Study: Siberian permafrost has been warming for 7000 years

Winters in Siberian permafrost regions have warmed since millenia

For the first time, researchers have reconstructed the development of winter temperatures in Russia’s Lena River Delta based on old ground ice data

From the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research

Exposed ice wedges at the coast of the Siberian island Muostakh. With this picture in mind, one can understand, why early researchers thought ice wedges could be nothing else than buried glaciers. Credit Photo: Thomas Opel, Alfred-Wegener-Institut

Exposed ice wedges at the coast of the Siberian island Muostakh. With this picture in mind, one can understand, why early researchers thought ice wedges could be nothing else than buried glaciers. Credit Photo: Thomas Opel, Alfred-Wegener-Institut

For the first time, researchers at the Alfred Wegener Institute have successfully decoded climate data from old permafrost ground ice and reconstructed the development of winter temperatures in Russia’s Lena River Delta. Their conclusions: over the past 7,000 years, winter temperatures in the Siberian permafrost regions have gradually risen. The study was published yesterday on Nature Geoscience‘s website.

You won’t find any glaciers in Russia’s Lena River Delta. Unlike in Antarctica or Greenland, in the Siberian tundra ice doesn’t form above ground on hillsides or elevated plains. Rather, it forms directly underground as ice wedges.

“Ice wedges are a typical feature of permafrost regions. They are formed when the permanently frozen soil contracts in response to intensively cold winter temperatures, causing it to crack. When the snow melts in spring, the melt water fills these cracks. Since the ground temperature is roughly minus ten degrees Celsius, the water refreezes immediately. If this process repeats itself winter after winter, over the decades and centuries an ice body shaped like a giant wedge is formed,” explains Dr Hanno Meyer, a permafrost researcher at the AWI Potsdam and first author of the study.

With a depth of up to 40 metres and a width of up to six metres, the ice wedges of the Siberian Arctic may not be as physically impressive as Antarctic glaciers. However the ice wedges, some of which are more than 100,000 years old, store climate information in much the same way, allowing scientists to investigate them using glacier research methods. “The melt water always comes from the snowfall of a single winter. Therefore, when it freezes in these frost cracks, information on the winter temperatures in that specific year is also preserved. We have now succeeded for the first time in using oxygen isotope analysis to access the temperature information stored in the ice and compile it into a climate curve for the past 7,000 years,” states AWI researcher and co-author Dr Thomas Opel.

The new information represents the first well dated winter-temperature data from the Siberian permafrost regions and indicates a clear trend: “Over the past 7,000 years, the winters in the Lena River Delta have steadily warmed – a trend we haven’t seen in almost any other Arctic climate archive,” says Hanno Meyer. As the permafrost expert explains, the likely reason is: “To date, primarily fossilised pollen, diatoms and tree rings from the Arctic have been used to reconstruct the climate of the past. But they mostly record temperature information from the summer, when the plants grow and bloom. Ice wedges are among the few archives that can exclusively record winter data.”

Further, the new data will allow the researchers to fill an important gap: “Most climate models indicate a long-term cooling in the summer and long-term warming in the winter for the Arctic over the past 7,000 years. But until now, there has been no temperature data to support the second claim, essentially because the majority of climate archives record information from the summer. Now we can finally demonstrate that ice wedges contain similar winter-temperature information as predicted by climate models,” says AWI modeller and co-author Dr Thomas Laepple.

At this point, the researchers can’t exactly determine yet how many degrees the Arctic winters have warmed. As Thomas Opel explains, “The results of the oxygen isotope analysis can only tell us whether and how the isotopic composition has changed. If it rises, it indicates a warming. But the exact extent of warming is something we can’t yet make a statement on.”

Nevertheless, the researchers found clear indications for the causes of this warming.

According to Hanno Meyer: “The curve shows a clear partitioning. Up to the dawn of industrialisation around 1850, we can attribute the development to changes in the Earth’s position relative to the sun. In other words, the duration and intensity of the solar radiation increased from winter to winter, causing temperatures to rise. But with industrialisation and the strong increase in the emissions of greenhouse gases like carbon dioxide, this was supplemented by the anthropogenic greenhouse effect. Starting at that point, our data curve shows a major increase that clearly differs from the gradual warming in the previous phase.”

In a next step, the researchers will investigate whether the same indicators for a gradual rise in winter temperatures in the Arctic can also be found in other permafrost regions around the globe. As Thomas Opel elaborates: “We already have data from an area 500 kilometres east of the Lena River Delta that supports our findings. But we don’t know how it looks for example in the Canadian Arctic. We suppose the development was similar there, but don’t yet have evidence to back up that assumption.”

The data for the new Lena River Delta temperature curve comes from 42 ice samples, which AWI researchers collected over the course of several expeditions from 13 ice wedges that the river had uncovered during flooding. “For the purposes of the study, we only included samples for which we could clearly determine the age. Fortunately, for ice wedges this is relatively simple as a large number of plant remains and other organic material enters the ground ice during snow melt- and we can use the radiocarbon method to precisely determine the age of this material,” says Hanno Meyer.

###

The study was supported by the Alfred Wegener Institute, the German Research Foundation and the Initiative and Networking Fund of the Helmholtz Association (grant VG-900NH).

The paper is going to be published on the 26th of January 2015 on Nature Geoscience‘s website under following title: Hanno Meyer, Thomas Opel, Thomas Laepple, Alexander Yu Dereviagin, Kirstin Hoffmann und Martin Werner (2015): Long-term winter warming trend in the Siberian Arctic during the mid- to late Holocene, Nature Geoscience, Vol 8, DOI: 10.1038/ngeo2349

76 thoughts on “Study: Siberian permafrost has been warming for 7000 years

  1. Let’s also hear from the scientists in Siberia monitoring the situation. What do they say?

    “Indeed above at the surface it has gotten warmer, but that’s just part of a normal cycle. The permafrost is rock hard, And that is how it is going to stay. There’s no talk of thawing.” Michali Grigoryev
    http://notrickszone.com/2012/11/19/russian-arctic-scientist-permafrost-changes-due-to-natural-factors-its-going-to-be-colder/

    “It seems that the permafrost should be melting if the temperature is rising. However, many areas are witnessing the opposite. The average annual temperature is getting higher, but the permafrost remains and has even started to spread. Why? An important factor is the snow cover. Global warming reduces it, therefore making the heat insulator for the permafrost thinner. Then even weak frosts are enough to freeze the ground deeper below the surface.”

    Nikolai Osokin is a glaciologist at the Institute of Geography, the Russian Academy of Sciences.

    http://en.rian.ru/analysis/20070323/62485608.html

    “The Russian Academy of Sciences has found that the annual temperature of soils (with seasonable variations) has been remaining stable despite the increased average annual air temperature caused by climate change. If anything, the depth of seasonal melting has decreased slightly.”

    “This is just another scare story . . . This ecological structure is balanced and is not about to harm people with gas discharges.”

    Vladimir Melnikov is the director of the world’s only Institute of the Earth’s Cryosphere. The Russian Academy of Sciences’ Institute is located in the Siberian city of Tyumen and investigates the ways in which ground water becomes ice and permafrost.

    “The boundaries of the Russian permafrost zone remain virtually unchanged. At the same time, the permafrost is several hundred meters deep. For methane, other gases and hydrates to escape to the surface, it would have to melt at tremendous depths, which is impossible.”
    Yuri Izrael, director of the Institute of Climatology and Ecology of the Russian Academy of Sciences.

    http://en.rian.ru/analysis/20050822/41201605-print.html

    • If you go to the reference paper sourced in the article above you can see the actual data.

      surprisingly, the temperature curve shows a hockey stick of warming over the last hundred years!!!

      WOW! who could have imagined!

      • Now 1850 is being dubbed the start of industrialization. And here I thought it was the end of the LIA! See how easy it is to get confused by labels?

  2. Well their figure three shows a very clear hockeystick.
    Are those black dots the actual readings? The jump before the last 3 is huge compared to everything else.
    And then it starts to drop again?

    How does this δ18O proxy work? And is it affected by nuclear testing?

    • I downloaded the data and would calibrate the do18 isotopes to temperature in this manner.

      First excluding two datapoints omitted in the study.

      And then including the two (which don’t appear to be that out of place, although the one at 2100 years ago was near the surface and may have been contaminated.)

      • So how do they know that the recent data isn’t reflective of the same phenomenon causing the data for 2100 years ago – that they discarded to fit their conception?

      • I would suggest that ten data points over the period 7500 to 2500 B2000 is hardly a valid statistical sample to prove anything

      • ” the one at 2100 years ago was near the surface and may have been contaminated”. Looks very much like they cheated. Can anyone explain how, in a process that forms ice in layers, ice from 2000 years ago can be nearer the surface than later ice?

      • [Bill Illis] “And then including the two (which don’t appear to be that out of place, although the one at 2100 years ago was near the surface and may have been contaminated.)”

        Especially in a study with this few data points, forthrightness needs to happen. If they had only the 35 datapoints I count (including the two), if they discard ANY, they need to be very clear on WHY they excluded any data points.

        Excluding ANY that show a peak higher than the ones they left in at the end – the ones that made their argument about the hockey stick – they have no business doing such a thing without explanation. And darned good reasons for the exclusion.

        With the big DOWNward spike at about 6100 ys, this number of data points seems to have a lot of WOW in it to begin with. Leaving in a LOW spike and excluding a HIGH spike – not kosher.

        Not only that, but the Y-axis is in “Winter Degrees C”, yet in the text as presented, the authors state clearly that,

        As Thomas Opel explains, “The results of the oxygen isotope analysis can only tell us whether and how the isotopic composition has changed. If it rises, it indicates a warming. But the exact extent of warming is something we can’t yet make a statement on.”

        Since “Winter Degrees C” MEANS “the exact extent of warming” relative to SOMETHING (like 0°C), if Opel is correct in his statement, what in the devil are they putting temperature NUMBERS on their data for?

        What are the numbers supposed to represent if not “Winter Degrees C”?

        I reference everyone to the first comment above, by Ron C. for another take on this from the scientists who deal with this stuff all the time.

    • How much data do they have? http://www.nature.com/ngeo/journal/vaop/ncurrent/carousel/ngeo2349-f3.jpg
      It looks like a 100 year gap in the period that has supposed to have warmed faster than any other period in human history.

      So they have more cracks when it gets very cold suddenly ie samples are biased towards years with a cold Autumn. Warming pushes snow and less compacted ice out of cracks so older data is most likely to retain ice that was from melting snow above a crack than snow falling into a crack. Basically, older samples have signal from spring rains and recent ones from Autumn snow.

      Any suspicions that in 100 years time that those recent samples would have given a lower result?

    • in their abstract they speak of the hockey stick nature to the results:

      The isotope values, which reflect winter season temperatures, became progressively more enriched over the past 7,000 years, reaching unprecedented levels in the past five decades

  3. Yet another study with a hammer & a whole bunch of nails! What have the warmists got? Puter models!

    • “According to Hanno Meyer: “The curve shows a clear partitioning. Up to the dawn of industrialisation around 1850, we can attribute the development to changes in the Earth’s position relative to the sun. In other words, the duration and intensity of the solar radiation increased from winter to winter, causing temperatures to rise. But with industrialisation and the strong increase in the emissions of greenhouse gases like carbon dioxide, this was supplemented by the anthropogenic greenhouse effect. Starting at that point, our data curve shows a major increase that clearly differs from the gradual warming in the previous phase.”

      Those hammer and nails folks are never wrong, right?

      • Look at Bill Illis’ post above. To conclude that the recent spike is unusual requires hammering down the spike 2000 years ago.

      • So, beginning conveniently at the end of the LIA the temperatures have shown an increased rise.
        Seems reasonable to me.
        Human CO2, according to CAGW folks, hasn’t had much of an impact until around 1950, so unless there is a significant change in the rate between 1850 to 1950 and the rate between 1950 and present, it would appear the anthropogenic impact is not discernable.

      • Phlogiston, it seems justifiable to take out the outlier as it was near the surface and suspected of contamination.
        It wasn’t removed just because it was an outlier.

      • Eliminating the datapoint showing an increase 250 years ago removes the trend of a natural increase in the last 500 years.

      • M Courtney

        So why don’t they delete the year zero outlier? That one looks very suspicious to me…… ;-)

        R

      • A mighty good point.

        I’d also like to know:

        WHERE IS THE MWP?

        2000 ya could be the Roman Warm Period. But if THAT is there (perhaps the spike/nail), then so should the MWP be in it.

        Like Mann’s erasure of the MWP and the RWP and the LIA, this one doesn’t appear to have any of those – except the one they left out.

  4. First sentence of the abstract
    Relative to the past 2,000 years, the Arctic region has warmed significantly over the past few decades

    • Not in the local inatrument record. The abstract first sentence just shows that the methodology is flawed. And the authors knew, or should have known. See main comment below.

      • Yes, but in a publish-or-perish world and in which they need to know which side their bread is buttered on, they HAD to both publish SOMETHING and to also wrap it around CAGW or risk losing the money that is out there for the picking.

  5. “… around 1850, we can attribute the development to changes in the Earth’s position relative to the sun…”

    Hmmm…they are moving the anthropo effect back a hundred years! And they support the models on temps over the past 7000yrs….so they must support the models since 1850 showing a big hockey stick? There likely is a “firn” effect with this too. Snow goes down the cracks when they open up….some of the ice at the top of the wedge remelts plus surface water collected and then refreezes. I would say the data isn’t as good as the ice cap type.

    • And would not spring rains melt the snow and fill in the crack too?
      I also like their Figure 1: Temperature and climate forcing, where they show GHG radiative forcing swelling from -0.5 W/m^2 to 2 W/m^2 and compare to the MJJASO Isolation dropping by 10 W/m^2. Some change.

    • Right. There wasn’t enough CO2 being added to the atmo by 1850. NOBODY is arguing that there was.

      What they really have is the warm-up after the end of the LIA. But they are too uninformed to know the difference.

  6. Last sentence of the abstract:

    Furthermore, our reconstructed trend as well as the recent maximum are consistent with the greenhouse gas forcing and climate model simulations, thus reconciling differing estimates of Arctic and northern high-latitude temperature evolution during the Holocene.

    • Hang on, If it is consistent with climate model simulations, then we know there is a problem because the models can’t simulate reality well at all.

      Canadian borehole studies showed years ago 6 deg C of Arctic warming over 150 years. This sort of, messily, confirms the borehole temperature study. There is a correlation with the rise from 1850 – no doubt about it.

      The leap to causation is a big one. I certainly would not have chosen ‘climate models’ as a yardstick of proof. They have been invalidated by their inability to deal with spikes and pauses and cooling.

      The last thing I want said about my study is, ‘It has as much value as a climate model.’ Crikey. Talk about faint praise.

    • Execpt there is no need for a recent reconstruction and there is no recent maximum. There is a weather station at the base of the delta. See main comment below.

  7. Global warming reduces it, therefore making the heat insulator for the permafrost thinner. Then even weak frosts are enough to freeze the ground deeper below the surface.”

    • The effect is always local. Have to know about WHEN the snow falls came. In Western Canada, we understand that the timing of snow cover is all important. Early cold, late snow and the frost line dives down and freezes our pipes. We don’t know about the freezing – and cracking – until springtime, when the ground melts and the pipes come apart.

      • Yes, and…

        Somebody trots a path on snow over a water pipe, and frost line goes down to the pipe. Then temps rise (‘heat’) from -35°C to +5°C and the pipe says bang because ice in the pipe expands in every direction before melting.

        BTW heating, hotter, etc. language is pretty funny when what we experience here is milder sub-minus-twenty-centigrade temperatures. And pretty much equal summer daytime highs.

  8. I know in northern Canada there are islands of “stranded Black Spruce” that indicate the tree line was north of todays tree line during the Holocene Optimum. The spruce in the islands cannot reproduce via seeds, so it as assumed they arrived during warmer times, but todays tree line is at least a few hundred kilometers south of Holocene Optimum. I’ll try and dig out references. I know it is mentioned by E.C. Pielou in her book:

    After the Ice Age: The Return of Life to Glaciated North America

  9. Compare:
    “…the exact extent of warming is something we can’t yet make a statement on.”
    with
    “Starting at that point (the dawn of industrialisation around 1850) our data curve shows a major increase that clearly differs from the gradual warming in the previous phase.”

  10. “Starting at that point (the dawn of industrialisation around 1850) our data curve shows a major increase that clearly differs from the gradual warming in the previous phase.”

    Grant money duly earned.

  11. According to Hanno Meyer: “The curve shows a clear partitioning. Up to the dawn of industrialisation around 1850, we can attribute the development to changes in the Earth’s position relative to the sun. In other words, the duration and intensity of the solar radiation increased from winter to winter, causing temperatures to rise.

    Say what? The intensity of the winter solar radiation has been decreasing since it peaked out about 11,000 years ago … I hate it when I find things like this, it means you truly can’t trust a word out of their mouths after such an egregious error.

    w.

    [UPDATE: as Steven Mosher and Nick Stokes pointed out below, I was 100% wrong in this claim. My thanks to them for checking my work. -w.]

      • Rud.

        1. they seem to be making a claim about particular seasons. So, you’ll have to parse the seasons out.

        2. I would look at the REGIONAL expectation and not a single station.

        3. If I were looking at local detail I would NOT use the global product. the global product is based
        on a regression equation. That regression equation explains 93% of the variance for an random
        station. HOWEVER, you will and you MUST find various stations where the local detail could
        be captured more accurately by doing a different regressions. This is especially true in regions
        close to coasts. In these cases if you are really interested in the local detail a TAILORED regression
        will give you something that is more accurate.

    • “Say what? The intensity of the winter solar radiation has been decreasing since it peaked out about 11,000 years ago”

      No. From Fischer and Jungclaus, 2011
      “The Earth’s obliquity was larger 6000 yr ago and, at the same time, summer solstice was closer to perihelion. In the Northern Hemisphere, this resulted in higher insolation in summer and lower insolation in winter, and, consequently, in increased seasonal differences in insolation.”

      • Ah, I see. Mosher and Nick are correct, and I was 100% wrong. They are using N-D-J-F-M-A for their winter calculations.

        Well done, gentlemen. You can see why I say I don’t trust anyone’s numbers … including my own.

        w.

    • Willis, you called it. phlogiston, you are right, they did. See my fact check below. Siberia is remote, but not uninhabited. So there is an instrumental record for the entire AGW period of interest for the Lena delta.

    • There is a much better refutation. See posts below.
      But regards to a fellow civilian scientist. I had to have Judith publish a corrigendum to one of my guest posts over there last year (maybe 2013?). Glad her denizens caught the goof. Simple writing stupidity on my part. Been there, done that.
      Here your instincts proved correct even if for the wrong reasons. See below. The funniest part is how Nick Stokes ‘rode to the rescue’, only to be unsaddled.

  12. Interesting. Over that same time span most geological data indicate that the globe as a whole has cooled. AND over all, sea levels may have declined about a meter. That would pretty much indicate that what we see at present is simply more of the same.

  13. Color me skeptical. Maybe someone can assure me that this data is sound. Measuring the O18 in ice wedges from melted snow in the Arctic is like a proxy of a proxy isn’t it? Doesn’t the O16:O18 ratio in the snow depend on what time of year the snow fell and where the atmospheric moisture came from in the first place? For instance, doesn’t lake effect snow give you much different results than snow from a Nor’easter? There seems to be so many variables here, it’s not like taking oxygen isotope ratios directly from a marine source, and their results are questionable. There was a sudden shift in the ratio as soon as the first combustion engine was invented? Very odd. Validation or correction would be appreciated.

  14. Well, the early ice core stuff is interesting, but the late stuff is wrong and completely unnecessary—and the authors know it. They got to the Lena delta from the Laptov sea port city of Tiksi, which is just southeast of the delta. Tiksi was founded in 1933 and there is a semi continuous station record from 1936. IPCC says AGW is only after 1950 or so. Tiksi is BEST station 169945. There has been essentially no meaningful change in mean temp. After QC, the mean anomaly change is 0.53C/century. After a single breakpoint alignment, 0.94C/century. No global warming, no arctic amplification, bad paper.
    Check Tiksi on a map of northeastern Siberia. Check BEST. Chuck paper conclusions.

    • “There has been essentially no meaningful change in mean temp.”
      TIKSI is just one place, and not a GHCN station. The GHCN Stations in the Lena valley show plenty of change:

      Dzardzan GHCN Trend 2.54°C/century since 1952

      Viljujsk GHCN Trend 4.15°C/Cen since 1952

      Zhigansk GHCN

      • Nick, you fell for a few little tricks exposing much, which were preplanned into the post about which you complain. Glad you took the bait.
        First, Had you checked, the Lena is the third largest river in Siberia, and tenth largest in the world. Obviously there are a lot of cities along such a waterway, all much further south, and bigger, with more UHI, than at the delta. (Check a map. The river flows basically due north across most of Russia, just like the Mississippi flows south across most of the US.) The population of Tiksi is 5000. You nicely illustrate the UHI problem BEST and other homogenizations do NOT fix in the bigger cities further south along it. Read essay When Data Isn’t for details, then get back.
        Second, GHCN has been greviously fiddled by homogenization. Again see essay When Data Isn’t for many irrefutable examples. I use the BEST analysis because although far from perfect (see book footnote on BEST 166900 and recent posts on BEST 157455), at least transparent and not obviously politically corrupted.
        Oh, and what about the famous Arctic amplification in you own recited data? Aint there, and should be according to your ‘religion’. Massive logic and data fails. Thanks again for highlighting both.

      • Oh, and Nick. Next time you pick comparable stations, mind the regional expectations fallacy exposed by BEST 166900. Over that, you can argue with Mosher (and probably will) til H freezes over. But your outcome wont change the basic flunk. Maybe you should start writing a book refuting mine just published after two years of research?

      • That file went missing pretty quickly. It was the old version of GISS (v2 from 2011) for Viljujsk, homogenized but still showing that there wasn’t a noticeable warming for the last 100 years, unlike the revised version.

      • Robert B, you just learned something. When challenging these types, always back up. Best is not a Wayback Machine, is a locked datestampted .pdf. Otherwise, the past might just continue changing.

      • Since most people would have seen examples of the huge differences between versions in Steve Goddard’s or Paul Homewood’s blogs, deleting the file is a bit of an own goal.

      • For crying out loud another argument abut when an inch isn’t an inch depending on what ruler you use and where you used the ruler.

        I am not sure which has more value, trying to figure out global temperature or trying to figure out how PSI in footballs in deflate-gate was measured.

  15. I thought it had been established that soot from fires deposited in high latitudes was a significant factor in local warming. It appears that this, along with factors such as major ocean current variation at long time scales are never included in studies such as this, or in the computer programs. It appears any results have to be due to AGW.

Comments are closed.