Solar influence on glaciation in Greenland

  • In the GISP2 ice core, Greenland summit, Dansgaard – Oescheger (D-O) warm events 2 to 8 [1] are all associated with low 10Be events most likely caused by active solar magnetic activity. The simplest explanation is that warm D-O events are caused by an active Sun.
  • The mechanism by which active solar magnetic activity causes warming remains speculative but could be one of or a combination of the following 1) variable solar magnetic activity is accompanied by variable irradiance; 2) variable solar magnetic activity is accompanied by variable spectral output [2]; 3) cosmic rays affect global cloud cover [3]; 4) cosmic rays affect the position and activity of the polar vortex [4].
  • There are at least 20 low 10Be events between 15,000 and 38,000 y before present with an average frequency of 2000 y. Not all low 10Be events are associated with warm D-O events, hence, other variables are also important in determining whether or not high solar magnetic activity results in warming at the Greenland summit.
  • Glaciation on Earth is known to be modulated by the 41,000 y obliquity and 100,000 y eccentricity orbital cycles [5, 6]. And yet physical changes to Earth’s orbit are insufficient to create the changes in insolation required to trigger and end glacial periods. It is tentatively suggested that it is orbital cycles of the solar system acting on the Sun that causes solar variability that in turn modulates glaciation.


At the extremes, there are two competing hypotheses to explain climate change on Earth. “Climate Science” led by the International Panel on Climate Change (IPCC) advocates a number of manmade forcings, particularly the accumulation of greenhouse gasses in the atmosphere, that has led to global warming in the last 100 years. “Climate sceptics” are drawn to natural climate variability to explain most of the recent temperature trends where changes to the Sun over time is considered to be the main driver, leading to cyclical warming or cooling.

There are 34 years of satellite observations of the Sun and over this time the Total Solar Irradiance (TSI) changed very little (Figure 1). On this basis the IPCC effectively set the role of the Sun in modulating climate change to zero (Figure 2). This lacks credibility for three reasons 1) assuming that 34 years of observations may be applied to the Sun for “all time” is naive, 2) sceptics do not normally consider TSI to be the key variable (although it could be important) but instead advocate changes to solar magnetic activity and / or changes to the spectrum of energy leaving the Sun to be the key natural drivers of Earth’s climate [7] and 3) the IPCC approach ignores abundant evidence from geological records that changes in solar activity are linked to past climate change [8, 9, 10, 11].

Figure 1 Total solar irradiance [TSI] as measured by satellites [12]. The cyclical variation in TSI follows the sunspot cycle. Sunspots are dark patches on the face of the Sun and one could reasonably expect that a large number of sunspots mid-cycle could lead to solar dimming. In fact the opposite occurs, when the Sun is active mid cycle, TSI increases. Note that the total variation in TSI is < 2 units (0.15%) which is insufficient to cause significant temperature variance on Earth. While TSI may not have changed much, declining solar activity had seen the pressure of the solar wind drop by 20% by 2008: “This is the weakest it’s been since we began monitoring solar wind almost 50 years ago.” [NASA 13]. Note how in 2009, TSI (and sunspot numbers) reached satellite age minima between cycles 23 and 24. Historically, low solar activity of this sort has been accompanied by occasional extreme cold winters in N America and N Europe.

Figure 2 Summary of climate forcing according to IPCC AR5 [14]. Note that the only natural forcing considered is the Sun which is set to +0.05 Wm-2 (effectively zero). The IPCC have ignored dCloud, dVolcanos, dOcean currents, dSolar wind which are all known to be associated with natural climate variability, their imprint embedded within the historical climate record but ignored in forecasts.

Understanding climate history normally depends upon proxy records, i.e. direct information may not be available but secondary information may be used to estimate the required variable. For example the oxygen isotope (d18O) composition of carbonate fossils can be used to estimate the temperature of the seawater that the creatures lived in. d18O is a proxy for temperature. Cosmogenic isotopes provide a proxy for past solar magnetic activity. The two main cosmogenic isotopes used in climate history studies are 14C and 10Be. Both are made by the action of galactic cosmic rays on Earth’s upper atmosphere. 14C is made by the action of cosmic rays on N. It is naturally radioactive and decays to 14N with a half life of 5730 years.

10Be is made by the spallation of N and O by cosmic rays. It is also radioactive and decays to 10B with a half life of 1.4 million years. The longer half life makes 10Be useful for studying events up to 10 million years ago providing a tool to look back in time on a scale relevant to the ice age that began in the N hemisphere about 2.7 million years ago.

The key variable that controls the production rates of both 14C and 10Be is the intensity of cosmic ray bombardment of Earth and the key variable that controls that are fluctuations in the intensity of the Sun’s magnetic field. Hence the abundance of cosmogenic isotopes in geological layers provide a proxy for investigating past solar magnetic activity.

This post is the first in a mini series on cosmogenic isotopes and climate change and focusses on the extraordinary 10Be record of the GISP2 ice core from the Greenland summit. Hat tip to Mark BLR who posted links to the data in this comment on Tallblokes Talkshop a couple of weeks ago. Some of the questions being asked are:

  1. What causes glacial periods to end abruptly?
  2. What causes ice sheets to rebuild so soon after they collapse and melt?
  3. Why is the pattern of ice sheet collapse and growth correlated with Earth’s orbital cycles?
  4. What causes transient warm periods recorded in Greenland ice cores? (Dansgaard – Oescheger (D-O events)
  5. Why are historic climatic phenomena like the Indian ocean monsoon [9], the flora of North America [10, 11] and drift ice extent of the North Atlantic [8] all linked to cosmogenic isotope abundances?

The concentration of cosmogenic isotope 10Be in the GISP2 ice core (Greenland summit) shows a very high degree of covariance with temperature at many scales (Figure 3). There are five key variables that controls the concentration of 10Be in ice cores:

  1. The strength of the Sun’s magnetic field (the Solar wind)
  2. The strength of Earth’s geomagnetic field
  3. Variations in the intensity of incoming cosmic rays
  4. The snow / ice accumulation rate
  5. The radioactive decay of 10Be

Figure 3 The 10Be data come from NCDC NOAA [15] and the temperature and accumulation data come from ref [16]. I was surprised to find that the 10Be is not archived in The Greenland ice core gateway [17]. Note the high degree of correspondence between 10Be and temperature. However, the raw 10Be data plotted here cannot be used since it must be corrected for the rate of ice accumulation (left hand panel) as plotted in Figure 4.

On the time scale of GISP2 10Be (40,000 years) variations in Earth’s magnetic field are not considered to be significant [18]; variations in incident cosmic rays upon the solar system are not considered significant [18]; and variation due to radioactive decay on this time scale is also insignificant. Therefore, the 10Be profile through GISP2 may be interpreted in terms of variations in the strength of the solar wind and in the snow / ice accumulation rate.

The raw 10Be data plotted in Figure 3 presents an over simplified picture since the accumulation rate of ice must be taken into account. I was more than a little surprised to learn that the rate of ice accumulation at the Greenland summit increased by a factor of 5 from the glacial to interglacial Holocene [19] (Figure 3). This is about as counterintuitive as climate data comes. The story goes that a warmer climate brought more precipitation (snow fall) but also disproportionally larger rate of ice loss around the margin of the ice sheet. The ice accumulation profile can be split into three parts: 1) largely uniform accumulation rate from 40,000 to 15,000 y; 2) large swings and step change up in accumulation rate at the end of the glaciation and associated with the Younger Dryas, 15,000 to 10,000 y and 3) largely uniform accumulation during the Holocene 10,000 to 0 y. The accumulation data are based on counting and measuring layers and then applying an ice flow / de-straining model [19].

In order to remove the effect of variable accumulation rate the 10Be data have been normalised to a datum of 0.1 m ice / y (Figure 4). In intervals where accumulation rate is uniform, the normalisation does not significantly impact the structure of the data. But in the 15,000 to 10,000 y interval it imparts some irrational structure to the data, for example a spike towards a “quiet Sun” at the end of the last glaciation. And data for the whole of the Holocene are displaced towards higher 10Be which is counterintuitive.

Figure 4 10Be normalised to an ice accumulation rate of 0.1 m / y. The large variations in accumulation rate in the time interval 15,000 to 10,000 y and their attendant uncertainties has likely given rise to spurious 10Be structure in this part of the sequence and above.

The 10Be story does not support one of the hypotheses I was wanting to test in that there is no evidence for increased solar activity being responsible for the termination of the glacial period (Figure 4). However, I suspect uncertainty in the ice accumulation model may impart spurious structure to the normalised 10Be profile and that hypothesis remains untested. There is variability in 10Be in the 40,000 to 15,000 and 10,000 to 0 y intervals that carry a solar signal [18]. The former is examined in more detail below, while the latter will be the subject of a subsequent post.

The Dansgaard – Oescheger events

The D-O events are pervasive throughout Greenland ice cores and represent transient phases of warming at the summit of about 5˚C (Figure 5). These events are also recognised in Antarctic ice cores but are less obvious on the southern continent.

Figure 5  The temperature record from [16] and labelled D-O events 2 to 8 in blue from [1]. Low 10Be events 1 to 20 labelled in red. These low 10Be events would equate to an active solar magnetic field, shielding Earth from Galactic cosmic rays. It is possible that another 3 weak D-O events are present at 10Be events 7, 10 and 15. A quirk of XL means that it is only possible to plot 2 variables against time by placing time on the x-axis with the present day to the left. On this chart 10Be is normalised to the mean value for the interval = 0.23 m / y.

D-O events 2 to 8 are labelled on Figure 5 in blue [1] where it can be seen that each of these events is associated with spikes towards low normalised 10Be concentrations. Low 10Be, normalised for accumulation rate, would equate to an active solar wind shielding Earth from Cosmic rays. Note how prolonged D-O event 8 is associated with a prolonged period of active Sun.

It is difficult to escape the conclusion that warm D-O events are linked to periods of active solar activity and are most likely caused by it. However, there are many more 10Be events in this interval than there are D-O events. I count 20 in all, there could perhaps be a couple more, giving a mean frequency of 2000 y. Notably, there could be some additional “weak” D-O temperature events than counted before [1] at low 10Be events 7, 10 and 15. Thus it may be concluded that changes in solar magnetic activity on occasions impacted temperature at the Greenland summit and on other occasions did not. It appears that the impact of solar variability has diminished with time post 28,000 y, after D-O event 3. This suggests that solar activity is not the sole variable controlling temperature in Greenland.  N hemisphere ice sheet mass and extent, ocean currents, orbital parameters and solar variability may combine to control atmospheric circulation patterns.


While a connection between variable solar magnetic field strength and climate change on Earth has been recognised for a long while understanding the exact mechanism has remained elusive. This to a large extent is down to Satellite Man not having direct observations of the quiet Sun. With the Sun now entering a slumber, both solar and consequential climatic observations are being made for the first time. Europe has been recently exposed to extreme cold winter conditions reminiscent of the Little Ice Age and North America has recently been exposed to extreme cold caused by an expanded, mobile polar vortex. Several, perhaps dependent process, linking solar magnetic activity to climate change are summarised briefly below. They fall into two broad categories: 1) the impact of variable cosmic ray flux on the atmosphere and 2) other changes to the Sun that accompany variable geomagnetic activity.

Cosmic ray flux and cloud formation. Svensmark [3] has proposed that increased cosmic ray flux may increase the nucleation rate of low clouds leading to a cooling of The Earth. This theory has been tested at CERN but remains largely unproven. We will need to wait and see if increased cosmic ray bombardment that is to be expected with the current solar slumber leads to an increase in cloud cover.

Cosmic ray flux, impact on stratosphere and polar vortex.

It was shown that long-term oscillations of the amplitude and sign of Solar Activity/Galactic Cosmic Rays effects on troposphere pressure at high and middle latitudes are closely related to the state of a cyclonic vortex forming in the polar stratosphere. [4]


The timing of the pine minima is correlated with a series of quasi-periodic cold intervals documented by various proxies in Greenland, North Atlantic, and Alaskan cores and with solar minima interpreted from cosmogenic isotope records. These events may represent changes in circumpolar vortex size and configuration in response to intervals of decreased solar activity, which altered jet stream patterns to enhance meridional circulation over eastern North America. [10]

Solar magnetic activity and energy spectrum.

Although sunspots themselves produce only minor effects on solar emissions, the magnetic activity that accompanies the sunspots can produce dramatic changes in the ultraviolet and soft x-ray emission levels. These changes over the solar cycle have important consequences for the Earth’s upper atmosphere. [NASA 13]


SORCE observations made during the decline of solar cycle 23 reveal a remarkably strong decrease in mid-ultraviolet flux, some four to six times greater than previous spectral irradiance reconstructions.


If the updated measurements of solar ultraviolet irradiance are correct, low solar activity, as observed during recent years, drives cold winters in northern Europe and the United States, and mild winters over southern Europe and Canada, with little direct change in globally averaged temperature [2 Met Hadley]


  • Warm D-O events at the Greenland summit are associated with low 10Be events most likely caused by high levels of solar magnetic activity. In other words, warm D-O events are linked to an active Sun.
  • To explain away the association of 10Be and temperature at the Greenland summit by some other process we would need to invent a new ice accumulation rate model that cancelled out the 10Be concentration profile.
  • An active Sun shields Earth from galactic cosmic rays and this can affect Earth’s climate in a number of benevolent ways. Our Sun is now in the process of going very quiet and that benevolence may abandon N Europe and N USA intermittently during winters to come for perhaps another 30 years or more.
  • There are two competing theories to explain climate change on Earth. One predicts that Earth will become monotonically warmer with increasing green house gas emissions. The other predicts climatic fluctuations linked to natural variability of the Sun’s activity where low solar activity, like we have now, may cause periodic extreme cold conditions in N America and N Europe.
  • Globally averaged lower troposphere temperatures over the last 100 years is not a sound metric for describing climate change on Earth.


I have a PhD in isotope geochemistry, but this does not make me an expert in the interpretation of cosmogenic isotope data from ice cores, although I have extensive experience in the interpretation of isotope geochemical data from bore holes. I have not found any reference linking directly low 10Be to the D-O events. If there are any papers or blog articles out there please let me know and I will ensure credit is given where credit is due.

The original paper by R. C. Finkel and K. Nishiizumi (1997) [18] does recognise the connection between 10Be and d18O variance but seemed reluctant to draw the conclusion that D-O events are caused by solar variability.


[1] Richard B. Alley et al (2010) History of the Greenland Ice Sheet: paleoclimatic insights. Quaternary Science Reviews 29 (2010) 1728-1756

[2] Sarah Ineson et al (2011) Solar forcing of winter climate variability in the Northern Hemisphere Nature Geoscience PUBLISHED ONLINE: 9 OCTOBER 2011 | DOI: 10.1038/NGEO1282

[3] Svensmark et al (2007) Experimental evidence for the role of ions in particle nucleation under atmospheric conditions Proc. R. Soc. A (2007) 463, 385–396

[4] S Veretenenko and M Ogurtsov (2013) The stratospheric polar vortex as a cause for the temporal variability of solar activity and galactic cosmic ray effects on the lower atmosphere circulation Journal of Physics: Conference Series 409 (2013) 012238

[5] Clive Best: Phenomenology of Ice Ages

[6] Euan Mearns: The Ice Man Cometh

[7] EDOUARD BARD et al (2000) Solar irradiance during the last 1200 years based on cosmogenic nuclides Tellus (2000), 52B, 985–992

[8] Gerard Bond et al (2001) Persistent Solar Influence on North Atlantic Climate During the Holocene VOL 294 SCIENCE

[9] U. Neff et al (2001) Strong coherence between solar variability and the monsoon in Oman between 9 and 6 kyr ago NATURE | VOL 411 | 17 MAY 2001

[10] Debra A. Willard et al (2005) Impact of millennial-scale Holocene climate variability on eastern North American terrestrial ecosystems: pollen-based climatic reconstruction Global and Planetary Change 47 (2005) 17–35

[11] Andre E. Viau  Widespread evidence of 1500 yr climate variability in North America during the past 14 000 yr Geology; May 2002; v. 30; no. 5; p. 455–458

[12] WoodForTrees

[13] NASA The Sunspot Cycle

[14] IPCC AR5 Summary for policy makers

[15] NCDC NOAA 10Be concentration (103 atom/g) in GISP2 ice, 719 – 2253 m

[16] NCDC NOAA GISP2 Ice Core Temperature and Accumulation Data

[17] NCDC NOAA The Greenland Ice Core gateway

[18] Finkel, R.C., and K. Nishiizumi. (1997). Beryllium 10 concentrations in the Greenland Ice Sheet Project 2 ice core from 3-40 ka. Journal of Geophysical Research 102:26699-26706.

[19] Cuffey, K.M., and G.D. Clow (1997), Temperature, accumulation, and ice sheet elevation in central Greenland through the last deglacial transition. Journal of Geophysical Research 102

This entry was posted in Climate change and tagged , , , , , , , , , , . Bookmark the permalink.

21 Responses to Solar influence on glaciation in Greenland

  1. Roger Andrews says:

    Eaun: A few preliminary questions and observations. Maybe more to follow.

    Have you plotted up the 386ka GRIP 10Be record? If not I might give it a shot, but it will be a pain because the 10Be numbers are given by depth and the depth/age data are in a separate file.

    On your comment: “I was more than a little surprised to learn that the rate of ice accumulation at the Greenland summit increased by a factor of 5 from the glacial to interglacial Holocene [19] (Figure 3). This is about as counterintuitive as climate data comes. The story goes that a warmer climate brought more precipitation (snow fall) but also disproportionally larger rate of ice loss around the margin of the ice sheet.” This will cause the ice sheet to get higher and higher and narrower and narrower, which is actually not as bizarre as it sounds. Something like it has been happening at Kilimanjaro, although the cause wasn’t warming. And a block of ice like the one shown in the picture can’t begin to expand again no matter how much it snows or how cold it gets.

    Temperature records show what appears to have been a mini-DO event in the North Atlantic starting in 1920, with maybe another beginning around 1995. Note that I’ve plotted the time scale backwards to match the backwards time scales on your graphs 😉

    • Euan Mearns says:

      The snow accumulation pattern is bizarre. You are right that the tendency would be to get higher and thinner, counteracted by a great increase in ice flow and flux of “fresh water” going through the ice sheet, which makes me suspicious of the dynamic flow part of the ice accumulation model.

      I had a look at the 386K GRIP record which is totally fragmented the way it is archived – you get the feeling they don’t want folks like us to use it 😉 I decided life was too short to try and fix the data, and so if you were willing to have a go, I’d love to see it.

      Tinypic was down for maintenance.

    • Roger Andrews says:

      I downloaded the GRIP 10Be data and this was what I got:

      I don’t think I’m going to bother to go any farther.

      (Hope tinypic is back up)

  2. A C Osborn says:

    Euan, isn’t it odd the way the Temperature has smoothed out over the last 25,000 years?

    • Euan Mearns says:

      It’s not so mush odd as a good observation. Before 28,000 y you had a system (on the Greenland summit) flipping between two states – cold and less cold – I’d argue in part controlled by Solar activity. And then it more or less stops flipping settling on the cold state. And then it does couple of whip saw flips in the Bølling-Allerød less cold interstadial and the Younger Dryas cold before settling on the less cold state of the Holocene – where we still are today.

  3. Roger Andrews says:


    I don’t know how this might affect your conclusions, but the preamble to the GRIP 10Be data package states that “the results confirm that the first-order origin of 10Be concentration variations is changes in precipitation rate associated with different climate regimes.” Solar, geomagnetic and (possibly) cosmic ray impacts on 10Be are observed “after taking into account variable snow accumulation effects”.

    Which raises the question, do the published 10Be records take snow accumulation effects into account? My guess is that they don’t. It’s certainly hard to relate the GISP2 10Be record as it stands purely to changes in orbital obliquity and eccentricity. Explaining the 0-50,000 year segment of the Taylor Dome 10Be record in terms of Milankovitch cycles is even harder:

    • Euan Mearns says:

      Roger, I think you need to set aside 20 mins to read the post in detail. Figure 3 plots the raw 10Be concentrations in ice. Figure 4 plots the data corrected for ice accumulation rate. The original paper describing this has charts that show what my charts show – just not so pretty. I’ll send you a copy of the paper.

      [18] Finkel, R.C., and K. Nishiizumi. (1997). Beryllium 10 concentrations in the Greenland Ice Sheet Project 2 ice core from 3-40 ka. Journal of Geophysical Research 102:26699-26706.

      The paper points out that the first order structure is down to ice accumulation rate, but the residual structure is a solar / correlated with climate effect. But they fall short of saying warm D-O events are linked to an active Sun.

      • Roger Andrews says:

        Euan: I apologize for not picking up on that. But now that I’ve looked into it I’m not sure that normalizing to 0.1m/yr is going to give the right answer. Back after I’ve read the paper 🙂

        • Euan Mearns says:

          Roger, the units “10^3 atoms / g” are somewhat arbitrary and controlled in this case by two main processes 1) rate of production of 10Be in the stratosphere and 2) the rate of snow / ice accumulation. Normalising to a uniform rate of ice accumulation removes the former effect leaving a residual that can be attributed to the latter. Where the rate of ice accumulation is <0.1 m / y concentrations are adjusted down to account for slow accumulation that would result in spurious high concentrations and vice versa. You can in fact normalise to any accumulation rate – it is the trend in normalised data that is important.

        • Roger Andrews says:


          Thanks for the Finkel paper. After reading it I still wasn’t sure how they removed the ice accumulation signal, so I decided to remove it the way I would do it, i.e. by defining a regression relationship between 10Be and ice accumulation and using this relationship to adjust the 10Be records.

          First I adjusted the ice accumulation and 10Be records to common reading times. Then I calculated running correlation coefficients for 3,000 years of data, which is about the length of the average DO event, and plotted R^2 against time. The first graph shows the results and the second plots the ice accumulation and 10Be data at common reading times (sampling points) for reference:

          For the first ~10,000 years BP there is no correlation between 10Be and ice accumulation. Between 10,000 and 16,000 years there is a strong one. Between 16,000 and 20,000 years there is again no correlation. After 16,000 years we lurch from high levels of correlation to low levels and back again. I can’t think of a physical mechanism that could explain these effects and suspect they’re an artifact of the data. Anyway, I made no further effort to “correct” the 10Be records.

          An argument can in fact be made for not correcting them at all. I’ve plotted graphs of 10Be, 18O, potassium, sodium, calcium, magnesium, chlorine, NO3, SO4 and methane for GISP2 and they all show matching DO events (the example graph below plots 10Be, 18O and potassium). There are no good CO2 records for Greenland cores – the values are reportedly contaminated by high carbonate levels – but the near-exact CO2-18O match in Antarctic cores suggests that Greenland CO2 would match the DO events too, other things being equal. So if the GISP2 10Be record is significantly distorted by ice accumulation variations then so are all the records that match it, and if we correct one of them we have to correct the lot.

  4. G. Watkins says:

    Very impressive – thanks for your hard work and lucid writing.

  5. Bernard Durand says:

    Interesting. However, I did’not see any 10Be measurements for the 1860-2012 warming period?.

  6. clivebest says:


    Very Interesting post!

    I only stumbled on these D-O climate events about a month ago, and it certainly looks like something dramatic is causing them. There is a good correlation with 10Be which is a proxy for solar activity. I think the only physical model that makes sense to explain them is the possible cosmic ray seeding of clouds. The CERN experiment is crucial to decide this. This is also very relevant for the current drop in solar activity and the origin of the little ice age.

    Note: The GISS people have tried to play down the role of 10Be implying that the signal itself can be effected by climate see: This may have been in response to a paper claiming a 10Be signal in sediments showing 100,000y cycles

    Ian Wilson claims the D-O events are instead caused by the 1870 year cycle of lunar extreme tides – but I haven’t yet understood how he manages their recalibration to 1470y cycles – ” Are the Dansgaard-Oeschger (D-O) Warm Events driven by Lunar Tides?

    Yet another mystery which climate models fail to explain.

    • Euan Mearns says:

      Clive, Standard operating procedure in Climate Science is to debunk everything that does not comply with gospel. Earth’s climate is sole influencer on 10Be, not cosmic rays and the solar magnetic field. I’m doing a second post on climate and 10Be in the Holocene next week, Roger posted link to one extremely useful paper (thank you Roger).

      The D-O and Bond cycles appear to be more strongly expressed in N hemisphere and there must be a clue to understanding the origin in that. I think I asked on your own blog, does the polarity of Earth’s dipole have an effect on shielding cosmic rays?

      I grabbed this chart from WUWT the other day which I think is fabulously enlightening – sorry I didn’t store the link. The D-O and Bond cycles are quasi periodic. The average may be 1470 y but for Bond it is effectively 1500±500 y.

      Cosmic rays seeding clouds is actually bottom of my list. I’m more drawn to spectral shifts, TSI shifts outside of the 34 years of observations and given very recent history, stratospheric impacts on the size and mobility of the polar vortex. Just watched “The Day after Tomorrow” for about the 10th time 😉

      • Clive Best says:

        “Standard operating procedure in Climate Science is to debunk everything that does not comply with gospel.”
        Yes unfortunately this seems to be true !

        The last magnetic inversion was 700,000 years ago so it is hard to see how they could be responsible. Whatever model we come up with must explain enhanced energy flux to northern latitudes. That could be ocean currents, higher insolation, enhanced gravitational tidal energy or reduced albedo.

        The “day after tomorrow” is the most ludicrous nonsense I have ever seen.

        Along the same lines I think we need a new parliamentary select committee on “Catastrophic Asteroid Impacts”. UN Scientists are now 100% convinced that major meteor impacts will happen in the future causing 100m high tsunamis and year long nuclear winters. It would be foolish not to act now as an insurance policy to such possible threats. To safeguard UK citizens we should therefore construct underground cities powered by geothermal power. Patent pending……

        • Euan Mearns says:

          Clive, the central plank of the story line in “The Day After Tomorrow” is very rapid climate change linked to the Gulf Stream closing down, based on the DO observations in Greenland ice cores and the apparent fact that Mammoths froze to death almost instantly. Bond interpreted his events as being due to the Labrador current periodically truncating the Gulf Stream. Bond counted 8 such naturally occurring events in The Holocene. The ludicrous thing about the movie (and climate science) is that when nature repeats it is going to be blamed on a few ppm of CO2 and not massive unstoppable natural forces.

  7. Roger Andrews says:

    I’ve been looking at ice core records again. I find that the following features are common to just about all of them:

    When 18O decreases (i.e. when we have cooling):

    * Atmospheric gas concentrations (CO2, CH4) decrease.

    * Terrestrial element concentrations (K, Mg, Ca, Cl, also NO3 and SO4) increase.

    * Cosmogenic isotope concentrations (10Be, also 14C) increase.

    * Dust increases.

    And vice versa when we have warming.

    How to explain these features? Thinking out loud here:

    The decreases in CO2 and CH4 can probably be explained by the impacts of cooling on the carbon cycle. The increases in K, Mg etc. are presumably a result of the increased dust in the atmosphere (it’s hard to see what else could have caused them). But what about the 10Be? It tracks the K, Mg etc. concentrations, so presumably it came from the dust too.

    And where did the dust come from? Either the cooling caused deserts to expand or the wind to blow more strongly, or a bit of both. So maybe the 10Be in the ice cores came from the Sahara, or the Gobi. And maybe it had been lying around there for a while. Or maybe it hadn’t.

    I looked briefly into the GISP2 ice accumulation rates. They were estimated by correcting the 18O record “for changes in density with depth in the firn using a Herron-Langway densification model and…. for flow-induced thinning using a 1-dimensional Dansgaard-Johnsen model” in the deeper layers. The results strike me as implausible (accumulation rates can be either positively or negatively correlated with 18O) but maybe that’s just because I don’t know enough about glaciology.

    So what’s the bottom line? Dunno. Is there an isotope chemist in the house? 😉

    • Euan Mearns says:

      Roger, good observations. I didn’t go into the depositional mechanics in the post since it was already long and complex enough. And as mentioned I am no expert on cosmogenic isotopes. But here goes…

      Be is an alkaline earth and so will behave like Mg and Ca etc – soluble and will form compounds easily. Its is also extremely light – 10 amu for the cosmogenic isotope. It somehow finds its way from stratosphere to troposphere where it will quickly be absorbed into clouds and dumped as rain or snow as a Be complex – chloride, sulphate, nitrate whatever. Most will be deposited on the oceans, some on land and some as snow on ice caps. It will likely stay in solution in the oceans for thousands of years. That which falls on land will I guess be quickly washed into water courses and will too end up in the oceans.

      The primary control on the concentrations of all the things you mention is accumulation rate. So they are all correlated because the concentrations are in first order, all controlled by the same variable – or are you looking at normalised data? I have heard an interpretation before that it was much stormier during the glacial events, and so higher concentrations of dust may be explained that way.

      You are right to draw attention to the complexities. What to believe? You could be like NASA and discount the data as unusable. I am inclined to believe the data for two reasons. 1) there is pretty good concordance between the 10Be and 14C records and since these follow totally different geochemical paths it provides quite strong evidence that both signals are underpinned by intensity of cosmic rays and 2) in the Holocene a number of climatic variables, like the pollen record of N America, follows the 10Be signal with variable degrees of conformance.

  8. Roger Andrews says:


    Thanks for your considered response. I have a couple more incisive observations/questions for the isotope doctor, if he’s in 😉

    You say “there is pretty good concordance between the 10Be and 14C records and since these follow totally different geochemical paths it provides quite strong evidence that both signals are underpinned by intensity of cosmic rays.” No argument there.

    However, if 10Be is a CR proxy and non-cosmogenic elements like K, Ca and Mg aren’t then 10Be should vary independently of K, Ca and Mg in the ice core records. But they all march pretty much in lockstep in GISP2. I’ve been trying to think of a mechanism that would explain this and still leave 10Be as a valid CR proxy, but without much success. The only explanation that seems to fit is that the 10Be, K, Ca, Mg etc. concentrations are all controlled by some external factor that isn’t related – or at least not directly – to CR activity.

    As to what this external factor is, according to Yiou et. al. it’s precipitation. They state that the results from the GRIP core (drilled next door to GISP2) “confirm that the first-order origin of 10Be concentration variations is changes in precipitation rate associated with different climate regimes.” (Link to abstract below but the article is paywalled, so I can’t find out what their rationale is, but “confirm” is a word you don’t see that often in scientific papers so presumably it’s a good one. On the other hand the paper was published in 1997 and things might have changed since then.)

    But if we accept for the moment that changes in precipitation really are the cause of the fluctuations in 10Be, K, Ca, Mg etc then 18O, which is inversely correlated with all of them, becomes a proxy for precipitation. Is this isotopically possible?

    Now the plot thickens. If 18O is a proxy for precipitation how do we explain the LR04 sea bed sediment 18O values, which match the ice core values, but which were measured on marine organisms that presumably aren’t affected by precipitation? Or are they?

    Getting in a little deep here. Please feel free to shut me up if you think I’m overdoing it. 🙂

Comments are closed.