is science.
    “Consensus” is NOT science.

  151. Jeremy (07:13:24) :
    “Consensus” is NOT science.
    But Consilience is. But that I mean that scientists by various methods and various data can eventually come to a ‘general agreement’ about a subject. There is, of course, no guarantee that that agreement reflects actuality, but it is a statement of ‘the best we can do with what we have got’. Purveyors of pseudo-science like to point to cases in the past where the ‘general agreement’ was wrong as an argument for the notions that any harebrained scheme will eventually be vindicated. This blog [and others] teems with examples of that.

  152. @Leif Svalgaard
    Thank you very much for taking your time to answer our questions. At least I know what you say an mean exactly! 😉
    I still don’t know if there is any mechanism that links the activity of the sun to “local” climate (in Europe for example).
    Maybe also that other strong climate mechanisms overway the influence of the sun sometimes, or maybe its just pure coincidence!
    The scientist Schlüchter also says this:
    http://www.sciencedaily.com/releases/2009/04/090430144535.htm
    “Southern Glaciers Grow Out Of Step With North”
    Maybe he doesn’t need any more funding?! 😉
    Thank you again 🙂
    Eddy

  153. Leif Svalgaard (06:51:59)
    Other people looking at Forbush Decreases find no such effect, e.g. Sloan and Wolfendale.
    No, I know. Sloan & Wolfendale didn’t find anything. Neither did Kristijansson (or almost nothing). But Svensmark did and that’s because he used a different methodology, e.g. he used the AERONET network for tracing aerosols. In my opinion this is a strong paper giving some real evidence for GCRs affecting aerosols and clouds. There’s nothing about correlations or sunspots or proxies or climate models or other ‘hot topics’ within this climate change debate. Just plain measurement data. And of course the conclusion, something that would be obvious for anyone looking at the plotted data, that GCRs affect clouds.
    I have not seen much critic on the internet other than ‘Slon & Wolfendale didn’t find anything’ or ‘Kristijansson didn’t find anything’. Maybe I am asking for too much, but it would be interesting to read a critical analysis of this very paper. What is wrong with Svensmarks Forbush paper?
    Svensmarks paper: http://wattsupwiththat.files.wordpress.com/2009/08/svensmark-forebush.pdf

  154. Lucy Skywalker (03:34:15) “Paul Vaughan, I’d love to see a post by you here.”
    Lucy, you might like figures 90-93 (pp.55-56) here:
    Semi, P.A. (2009). Orbital resonance and solar cycles.
    http://arxiv.org/ftp/arxiv/papers/0903/0903.5009.pdf
    Thanks to Geoff Sharp & G.E. Pease for drawing my attention to this article. I ran a time-normalized wavelet-power analysis yesterday and it was a cinch to see that Uranus periodically interrupts power at the Jupiter-Neptune frequency to produce the 11.1 year pattern that is visually evident in Earth-Moon-system angular momentum with respect to the center of the Sun.
    For anyone who is handy with the Helmholtz acoustic equation, this is simply U-(N-J):
    (11.86630899)*(164.888325) / (164.888325 – 11.86630899) = 12.78649873
    (12.78649873)*(84.05119028) / (84.05119028 + 12.78649873) = 11.0981628
    [in years]
    Note that this is equivalent to J-(N-U), which I have already addressed here:
    http://www.sfu.ca/~plv/VaughanPL2009_11.1aCycleSSD.htm
    People relying on traditional FFT analysis (& even those relying on non-time-normalized wavelet-power analysis) could easily miss the obvious.
    Anyone handy with acoustics will see the connection with the following:
    http://www.sfu.ca/~plv/absaayoy1111.1111.png
    Some clues are provided here:
    http://www.sfu.ca/~plv/DRAFT_VaughanPL2009CO_TPM_SSD_LNC.htm
    [There’s one loose end related to nutation that I haven’t gotten around to yet.]
    I again ask that people not misrepresent my work; it is true that I investigate synchrony. (People might find it amusing to see some of the e-mails I’ve received – my response generally goes like this: “I have made no such claim.”)

  155. Leif Svalgaard (18:37:00) :

    yonason (17:41:41) :
    the fact is that clouds do seem to empirically respond to cosmic rays.
    Have a link to that? That clouds correlate with the solar cycle does not show that they correlate with cosmic rays.


    http://www.tau.ac.il/institutes/advanced/cosmic/Conferences/2002_COSPAR_Houston/CosmRays_WheatPrice1.pdf
    See figure 2, and related discussion. Also, I already gave a link above. Here it is again.
    http://static.icr.org/i/articles/af/figure_2.jpg
    see also here, where it’s figure 3.
    http://physicaplus.org.il/zope/home/en/1105389911/1113511992_en/
    (same figure, but with whole paper)
    And clouds are inversely related to the solar cycle, in that when TSI is max, CRF, which is attenuated by solar wind, decreases. When solar wind is weak, CRF increases, and so do clouds, at least according to the data in the links above and below. It remains to elucidate the mechanism, or to disprove the apparent connection, but until then there’s sufficient reason to believe one exists. CERN is taking it seriously.
    http://public.web.cern.ch/public/en/Research/CLOUD-en.html
    As I said in another comment, if you can provide a better explanation, I’m all ears, as I’m sure these folks at NASA would be, as well.
    http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/20689/1/98-1743.pdf
    (NOTE – those guys also invoke the geomagnetic field, which ties it back in with the the initial topic of the thread.)

  156. One final comment. A “proof” based on the business model.
    The first stage is determining if there is a market, which, in the case of CRF and clouds, is equivalent to determining if there is a likely connection by showing correlation. That having been done, the company moves on to the next phase of bringing the product to that market, a much more expensive task, and a waste if there is no market.
    In the case of CERN, they are investing in the next phase, looking for a specific mechanism for the process. There would be no point if they didn’t already believe there was a connection.
    “The CLOUD experiment involves an interdisciplinary team of scientists from 18 institutes in 9 countries, comprised of atmospheric physicists, solar physicists, and cosmic-ray and particle physicists.”
    http://public.web.cern.ch/public/en/Research/CLOUD-en.html
    That’s expensive and time consuming. You don’t spend a lot of money and time looking for a mechanism for a process for which there is no evidence, and in which you don’t believe.
    As you well know, scientists and engineers don’t like wasting time or resources, especially when resources are limited and their careers depend on constantly focusing on what they think will be most important.
    If the correlation didn’t exist, and wasn’t compelling, they wouldn’t be doing those experiments.

  157. Paul Vaughan (10:42:33)
    U-(N-J):
    (11.86630899)*(164.888325) / (164.888325 – 11.86630899) = 12.78649873
    (12.78649873)*(84.05119028) / (84.05119028 + 12.78649873) = 11.0981628
    [in years]
    Note that this is equivalent to J-(N-U)

    Pure gold. Thank you.

  158. Leif Svalgaard (18:04:14) :
    Geoff Sharp (17:12:09) :
    Not by coincidence, we had 3 of the strongest grand minima in a row all separated by 172 years average. This is the lowest and longest period of solar activity during the 11,000 year Holocene record.
    Not at all:
    http://www.leif.org/EOS/2009GL039439.pdf
    Seems like the data in the graph has been detrended (5th°Poly), and clipped from detrending. Curious data…

  159. yonason (11:27:42) :
    http://public.web.cern.ch/public/en/Research/CLOUD-en.html
    You were claiming this was an LHC experiment, but in the left side bar it explicitly says that it is not an LHC experiment, which is what I pointed out.
    The money for CLOUD does not come from CERN or LHC, but from other institutions.
    That’s expensive and time consuming. You don’t spend a lot of money and time looking for a mechanism for a process for which there is no evidence, and in which you don’t believe.
    Do you know how many billions are spent on the AGW boondoggle?
    ET (13:41:16) :
    Leif Svalgaard (18:04:14) :
    Seems like the data in the graph has been detrended (5th°Poly), and clipped from detrending. Curious data…
    One has to do this as the background varies in different ways, so cannot be related to solar activity and has to be removed. Not curious, important.

  160. Paul Vaughan (15:44:22) :
    The observation is empirical, based on NASA Horizons output.
    The cyclomania is not in analyzing the cycles [we know the planets cycle], but in ascribing physical significance to these [which I know you don’t, so my comment was not directed at you].

  161. The cyclomania is not in analyzing the cycles [we know the planets cycle], but in ascribing physical significance to these [which I know you don’t, so my comment was not directed at you].
    Once again another emerging area of science, and there just maybe a physical link in the making, has the missing AM been found?
    http://www.landscheidt.info/?q=node/79

  162. Geoff Sharp (20:08:26) :
    Once again another emerging area of science, and there just maybe a physical link in the making, has the missing AM been found?
    since there is no missing AM [as we have shown you repeatedly] this is an emerging area of pseudo-science [although it has old roots].

  163. Paul Vaughan (00:19:27) :
    Geoff Sharp (20:08:26) ” http://www.landscheidt.info/?q=node/79
    Here’s a suggestion Geoff: Using the wiggle-matching technique you describe, compare your “planet” curve here …
    http://www.landscheidt.info/images/sun_planet_am.png
    …with Sun_SSB distance (instead of with Sun_SSB AM). [You’ll get a better shape match – could become one of the clues you’re looking for, in the mystery you’re building for us…]
    Cheers,
    Paul.

    I noticed the same shaped curves as well…I have compared them and the difference is still there. Difference is good.

  164. Leif Svalgaard (09:48:20)
    What is wrong with Svensmarks Forbush paper?
    That was discussed on WUWT a while ago [August?]

    In that thread from august I find two objections (Maybe this isn’t the place to continue that discussion, anyway here I go): 1. You seem to think that Svensmark used an invalid method to select Forbush events. 2. You seem to think that there a too few events in the analysis.
    Regarding number 1. The events were ranked according to the level of ionization of the lower atmosphere. I don’t see what’s wrong with that? Would GCRs count be a better method of selection or how are you supposed to do this?
    Regarding number 2. There are not that many events to analyze to start with and the Aeronet network has taken measurements only from 1998 onwards. In other scientific areas it is a valid method to look at strongest events while studying a phenomenon, e.g. you can study the strongest earthquakes, strongest hurricanes, strongest meteorite impacts etc. Why isn’t it OK to analyze the five strongest Forbush decreases?

  165. Bengt A (04:57:42) :
    Regarding number 1. The events were ranked according to the level of ionization of the lower atmosphere.
    A Forbush decrease is a DECREASE of the cosmic ray intensity, so one should select according to the lowest ionization, no?

  166. Leif Svalgaard (07:54:06)
    Yes, of course. Maybe I was a bit unclear, but that was what I tried to say. I don’t have access to the on line material for this article so I don’t know exactly how they did this. I hope they have looked for the lowest level of ionization and ranked that event the highest. Otherwise it makes no sense!
    Sorry to be going on about this, but I think this is Svensmarks strongest paper (for some of you maybe that doesn’t count for much). That’s why I am keen to understand the detailed critique.

  167. Bengt A (09:17:05) :
    I hope they have looked for the lowest level of ionization and ranked that event the highest. Otherwise it makes no sense!
    they did not. Anyway, the statistics is too poor for any firm conclusions. The strongest argument against the cosmic ray idea is that temperatures the past 60 years [since we have good GCR measurements] simply do not correlate with the GCR flux.
    If they did, we would not have this discussion.

  168. Is Motl talking about the same paper?
    http://motls.blogspot.com/2009/07/forbush-events-confirm-cosmoclimatology.html
    “An independent set of measurements has also shown that the amount of aerosols, i.e. potential nuclei of the new clouds, also decreases. All these “strength vs decrease” graphs display a lot of noise but the negative slopes are almost always significant at the 95% level (with one dataset being an exception, at 92%, which is still higher than the official IPCC confidence level that climate change is mostly man-made).”

  169. Leif Svalgaard (10:29:42)
    So I’ve read Svensmarks article once more and found this (my bold). Clearly says that they picked decrease/minimum ionization. (Still would like to know if there is a possibility to access the on line material for a non-subscriber?) :
    Table 1 lists the strongest FDs, 1987-2007. The first and second columns give the numerical order and the dates of the Forbush minima in the daily averaged GCRs . The third column is the strength of the FD, defined by the change in the ionization at the minimum, relative to a base period 14 days before the minimum. The value of the ionization decrease is normalized to be relative to the variation in ionization during the solar cycle at a latitude of 45 deg. On average the solar cycle variation in GCR ionization is 10 -15 % below 6 km altitude [Bazilevskaya et al., 2008].
    I understand your argument about a lack of correlation, but in my opinion this paper shows a cause and effect relation between GCRs and clouds whether or not we find correlations to climate/temperature. Maybe this effect is of minor importance, nevertheless it’s observable and thus need to be investigated.

  170. Bengt A (13:34:43) :
    in my opinion this paper shows a cause and effect relation between GCRs and clouds
    It shows at best a correlation. What weakens the argument for cause/effect is the lag between cause and effect. The causal link would be expected to be immediate, not delayed for days.

  171. One last note:
    Sunspot number may appear to be the “better” variable for some purposes, but aa is “better” for comparisons with a variety of terrestrial variables at interannual timescales.

Comments are closed.