RealClimate logo

Warming, interrupted: Much ado about natural variability

Filed under: — raypierre @ 12 July 2009

A guest commentary by Kyle Swanson – University of Wisconsin-Milwaukee

I am quite humbled by the interest that has been generated by our paper “Has the climate recently shifted?” (Swanson and Tsonis, 2009), and would like the thank the RealClimate editors for the opportunity to give my perspective on this piece.

Before delving into the paper itself, a few words about the place of our work in the global warming “debate” are in order. A quote from the early 20th century Viennese polymath Egon Friedell (which I ran across in the wonderful book Cultural Amnesia by Clive James) captures the situation better than any words I could ever weave;

Electricity and magnetism are those forces of nature by which people who know nothing about electricity and magnetism can explain everything.

Substitute the words “modes of natural climate variability” for “electricity and magnetism,” and well…, hopefully the point is made.

It first needs to be emphasized that natural variability and radiatively forced warming are not competing in some no-holds barred scientific smack down as explanations for the behavior of the global mean temperature over the past century. Both certainly played a role in the evolution of the temperature trajectory over the 20th century, and significant issues remain to be resolved about their relative importance. However, the salient point, one that is oftentimes not clear in arguments about variability in the climate system, is that all else being equal, climate variability and climate sensitivity are flip sides of the same coin. (see also the post Natural Variability and Climate Sensitivity)

A climate that is highly sensitive to radiative forcing (i.e., responds very strongly to increasing greenhouse gas forcing) by definition will be unable to quickly dissipate global mean temperature anomalies arising from either purely natural dynamical processes or stochastic radiative forcing, and hence will have significant internal variability. The opposite also holds. It’s painfully easy to paint oneself logically into a corner by arguing that either (i) vigorous natural variability caused 20th century climate change, but the climate is insensitive to radiative forcing by greenhouse gases; or (ii) the climate is very sensitive to greenhouse gases, but we still are able to attribute details of inter-decadal wiggles in the global mean temperature to a specific forcing cause. Of course, both could be wrong if the climate is not behaving as a linear forced (stochastic + GHG) system.

With that in mind, our paper is fundamentally about inter-decadal variability in the climate system and its role in the evolution of the 20th century climate trajectory, as well as in near-future climate change. The climate system has well known modes of variability, such as the El Niño/Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO), that are active on inter-annual time scales. We are interested in how this short time-scale (from the climate perspective!) variability impacts climate anomalies over multi-decadal time periods.

What we find is that when interannual modes of variability in the climate system have what I’ll refer to as an “episode,” shifts in the multi-decadal global mean temperature trend appear to occur. I’ll leave the details of these episodes to interested readers (here and here), as things get pretty technical. It’s sufficient to note that we have an objective criteria for what defines an episode; we aren’t just eyeballing curves. The climate system appears to have had three distinct “episodes” during the 20th century (during the 1910’s, 1940’s, and 1970’s), and all three marked shifts in the trend of the global mean temperature, along with changes in the qualitative character of ENSO variability. We have also found similar types of shifts in a number of model simulations (both forced and unforced) that were run in support of the IPCC AR4 report.

The contentious part of our paper is that the climate system appears to have had another “episode” around the turn of the 21st century, coinciding with the much discussed “halt” in global warming. Whether or not such a halt has really occurred is of course controversial (it appears quite marked in the HadCRUT3 data, less so in GISTEMP); only time will tell if it’s real. Regardless, it’s important to note that we are not talking about global cooling, just a pause in warming.

What’s our perspective on how the climate will behave in the near future? The HadCRUT3 global mean temperature to the right shows the post-1980 warming, along with the “plateau” in global mean temperature post-1998. Also shown is a linear trend using temperatures over the period 1979-1997 (no cherry picking here; pick any trend that doesn’t include the period 1998-2008). We hypothesize that the established pre-1998 trend is the true forced warming signal, and that the climate system effectively overshot this signal in response to the 1997/98 El Niño. This overshoot is in the process of radiatively dissipating, and the climate will return to its earlier defined, greenhouse gas-forced warming signal. If this hypothesis is correct, the era of consistent record-breaking global mean temperatures will not resume until roughly 2020. Of course, this contrasts sharply with other forecasts of the climate system; the purple line roughly indicates the model-based forecast of Smith et al. (2007) , suggesting with a warming of roughly 0.3 deg C over the 2005-2015 period.

Why would anyone in their right mind believe what I’ve just outlined? Everything hinges on the idea that something extraordinary happened to the climate system in response to the 1997/98 super-El Niño event (an idea that has its roots in the wavelet analysis by Park and Mann (2000)). The figure to the left shows the spatial mean temperature over all grid boxes in the HadCRUT3 data set that have continuous monthly coverage over the 1901-2008 period. While this provides a skewed view of the global mean, as it is heavily weighted toward North America, Europe and coastal areas, unlike the global mean temperature it has the cardinal virtue of being a consistent record with respect to time. The sole exclusion in the figure is the line connecting the 1997 and 1998 temperatures.

Now, anomalous behavior is always in the eye of the beholder. However, the jump in temperature between 1997 and 1998 in this record certainly appears to pass the “smell test” (better than 3 standard deviations of interannual variability) for something out of the ordinary. Nor is this behavior dependent on the underlying time interval chosen, as the same basic picture emerges for any starting time up until the 1980’s, provided you look at locations that have continuous coverage over your interval. Again, as the temperature anomaly associated with this jump dissipates, we hypothesize that the climate system will return to its signal as defined by its pre-1998 behavior in roughly 2020 and resume warming.

What do our results have to do with Global Warming, i.e., the century-scale response to greenhouse gas emissions? VERY LITTLE, contrary to claims that others have made on our behalf. Nature (with hopefully some constructive input from humans) will decide the global warming question based upon climate sensitivity, net radiative forcing, and oceanic storage of heat, not on the type of multi-decadal time scale variability we are discussing here. However, this apparent impulsive behavior explicitly highlights the fact that humanity is poking a complex, nonlinear system with GHG forcing – and that there are no guarantees to how the climate may respond.


Park, J. and M.E. Mann, 2000: Interannual Temperature Events and Shifts in Global Temperature: A Multiple Wavelet Correlation Approach. Earth Interactions, 4-001, 1-36.

Swanson, K.L. and A.A. Tsonis, 2009: Has the climate recently shifted? Geophysical Research Letters, 36, doi:10.1029/2008GL037022.

388 Responses to “Warming, interrupted: Much ado about natural variability”

  1. 251
    manacker says:

    Doug Bostrom

    Lots of philosophical talk there, Doug, but all a bit off-topic here.


  2. 252
    manacker says:

    Hank Roberts

    You asked: “How long do you need to collect temperature data to say with statistical confidence that there is either an up or down trend, given the variability we know about?”

    I’d say a century would be a good start. 8-1/2 years (2001-2009) is definitely too short, as is 30 years (1976-2005).


  3. 253

    manacker writes:

    The recorded fact is that the period 1910-1944 saw a linear warming of 0.53C (Hadley), 1944-1975 saw a very slight cooling and 1976-2000 (or 2005) saw warming of 0.37C (or 0.43C) (also Hadley). The warming has been interrupted, although the current years are still among the warmest recorded, as has been noted above.

    Prior to that, the short period right after modern measurements started showed a cooling trend (1850-1857). This was followed by a warming period from 1858 to around 1879, which was followed by a cooling trend from 1880-1909.

    The fact that you can refer to 1850-1857 as a “trend” shows you don’t know what a trend is or how to measure it. Until you learn some basic statistics, your analysis isn’t very useful.

  4. 254


    That coefficient should be 5.35 (Myhre et al. 1998), not the earlier value of 6.3.

    [Response: Yes, please note my own correction of this error further up in the comment thread, including the sign error (negative would describe the change in upward long wave flux from the surface, which is the opposite of what we usually mean by the ‘forcing’). This is starting to feel a bit like the cover sheets on the TPS reports in ‘Office Space’ ;) -mike]

  5. 255

    max writes:

    Your analysis (206) of UAH temperature anomalies of the 90s vs. the 00s is interesting, and is no doubt correct.

    Another way of looking at it is the method preferred by IPCC: determine a linear trend line over the period measured.

    Doing it this way shows a linear cooling rate of around 0.1°C per decade on average over the 4 main records: (Hadley, GISS, UAH, RSS) from January 2001 to today (the 21st century, so far).

    How long this trend will continue is anyone’s guess, as is the warming rate that will ensue when the current warming interruption has ended.

    A “trend” has to have a statistically significant slope, max. Read:

  6. 256
    manacker says:

    Tamino is right when he writes (240):

    “There are equally long periods both before and after with considerably higher warming rate.”

    The period 1910-1944 lasted 35 years with a linear warming rate of 0.154°C per decade and 0.54°C over the period (Hadley).

    The period 1858-1879 was a bit shorter at only 22 years with a linear warming rate of 0.173°C per decade and 0.38°C linear increase over the period.

    Then there is the well-studied period 1976-2005, with a linear warming rate of 0.143°C per decade and 0.43°C linear warming over the 30-year period.

    But due to all the internal variability and intermittent cooling periods, it probably makes sense to look at the overall linear trend, which was 0.041°C per decade and 0.65°C over the entire 159-year period.


  7. 257
    manacker says:

    dhogaza (242)

    Were the LIA and MWP global or not?

    Since there was neither a global network of surface thermometers nor global satellite measurements of tropospheric temperatures, we need to resort to proxy data.

    But let’s assume that there had been a network of surface thermometers (as there are today). Let’s assume that the LIA (or MWP before) was restricted to Eurasia and the North Atlantic (where there are extensive historical data pointing to both a strong LIA and a MWP).

    So if the coldest part of the LIA, for example, was on average 1.5°C colder over Eurasia and the North Atlantic than today, the globally and annually averaged reading would by definition also have shown a lower temperature than today, once the Eurasian and North Atlantic temperatures were averaged in statistically.

    So one could say that the LIA (and MWP) were, indeed, “global”, even if the temperatures were only strongly cooler (resp. warmer) in an extended geographical region, such as Eurasia and the North Atlantic, just as one can say that the current warming is global, even if some geographical regions do not show such a strong warming trend as others.

    Just another way to look at it.


  8. 258
    D Robinson says:

    In establishing what the long term trend is [or was] they “remove” the 1998 El Nino but don’t do anything about variations from previous El Nino’s and La Nina’s? It cannot be that only the 1998 Super El Nino effected temperatures. Removing one ENSO event instead of all ENSO events is cherry picking.

    Should they have removed volcano forcing as well? Since some of the cooling episodes mentioned in these comments were due to volcanos but the last 10 years of weather were not.

    [Response: We discussed an ENSO corrected time series last year. – gavin]

  9. 259
    Martin Vermeer says:

    Rod B #220:

    Thanks, Martin for opening the door.

    What door? Don’t put words into my mouth please. Mike’s description was/is accurate (minus typoes), I just wanted to point out a perhaps more fruitful way to look at it.

  10. 260
    Richard Steckis says:

    Tamino #242 Says:

    “Re: #235 (Richard Steckis)

    For the lower troposhperic data from satellites the trend becomes a statistically significant decline (p<=0.01).

    No, it doesn’t, neither for RSS nor for UAH data."

    Well. Yes it does. Would you like to see my R code. I don't know how you calculated it but GLM clearly shows a significant decline in temperature from 2001 to present (p <= 0.01). Of course I use monthly data.

    Kevin McKinney:

    2001 is the first year of the 21st Century. But that is not the reason I chose it. I chose it because it is the first year that the immediate influence of the La Nina that followed the super El Nino declined. It is well known that extreme El Ninos are often followed by deep La Ninas.

  11. 261
    G. Karst says:

    Brian Dodge says: 15 Jul 2009 at 9:44 pm

    “there is a very NOT hidden downward trend in ice; arctic summer ice cover, glacier mass, increasing Greenland yearly ice melt, and a march of disintegrating ice shelves further south around Antarctica.”

    It is so much less alarming, if one realizes, no ice has disappeared. It has merely transported it’s self, through the ocean and atmosphere to the southern pole.

    JUNE (month end averages) NSIDC

    1980 Southern Hemisphere = 13.2 million sq km
    1980 Northern Hemisphere = 12.3 million sq km
    Total = 25.5 million sq km

    2009 Southern Hemisphere = 14.4 million sq km
    2009 Northern Hemisphere = 11.5 million sq km
    Total = 25.9 million sq km

    Total ice extent remains relatively constant. A 400,000 sq km increase in sea ice area (30 yr. net ice gain) should balance out any thinning. Remarkably stable total ice (volume) results.

    This slinky toy action of polar ice has yet to be definitively explained. One of the perturbations in the earths “wobble” would be my best bet. Ocean oscillations and orbital variance are just as valid. What fascinates me, is that this tremendous mass transfer from one pole to the other, in itself, will/must perturbate the earth’s eccentricity etc.

    [Response: Even if this theory had the slightest connection to reality, there would be no perceptible mass change associated with floating sea ice. Ice sheets would be a different issue – but they are decreasing at both poles. – gavin]

  12. 262
    Chris says:

    Re #234

    That’s still not correct I think Alex – two errors:

    Firstly, the main point of my original post was to point out the error in your assertion that:

    Even the IPCC2007 has said somewhere, I believe, that the levels of CO2 were probably not high enough until about 1950 to have had much effect on the global average temperature. So I am having a lot of trouble in understanding how, in the Swanson & Tsonis theory, these authors can genuinely believe in an underlying “true forced warming signal” from GHG emissions that began 1850 and continues today

    It’s easy to determine (read their reports!) that the IPCC consider that the median value of the likely range of climate sensitivity to raised CO2 is near 3 oC (of warming per doubling of atmospheric CO2). There’s no controversy about the atmospheric concentrations near the time of the Maunder minimum (around 276) and the mid 20th century (310 ppm around 1940). The anthropogenic contribution to raised CO2 in this period is near 30 ppm. It’s straightforward to calculate a likely anthropogenic contribution to warming (280-310 ppm) is 0.45 oC at equilibrium. So we (and the IPCC!) expect a substantial contribution to warming before 1950.

    The second error relates to your statement just below and the aditional comments in your post #234

    It seems fairly reasonable to suppose that had humans emitted no CO2 at all, temperatures would have risen again anyway in the same pattern that they had fallen previously, i.e. gradually and over a number of centuries.

    Have a look at Moberg et al again. You’ll see that during the period before the so-called Medieval Warm Period (MWP) the paleoreconstruction give (N. hemisphere) temperatures around -0.4 oC below mid 20th century. They went up a tad during the MWP (to around 0 oC relative to mid 20th century) and fell to around -0.6 oC during the Little Ice Age (LIA).

    So, yes it’s quite reasonable to propose that when the reduced solar forcing and negative volcanic forcing (and whatever other factors that might have contributed to LIA cold) dissipated, the temperature would have slowly recovered to it’s pre-perturbation state. This state seems to have been a temperature around -0.4 oC below the mid 20th century level.

    One can draw pretty much the same conclusion from the recent paleoreconstruction of Mann et al. in PNAS [*].

    So my argument in post #132 applies as does your suggestion that the LIA cold was likely to have recovered by itself towards the pre-exisiting equilibrium state. It’s just that the pre-equilibrium state was (according to the paleoreconstructions) quite a bit cooler than the mid 20th century state. We can understand how the temperature rose well past that state (by perhaps around 0.4 oC) from our (and IPCC’s) understanding of the greenhouse effect. Your point about what Dr. Pierrehumbert said on a different thread is not relevant – the straightforward understanding of the recovery and overshoot from the LIA is independent of any causal mechanism(s) of the LIA.

    So again everything is quite consistent with your interpretation of Swanson & Tsonis theory, namely an underlying “true forced warming signal” from GHG emissions that began 1850 and continues today.

    M. E. Mann et al. (2008) Proxy-based reconstructions of hemispheric and global surface temperature variations over the past two millennia Proc. Natl. Acad. Sci. USA 105, 13252-13257

  13. 263
    Mark says:

    “it probably makes sense to look at the overall linear trend, which was 0.041°C per decade and 0.65°C over the entire 159-year period.


    Except that 159 years ago, the amount of CO2 being produced by industry and man in general was much lower than it is today.

    Just ask ExxonMobil how much more oil they are dealing with than even 20 years ago.

    Or are they in on the great conspiracy too?

  14. 264
    G. Karst says:

    Gavin: “no perceptible mass change associated with floating sea ice.”

    Not a theory really, merely an observation. Your response is appropriate and somewhat valid. It does not, however, explain the phenomena. BTW “observations” do have a connection with reality. Slight or not. Thanks for discussing it with me.

  15. 265
    Mark says:

    “Of course I use monthly data.”

    And such data will have sqrt(12) times more noise in them.

    Therefore you need many more months.

    Note also that the data you’re using includes the effect of stratospheric cooling. What do you think the point of that will be?

    Here’s an idea: check the record for June this year.

    Find out the average temperature.

    Check to see if this June is cooler than many other Junes in the past 30 years.

    Now check the June figure in that graph.

    How many Junes are there that are warmer?

    Now why the lack of skepticism? That graph of yours is from wattsupwiththat. Originally Roy Spencer and a pal.

    Where he’s taken 11 instruments (there are more than that up there), mentions that AMSU-A is the most steady and then ignores

    a) what does he mean by steady?
    b) what about the other 11 less steady

    and then creates a graph with a zero that looks like the average temperature on the graph.

    Well of course the end point will come toward zero. If it moves away, the zero line will follow it slowly.

  16. 266
    manacker says:


    It appears that you are confusing empirical evidence (physically observed temperatures) with possible causes for temperature increase.

    The warming period from 1858 to 1879 was certainly not caused by human CO2 emissions, but it was physically observed, nevertheless, as was the warming from 1910 to 1944, when there was hardly any human CO2.

    The 159-year record tells us that it has warmed in three multi-decadal “spurts” (of which the latest 1976-2005 period was one), with cooling periods in between, and an average underlying warming trend of 0.041C per decade.

    These are the physically observed facts. How one wants to explain these is another story, with or without ExxonMobil.


  17. 267
    tamino says:

    Re: #258 (Richard Steckis)

    You’re wrong. The GLM routine in R will compensate for non-Gaussian errors (if so specified), but will not compensate for autocorrelation.

    No, I don’t care to see your R code.

  18. 268
    Richard Steckis says:

    Mark says:

    “Note also that the data you’re using includes the effect of stratospheric cooling. What do you think the point of that will be?”

    No. It does not include stratospheric cooling. The data are from the AMSU TLT channel which maximises in the lower troposphere (approximate 1 – 2 km above the surface. The influence of the lower stratosphere is negligible for this channel.

    And no. The data do not have sqrt(12) times more noise in them. I do not know where that idea comes from. The more data you have the higher the degrees of freedom for the statistic. The higher the degrees of freedom the greater the confidence of the trend.

  19. 269
    Doug Bostrom says:

    Manacker, any word on your assertions about government and the scientific community? Slander or true? Baseless or founded? Paranoid or grounded in reality? Have you changed your mind?

  20. 270
    manacker says:

    Doug Bostrom

    Sorry, your query is OT at the time. We are discussing temperature trends, warming interruptions, etc.

    An open-minded person is always open to changing his/her mind as new data become available, right?


  21. 271
    manacker says:

    G. Karst

    On sea ice the June NSIDC data show that the losses in the Arctic are offset by gains in the Antarctic.


    Click June and then:
    N_06_area.txt for Arctic sea ice extent and area
    S_06_area.txt for Antarctic sea ice extent and area


  22. 272
    Mark says:

    Doug, it’s all #1’s…

    Except the last one. There’d need to be something to change for a start…

  23. 273
    Mark says:

    “And no. The data do not have sqrt(12) times more noise in them. I do not know where that idea comes from.”

    Binomial statistics.

    The difference of each month around the year average is much higher than the change between each year adjacent.

  24. 274
    co2isnotevil says:


    The GISS surface data is broken. There is a significant data anomaly cause when NOAA-14 was replaced with NOAA-16 (10/01). [edit]

    [Response: The GISTEMP data is not the same as the ISCCP analysis – it is instead derived from met stations. The issues with NOAA 14 and 16 are relevant to the ISCCP data which is very clearly caveated to indicate that it should not be used for trends because of the inter satellite calibrations and inhomogeneities. – gavin]

  25. 275
    Doug Bostrom says:

    Manacker, sorry I missed your earlier post, wherein you said:

    “Lots of philosophical talk there, Doug, but all a bit off-topic here.”

    “There” being here, presumably:


    –You’ve established a clear track record of what a reasonable person would consider to be a disingenuous and self-contradictory approach in your communications style. You employ demagogic, inflammatory rhetoric against climate science and climate scientists when attempting to influence selected audiences. Elsewhere you then engage in seemingly reasonable and judicious discourse with members of those same groups you disparage. You apparently expect to be treated as a peer in both arenas.

    –You’ve produced a history of making broadly accusatory pronouncements of malfeasance and corruption germane to the general topic under discussion here. Specifically, you have made statements that a reasonable person would interpret as being accusations of climate scientists reporting scientific results that have been shaped by avarice on the part of climate scientists. You are charging these scientists with being guilty of professional misconduct. At the same time, you do not appear able to verify the accusations you’ve leveled.

    –The history of your behavior, the history that you have created, would lead a reasonable person who is bothered by the deficiencies you have displayed to cast a jaded eye on opinions and statements you may express concerning the topic under discussion here. Everything you say about climate science must therefore be assessed not only on the superficial merits of your words but also with the certain knowledge as established by you yourself that you appear either to suffer from derangement in your perceptions or have a habit of practicing deceit in your communications.

    As I say, this is the appearance you give. You’ve made bold and astounding claims about this topic for which you appear to have no evidence. Perhaps you can change that appearance by justifying your demagoguery with some evidence.

    It’s really not about you specifically in any case. You’re a phenotype of what is commonly termed a “denialist”, with a particularly rich and generous historical track record available for scrutiny. Viewed from that perspective, you’re highly topical.

  26. 276
    chris says:

    re #266

    The warming period from 1858 to 1879 was certainly not caused by human CO2 emissions, but it was physically observed, nevertheless, as was the warming from 1910 to 1944, when there was hardly any human CO2.

    manacker, there was rather a lot of anthropogenic CO2 in the period 1910-1941. The anthropogenic increase in atmospheric CO2 gave a rise from around 280 ppm in the late 18th century to around 297 ppm in 1900 to nearly 310 ppm around 1940 [*]

    The total expected anthropogenic global temperature contribution over this full period within a climate sensitivity of 3 oC is easily calculated. It’s 0.44 oC at equilibrium.

    So human emissions made a very significant contribution to the temperature rises during the periods you specified. In fact they could account for most of the temperature increase during these periods. Of course the anthropogenic contribution was “mixed in” with natural solar and volcanic contributions so one doesn’t expect to observe a simple temporal relationship, any more than we expect to see a simple relationship nowadays (which is what this thread is largely about of course).

    [*]D. M. Etheridge et al (1996) “Natural and anthropogenic changes in atmospheric CO2 over the last 1000 years from air in Antarctic ice and firn J. Geophys Res. 101, 4115 -4128

  27. 277
    RichardC says:

    Thomas Lee Elifritz says:

    Note that the largest disturbance in the examples began to degrade within 100 days.

  28. 278
    Jacob Mack says:

    This warming pause I suspect will be for sometime even if not for another 12 years.

  29. 279
    manacker says:

    Doug Bostrom (275)

    You are getting a bit repetitive, Doug, but still OT.

    Try getting back to the topic here, which is the “warming interruption” since 2001 and its significance (if any) in the long-term picture.


  30. 280
    David B. Benson says:

    Alex Harvey (234) — As I understand it, LIA was almost global in extent, (It seems that Antarctica either did not participate or maybe had a little warm reversal, I’m unsure.)

    The causes include significant volcanic activity and the Maunder and Dalton Minima. Both of these lead to less insolation. When done, the temperature retuns to something more normal. See

    Regarding the subtopic of global temperature trend over the instrumental period (beginning at the earliest in 1850 CE), see the UPDATE on that same thread over at Tamino’s Open Mind. I’ll have something more to post about it later.

  31. 281
    gp2 says:

    Re: 231
    MSU records temperature of the entire troposphere with different weights, this is not the same as global ocean surface temperature…for instance free tropical tropospheric temperature depends upon heat transfer due to deep convection i.e. a relative small area with high sst and strong deep convection may influence the free troposheric temperature accross the tropics much more then large area with high ssta but lower sst.

    “It is expected that the free tropospheric
    temperature should be sensitive primarily to SST anomalies in regions in which the mean SST is
    high and deep convection is frequent, rather than to the tropical mean SST.”

    Furthermore there are major uncertain for long term trend in the tropics…different analysis shows very different trend and UAH trend is the lowest.
    An interesting paper that covers uncertain and propose a new intercalibration methods for MSU:

    “However, temperature
    trends obtained from these observations are still
    under debate; different results are obtained by different
    investigators. Further investigation is required to reconcile
    these differences.”
    They found a global ocean T2 trend of +0.21K/decade between 1987 and 2006 compared to +0.14K/decade (RSS) and +0.08K/decade (UAH).

  32. 282
    manacker says:

    Chris (276)

    You wrote:

    “there was rather a lot of anthropogenic CO2 in the period 1910-1941. The anthropogenic increase in atmospheric CO2 gave a rise from around 280 ppm in the late 18th century to around 297 ppm in 1900 to nearly 310 ppm around 1940 [*]
    The total expected anthropogenic global temperature contribution over this full period within a climate sensitivity of 3 oC is easily calculated. It’s 0.44 oC at equilibrium.”

    Sorry, Chris, but your calculation is incorrect.

    Let’s do it again
    C1 = 297 ppmv (concentration in 1910)
    C2 = 310 ppmv (concentration in 1944)
    C2/C1 = 1.04
    ln(C2/C1) = 0.0428
    ln(2) = 0.6931
    dT (2xCO2) = 3.2°C (at equilibrium)
    dT (1910-1944) = (3.2 * 0.0428) / 0.6931 = 0.20°C (at equilibrium)

    So anthropogenic CO2 could only have caused 0.20°C warming over the period (not 0.44°C), while 0.54°C were physically observed.


  33. 283
    Brian Dodge says:

    G. Karst says 17 Jul 2009 at 9:18 am;
    “Total ice extent remains relatively constant. A 400,000 sq km increase in sea ice area (30 yr. net ice gain) should balance out any thinning. Remarkably stable total ice (volume) results.”

    “Because sea ice does not stay in the Antarctic as long as it does in the Arctic, it does not have the opportunity to grow as thick as sea ice in the Arctic. While thickness varies significantly within both regions, Antarctic ice is typically 1 to 2 meters (3 to 6 feet) thick, while most of the Arctic is covered by sea ice 2 to 3 meters (6 to 9 feet) thick. Some Arctic regions are covered with ice that is 4 to 5 meters (12 to 15 feet) thick.”

    If one only remembers to subtract the ice lost by the collapse of Larsen B-(3,250 km^2) and Wilkins-(more than 2000 km^2 so far; about 10k km^2 when it all disappears) ice shelves, more than 200 meters thick; or the equivalent of 500,000+ square kilometers at 2 meter thickness; then the plus 400,000 km^2 sea ice area doesn’t “balance out any thinning”.

    When one considers the following:
    “Satellite radar altimeter measurements show that between 1992 and 2001 the Larsen C Ice Shelf lowered by up to 0.27 ± 0.11 meters per year.”

    “In the last 5 years, the picture of a slowly changing Antarctic ice sheet has radically altered. It is now realised that ice shelf basal melting may account for up to one third of the loss from the grounded ice; extensive, rapid thinning is occurring in one part of the West Antarctic ice sheet interior; and the collapse of the Antarctic Peninsula ice shelves is accelerating the grounded ice discharge.”

    “In order to better constrain the coastal element of the problem, Rignot et al. have analyzed satellite interferometric synthetic-aperture radar observations of Antarctica’s coastline from 1992 to 2006 to estimate ice flux to the oceans. … Ice mass loss from the coasts increased by 75% over the period of the study.”;319/5861/259d

    It is obvious that total ice volume is not “remarkably stable”, but declining.

    Perhaps the ice melting has resulted in a lower salinity surface layer in the seas surrounding Antarctica, which freezes more easily and extensively?

  34. 284
    chris says:

    re #282

    careful manacker. You didn’t read my post correctly. My calculation of 0.44 oC corresponds to the anthropogenic contribution during the full period of late 18th century to 1940 (280 – 310 ppm).

    That encompasses both of the periods you specified. The anthropogenic contribution accounts for a large proportion of warming during this period. Since it’s mixed in with the natural variaiton (volcanic/solar) we don’t expect to observe a stesdy temperature rise that follows the excess CO2 forcing perfectly. But the warming contribution will be there.

    That’s relevant to your rather cherry-picked start date for the later warming period of 1910-1944. The cooling period previous to 1910 was most likely a result of extensive volcanic forcing, so that some of the post-1910 warming was a recovery from this.

    Your own calculation shows that your statements in post #266 are wrong. There was “human CO2” in the periods you specified. This “human CO2” did contribute to warming.

  35. 285
    manacker says:

    Brian Dodge (283)

    Speculations on relative sea ice thickness in the Arctic and Antarctic are interesting, but not that relevant.

    Changes in floating sea ice (even including “dramatic” events, such as the collapse of Larsen B and Wilkins) have no substantial impact on sea levels.

    These are primarily of interest as they may affect surface albedo, and in that respect thickness has no impact, but it is total sea ice extent (or area) in both the Arctic and Antarctic that counts.

    Based on the latest NSIDC data for June, these appear to have been constant, with net losses in one region essentially offset by net gains in the other.


  36. 286
    co2isnotevil says:

    Max, re 282

    Your calculation is incorrect as well. The proper calculation is to first determine the power flux in and out of the surface, which for an average surface temperature of 288K is 390.11 W/m^2 from Stefan-Boltzmann. The forcing power (using the IPCC heuristic) is,

    5.35 * (ln(310/280) – ln(297/280)) = 0.23 W/m^2

    Add this to 390 (390.23) and convert back to a temperature, we get 288.02K, for a net increase of 0.042K. Now, the sensitivity coefficient of 0.8, which would predict a temperature increase of 0.18K from yet another heuristic, is after the effects of feedback. However, part of the problem today, is that the warming this heuristic predicts hasn’t happened, so the presumption is that this energy is hiding out some where and some should have been hidden back then too.

    Of course, this is all moot, because there’s a simpler explanation that is independent of CO2. Around 1910, there was record sustained level of low sun spot activity, the likes of which we haven’t seen since, and which resulted in significant cooling. This cooling rebounded as the sunspot activity increased up to it’s record high levels in recent history. Today, the current 11-year sun spot minimum is deeper than we have ever measured with modern techniques, but it remains to be seen if it will reach the depths of the minimum seen around 1910-1911.

    There’s an easy way to test this correlation, which is to plot monthly averaged sun spot numbers against the temperature record. I’ve done this as have others, but you will need to do this for yourself to believe it. I suggest plotting an 11-year running average of both temperature and the sunspot number.


  37. 287
    Doug Bostrom says:

    manacker 17 Jul 2009 at 3:35 pm

    “You are getting a bit repetitive, Doug, but still OT.”

    Not nearly as robotically, verbatim repetitive as you’ve been elsewhere, not to mention off-topic by your oddly elastic standards. How about your “A Layman’s Review of the latest IPCC report”, which you dumped regardless of context on at least half a dozen sites? That was a real gem of the pasty, bogus variety, containing as it did what appeared to be catchy talking points distilled from multiple sources, thrown together into a stream-of-consciousness rant and then inserted in multiple locations.

    That’s not the only time you’ve done that, either. I’ll spare everybody the details.

    Given the hopeless hole you earlier dug for yourself and seem doggedly reluctant to climb out of, I’m not surprised you’re insisting on remaining “on-topic” here. Your embarrassing past can’t really be put to rest without an explicit mea culpa and apology.

    After all, here you are, enjoying the hospitality of folks you elsewhere essentially called liars and charlatans. No wonder you’re skulking and would rather avoid discussing that uncomfortable topic. You’re dining in the house as a guest, but you shat in the yard as a vandal.

    Why not just say “Gee, I’m genuinely sorry, I was just wrong when I made a blanket accusation of theft directed at the climate science community, I’ve since come to realize that those words were hasty, ill-considered and had no foundation of truth”?

    Really nothing more to add here, other than a reminder that accountability in social matters is a form of algebra where it’s painfully obvious when an equation does not balance. At least you do consistently identify yourself; hats off to you for making it possible to sum you up.

  38. 288
    Steve E says:

    In #70 Ike Solem says:

    “Volcanic events are…are insensitive to the state of the ‘the climate system’.”

    I realize there are much higher-priority unanswered questions out there, but I’ve often wondered about this: If the Antarctic or Greenland ice sheets melt faster due to continued warming, one would think the disappearance of that ice would allow for faster isostatic rebound, and that in turn may have some impact on volcanism. And if so, one could then say there’s a global warming connection to volcanoes. But I agree with the gist of your post.

  39. 289
    David B. Benson says:

    Following up on my comment #280, assume a linear system (sorta dubious for climate, but the assumption means some suggestive answers are readily obtained by hand). Assume the very reasonable system response to a unit step function given in
    which has a LaPlace transform of the form
    1/s – 1/2(s+a) – 1/2(s+b)
    with a = 1 and b = 1/30 (using Tamino’s values) and where, for simplicity, the two weights are both 1/2, summing to 1 as required.

    Now for a ramp forcing, linearly increasing over time (as has been assumed in comments on this thread), the LaPlace transform is
    1/s^2 – 1/2s(s+a) – 1/2s(s+b)
    and the system response is, as expected, of the form
    k[t – c + transient]
    where the transient response function is initially c and decays as a sum of two dying exponentials. With b representing thirty years, one needs to take this transient respnse into account even for rather long intervals.

    The point is that it is not strictly correct to simply plug into the equilibrium response logarithmic equation to compute the temperature gain over any decadal, even centennial, interval. One needs to determine c depending upon a, b and the two weights. For the stated values, c is about 0.517.

    Now there is an arbitrary scale factor, k, to convert time to temperature. I suggest using parameter estimation methods on, say, HadCRUT3g to find the best fitting scale factor (for those who wish to carry out modestly accurate such comparisons between the historical record and the assumed linearization of the CO2 forcing). If one also uses paramter estimation simultaneously for k, a, b and the two weights, there are then five parameters to be esimated; not for the impatient or impersistent. As responses above have indicated, you’ll not be able to determine Charney sensitivity with any precision this way.

  40. 290
    BobFJ says:

    Brian Dodge (283) makes reference to ice shelf collapse in Antarctica:
    May I point out that this phenomena and also the calving of icebergs are fundamentally mechanical fracturing failures. If you study the canyons developing in ice shelves, before break-up, they appear to run parallel to the influence of the ocean, primarily because of tidal and wave action, resulting in hinging at the sites of the canyons. As an engineer, I find it hard to see any climate change effect that would significantly accelerate this normal process. Changed wind strength or circulation? Increased melt within the canyons? Maybe…. but I doubt it.

    [Response: Actually, temperature increase has been shown to have a very pronounced effect on the breakup of ice shelves. It is indeed mechanical fracture that does the job, but what warming does is to form massive melt ponds at the surface, which then cause hydraulic fracture. –raypierre]

  41. 291
    David B. Benson says:

    co2isnotevil (286) — Alas, correlation is not causation. Better, I opine, is to read Tung & Cabin (2008), available from Prof. Tung’s UW web site. There you will find an analysis of about 50 years of data and a determination that global temperatures go up and down over solar sunspot cycles. What is surprising is their technique shows considerably larger amplitude to the cycling than previous researchers have found.

  42. 292
    BobFJ says:

    Martin Vermeer,
    Sorry, but I should have included you as an addressee having a special interest in my post 191 above, (the bit below the wiggly line)

  43. 293
    Richard Steckis says:

    Tamino #267 says:

    “You’re wrong. The GLM routine in R will compensate for non-Gaussian errors (if so specified), but will not compensate for autocorrelation.”

    You are right. Does your criticism apply to all the linear regressions that you present at your site? Do you always test for autocorrelation before presenting linear regression models?

  44. 294
    Richard Steckis says:

    Tamino #267

    Further. Do you suggest an alternative that corrects for autocorrelation. Perhaps GLS?

  45. 295
    Brian Dodge says:

    manacker says 17 Jul 2009 at 5:21 pm
    “These are primarily of interest as they may affect surface albedo, and in that respect thickness has no impact, but it is total sea ice extent (or area) in both the Arctic and Antarctic that counts.”
    The decrease in albedo is in the Arctic, where it’s summertime and the sun is shining; the supposedly compensating Antarctic increase in extent/albedo is occurring where there is much less or no insolation.
    (And I was specifically discussing the claim that “Remarkably stable total ice (volume) results.”, where thickness does count.)

    BobFJ says 17 Jul 2009 at 8:20 pm
    “I find it hard to see any climate change effect that would significantly accelerate this normal process.”
    Warmer temperatures cause larger and longer duration of surface melt on the ice shelves.
    Hydrostatic pressure from the ponding of meltwater on the surface of ice shelves causes cracks to propagate; shortly after the disappearance of meltwater ponds on the surface of Larsen B was observed, indicating that the cracks had gone full thickness, the shelf abruptly collapsed. Stress from tides, waves, and winds play a role (and the increase in the strength of the circumpolar vortex predicted by climate models probably contributed), but warming has set the stage for the demise of ice sheets that have existed for 6-10 thousand years. Start with to learn more.

  46. 296
    Brian Dodge says:

    @ David Horton 16 Jul 2009 at 3:35 am
    Re endlessly pushing a boulder up a hill – I think the phrase you’re looking for is “Sysyphean whack-a-mole”

  47. 297
    Hank Roberts says:

    BobFJ, would it safe to assume that as an engineer, you find it hard to see any climate change effect, and haven’t looked for anyone else’s work to inform your opinion?

    If you have read on the subject, what have you read?
    Have you searched (search box, top of page) here?

  48. 298
    Richard Steckis says:

    Further again,

    I tested the RSS time series from 2001 to present using ACF and PACF in R and could not find any significant autocorrelation or partial autocorrelation for that time series. This is even though the Durbin Watson test was significant (p<=0.001) with a DW statistic of 0.85

  49. 299
    Martin Vermeer says:

    BobFJ #290, eh… “ice flow”… “basal melting”… “rapid thinning”… not just mechanics. You can read up on these things.

  50. 300
    Doug Bostrom says:

    BobFJ 17 Jul 2009 at 8:20 pm:

    “If you study the canyons developing in ice shelves, before break-up, they appear to run parallel to the influence of the ocean, primarily because of tidal and wave action, resulting in hinging at the sites of the canyons. As an engineer, I find it hard to see any climate change effect that would significantly accelerate this normal process.”

    Climate change effect? Temperature? Melting point of ice? Loss of mass? Reduction of beam cross section? Deformation of beam leading to strain and failure?