It’s long been known that El Niño variability affects the global mean temperature anomalies. 1998 was so warm in part because of the big El Niño event over the winter of 1997-1998 which directly warmed a large part of the Pacific, and indirectly warmed (via the large increase in water vapour) an even larger region. The opposite effect was seen with the La Niña event this last winter. Since the variability associated with these events is large compared to expected global warming trends over a short number of years, the underlying trends might be more clearly seen if the El Niño events (more generally, the El Niño – Southern Oscillation (ENSO)) were taken out of the way. There is no perfect way to do this – but there are a couple of reasonable approaches.
In particular, the Thompson et al (2008) paper (discussed here), used a neat way to extract the ENSO signal from the SST data, by building a simple physical model for how the tropical Pacific anomalies affect the mean. He kindly used the same approach for the HadCRUT3v data (pictured below) and I adapted it for the GISTEMP data as well. This might not be ideal, but it’s not too bad:
(Each line has been re-adjusted so that it has a mean of zero over the period 1961-1990).
The basic picture over the long term doesn’t change. The trends over the last 30 years remain though the interannual variability is slightly reduced (as you’d expect). The magnitude of the adjustment varies between +/-0.25ºC. You can more clearly see the impacts of the volcanoes (Agung: 1963, El Chichon: 1982, Pinatubo: 1991). Over the short term though, it does make a difference. Notably, the extreme warmth in 1998 is somewhat subdued, as is last winter’s coolness. The warmest year designation (now in the absence of a strong El Niño) is more clearly seen to be 2005 (in GISTEMP) or either 2005 or 2001 (in HadCRUT3v). This last decade is still the warmest decade in the record, and the top 8 or 10 years (depending on the data source) are all in the last 10 years!
Despite our advice, people are still insisting that short term trends are meaningful, and so to keep them happy, standard linear regression trends in the ENSO-corrected annual means are all positive since 1998 (though not significantly so). These are slightly more meaningful than for the non-ENSO corrected versions, but not by much – as usual, corrections for auto-correlation would expand the error bars further.
The differences in the two products (HadCRUT3v and GISTEMP) are mostly a function of coverage and extrapolation procedures where there is an absence of data. Since one of those areas with no station coverage is the Arctic Ocean, (which as you know has been warming up somewhat), that puts in a growing difference between the products. HadCRUT3v does not extrapolate past the coast, while GISTEMP extrapolates from the circum-Arctic stations – the former implies that the Arctic is warming at the same rate as the rest of the globe, while the latter assumes that the Arctic is warming as fast as the highest measured latitudes. Both assumptions might be wrong of course, but a good test will be from the Arctic Buoy data once they have been processed up to the present and a specific Arctic Ocean product is made. There are some seasonal issues as well (spring Arctic trends are much stronger the summer trends since it is very hard to go significantly above 0ºC while there is any ice left).
Update: A similar analysis (with similar conclusions) was published by Fawcett (2008) (p141).
The ENSO-corrected data can be downloaded here. Note that because the correction is not necessarily zero for the respective baselines, each each time series needs to be independently normalised to get a common baseline.
#194
Yes, I am looking at the 30 year trend. And the first 20 years (1978-1998) show a steady climb upwards, interrupted only by a volcanic eruption. Then the last 10 years the upward trend disappears. The whole 30 year period has been dominated by +PDO and El Nino ENSO, natural factors that favor warmer temperatures. So what changed in the past 10 years to stop the temperature rise? I still haven’t seen that question answered.
#190
Yes, David, El Nino has dominated the past 30 years. But La Nina dominated the previous -PDO period from 1946-76. And then before that, El Nino dominated the 1925-45 period. So there really isn’t anything too unusual about El Nino dominating this latest +PDO phase.
“…The World Meteorological Organization. Did you actually read what I posted?…”
(Barton Paul Levenson)
It was a rhetorical question, I’ll well aware of the orthodoxy on the definition of climate change, I just don’t agree. See my somewhat sarcastic response to another poster at #180 or so.
Regards, BRK
Rod, enough with the straw men.
You made up something Tamino did not say — a caricature, distorted, incorrect — and attribute it to him, and then argue with it.
Read what he wrote. Get real.
He’s trying to teach you one of the hardest lessons everyone learns who passes Statistics 101.
And you’re failing to learn it.
If you don’t understand this you won’t understand climate research.
Re: #199 (Rod B)
Visual inspection (looking at the graph) is one of the best ways to gain insight — I often scold scientists for failure to appreciate the importance of actually looking at the data (you might be surprised how often researchers simply run the data through a “black box” without actually looking at it). The eye/brain combination is one of the most potent pattern-recognition systems ever. But it’s very far from infallible, and what a non-analyst “sees” is usually different from what an experienced analyst sees; years of experience combining visual inspection with numerical analysis actually improves the reliability of visual inspection.
While visual inspection is perhaps the most powerful general-purpose pattern recognition method, it may also be the most susceptible to “false alarms.” So, conclusions based on visual inspection alone are suspect, even those coming from an experienced analyst. Only the combination with numerical analysis can provide the level of confidence needed to make assertions.
Analysis will rarely make the results “do a 180”; I seriously doubt that analysis will indicate the apparent increases in the 1950s and 1890s are actually decreases! But you’d be surprised (very surprised, I’ll venture) how often analysis makes the result “do a 90.” The increase you thought couldn’t be by accident turns out to have no significance at all, or the blip you were sure is meaningless turns out to be an undeniable physical change.
If you want to know how the recent decades compare with previous ones, I’d suggest acquiring the data and helping yourself with some basic analysis. One of the simplest is to smooth the data, and one of the simplest and most reliable methods of doing so is to compute running averages. It’s also a very accessible method, since ExCel will compute (and graph) running averages. Of course, one then has to contend with the issue of how long a time period to average, and how to interpret the results of that analysis. But it will surely provide more information, and will help assess whether the modern era is “just like the others” or something really different.
To the lay readers I say: don’t trust what the graph looks like without some analysis to assist your evaluation. To the scientist I say: don’t trust analysis without careful and detailed visual inspection. And to both: it always helps to get a 2nd opinion (but try your best not to bias that opinion by expressing your belief or your preference).
Boyle, #201.
No, don’t look IN the 30-year period. ADD UP each thirty year period.
Saying that the last 10 years is flat is what WEATHER is happening. 10 years can’t tell you squat about climate.
So, add up the last 30 years average temperatures.
Add up the 30 before that.
And the 30 before that.
Go back 210 years.
Plot.
See a trend.
PS “we’re in a cycle” a’ la Steven Goddard is silly: why is there a 15-year cycle THIS last 15 years but not before then? Because nobody has ever shown any, y’know, *cyclic* figures before that showing the same periodicity.
Maybe those saying have made up whatever period fits the bit they are showing. But that’s not science. Heck, it’s not even honesty.
#189 sidd
If Prof. ‘Twirlip of the Mists’ was trying to represent unsubstantiated opinion as fact (especially in the face of solid evidence that contradicts his assertion), yet, I would dismiss his reasoning.
My point is simple, if she is going to say she represents the truth, then she should tell us who she is. She is making definitive claims that are simply wrong. Any kindergarten student can probably figure out that 1.9 is larger than .3 if you show them the difference.
As to the poster in question being a victim of stalking or domestic violence? Maybe. Personally, I see her driving downtown in Austin, TX in a corvette (when she wants to strut) or greener vehicle when she wants to show how green she is, feeling pretty confident she has a handle on pretty much everything.
The main point is I would not have asked her name if she had answered my question and shown us here evidence that proves she is at the very least reasonably right about her claim that this is all solar. Instead she makes vague references to solar cycles that prove absolutely nothing because they don’t have context, nor relevance to known values of forcing of solar and AGW GHG’s.
She continues to ignore the forcing levels of current AGW estimated around 1.9 W/m2 and still won’t explain to us why .3 is not less than 1.9 W/m2.
Said another way, we don’t need sunspots to keep warming, it just means the warming will be a little slower, and when sunspots return, the warming will be a little faster, then compounded by positive feedbacks as well. If FurryCat can lay some concrete in here, I’m sure we’d all be very impressed.
#200 Rod B
I take Taminos post to mean that correlation does not mean causation like in the GCR’s argument that FurryCat presented on the aa index https://www.realclimate.org/index.php?p=42 saying that it matches, so it must be…
But according to FurryCatHerder I don’t know what that means; but I don’t think Tamino is being excessive, I think he is being considerate of context and relevance. But, hey, that’s just my opinion, and anyway, I’m a putz, from what I hear.
Rod,
Read what Tamino said in #205. It is important. When you get a dataset, the first step is exploratory data analysis–plotting data in different ways, comparing it to known distributions, looking for periodicities, trends, etc. The human eye is one of the most powerful tools for spotting such trends. Hundreds of millions of years of evolution have sculpted a tool with tremendous sensitivity to patterns, but it takes statistical analysis to determine if the patterns are really there.
Boyle (201) — Here are the decadal averages from the HadCRUTv3 global temperature product:
http://tamino.files.wordpress.com/2008/04/10yave.jpg
It is unfortunate that Ockham’s quite reasonable KISS admonition has been so widely misunderstood as “the simplest explanation is most likely to be correct.” Keeping your theories as simple as possible is an excellent rule of thumb for scientific investigation, because it generally takes much less data to exclude a simple model that a complex one, so working from simple to complex is the most efficient way of progressing through a series of possible models/explanations. Turning this from a practical admonition about model-building into an assertion about probability is utterly without foundation, and flies in the face of a long, long scientific history of simple models being supplanted by more complex one.
Mr. Reisman wrote on the 10th of July, 1400:
“if she is going to say she represents the truth, then she should tell us who she is.”
i do not see the syllogism. i might say that i represent the furry aliens from Tralfamadore, and you are free to believe or disbelieve me according to your view of supporting evidence, if any. Such supporting evidence has nothing to do with the name i chose to post under.
similarly if i claim that statistics governing fermions differ from those governing bosons, the truth of my statement is not affected whether i post this as Satyendranath or as Bozo. Else discussion degenerates into ad hominem and argument from authority.
I am deeply disinterested in the identities of posters, and very much against forcing participants to reveal their names before their arguments are considered. But this is getting very far afield, so i will not object if our gracious moderators choose to delete this thread entirely.
To drag the discussion back to climatology, I have a question. I am aware of some data measuring heat fluxes into deep ocean. Are there data accurate enough to tell if these fluxes have changed appreciably over, say, the last three decades ? And more sharply, is there any evidence that these fluxes are at all correlated with AMO/PDO/… ?
Hank (204), while I shortened it for essentials, I came close to quoting Tamino verbatim. What is it you think I should read?
Tamino (205), While I still maintain your assessment of the graph in question is a gross over-reaction and incorrect, and FurryCatHerder’s point has merit, I do agree with almost every word of your 205 post.
#209 sidd
Luckily this is America and you are free to be disinterested in anything you like.
But your disinterests are not my disinterests. That’s the beauty of individuality and freedom of thought. Please feel free to think as you wish and not let my interests impede your thoughts.
#206
Mark,
Yes, and I find ten years a long enough time to distinguish between “noise” and “trends”. Why? Because there are no prior ten year periods during the warming signal that can be equated with this one. If it is noise, it should appear periodically. Show me another flat ten year trend during a +PDO/+ENSO phase the past 100 years. In addition, there should be some sort of scientific explanation for the “noise”…if there is not, then can we not discount the previous ten year trend (1988-1998) as just noise as well? We must have some faith in the accurracy of global temperature readings, or else there is no way to be certain of any trends.
In re 205:
I know what to look for to know if a single outlier, or small number of points, represents a trend or an outlier(s). I’d be surprised if a 10 year moving average failed to show anything but what the graph appears to show. I don’t have time at this precise moment to do that, but I’d be surprised if a 10 year moving average even showed a 45 degree turn, much less the dreaded 90 or 180 degree variety.
But I’ll make my point to you that I made earlier to Reisman — whatever natural trends caused those cycles through 150 or whatever years of industrialization has to still be at work. The alternative is that they are purely chance, and chance seldom makes pretty charts with nice cycles.
In addition, Mark, doesn’t it seem somewhat arbitrary to label the latest ten-year period as “noise”, and yet we seem to accept any other ten year period as indicative of the overall trend?
[edit – don’t just repeat yourself all the time – it’s dull]
Boyle (218) — I left this for you on another thread, but you seemed to have wandered over here.
Here is the decadal averages of the HadCRUTv3 global temperature product:
http://tamino.files.wordpress.com/2008/04/10yave.jpg
Study it this time, please.
re: 218. 1998 was an exceptionally warm year due to the AGW and the additional influence of the historically strong El Nino on top of it. Using 1998 as a reference point for a any trend is classic cherry-picking of the data. The long-term trend is unequivocal.
“Using 1998 as a reference point for a any trend is classic cherry-picking of the data. The long-term trend is unequivocal.”
This is false. The 1998 El Nino was immediately followed by a long La Nina. There were 7 El Nino/La Nina cycles in that ten year period. As Gavin’s ENSO adjusted data shows, The decadal flat trend is not due to ENSO or any cherry picked ENSO endpoints.
http://reallyrealclimate.blogspot.com/2008/07/gavin-schmidt-enso-adjustment-for.html
So far no one has suggested any other observed natural variation that explains the decade long flat trend.
[Response: Well, discarding all the data that don’t support your point is also a classic cherry pick – what happened to the GISTEMP or NCDC? – gavin]
Re: Accumulated Cyclone Energy (ACE) graph for the Atlantic basin
I tried to explain that visual inspection of a graph can be very misleading, and that my professional eye indicated a different conclusion, but that I wouldn’t trust it because it wasn’t coupled to any analysis. I emphasized that this opinion isn’t a knee-jerk reaction, it’s based on decades of experience. I also suggested at least using some analysis, something as rough but simple as moving averages, to get more insight.
While my thesis garnered only agreement, regarding the specific case at hand it seems I haven’t persuaded those who believe the modern era is “just like” the 1950s and 1890s to give serious consideration to the possibility that it might not be the case. Instead I hear “I still maintain your assessment of the graph in question is a gross over-reaction and incorrect” and “I’d be surprised if a 10 year moving average failed to show anything but what the graph appears to show. I don’t have time at this precise moment to do that…”
The data are available here. I calculated 11-yr moving averages. The early peak moving average is 124.9 for the 11-year period centered at 1892; the mid-century averages peak at 122.4 centered at 1953; the lowest value in the trough between them is 58.9 centered at 1915. Peak-to-trough amplitude for the change in 11-yr average for this period of time is therefore 66.0.
The recent peak is 156.9 for the 11-year average centered at 2000. This is 32.0 higher than the highest previous peak, or 48% of the preceding peak-to-trough amplitude greater than the previous maximum.
In fact, the average for the 11-yr period centered on 1999, which excludes the phenomenal 2005 Atlantic hurricane season, is 137.2, still higher than the preceding peak by 12.3, or 19% of the peak-to-trough amplitude.
More sophisticated smoothing methods indicate an even greater difference between recent activity and previous peaks.
Tamino,
No, I don’t think the modern era is “just like” anything.
That would be called “all or nothing thinking” on your part.
Could you answer me a simple question — why do you do that? Why does it seem that otherwise smart people can’t comprehend a periodic signal summed with a linear signal to produce a monotonically increasing periodic signal?
Is this some kind of really difficult thing to grasp? Like, y = mx + b and y = sin(x) somehow don’t go together?
Re: #223 (FurryCatHerder)
Here’s what you said in #191:
The words are yours (the emphasis is mine). Yet you want to accuse me of “all or nothing thinking” and ask “why do you do that?”
You’re not just being absurd. You’re being dishonest.
Tamino,
“Like” is not “exactly the same value”, which is what you complained about. That, while going up and down — which you seem to comprehend the ACE value does — you then said “Ah ha! The high is higher!” Up and down is … up and down. That’s periodic part — “up” like before, “down” like before.
It’s the “down” part you and Reisman have a problem with. [edit]
Please — let’s keep all the context together. It went up, it’s going down. Yes?
Feel free to just, you know, agree with the obvious. 2008 is going to be a less active year in the Atlantic basin than 2005, 2006 and very likely 2007.
[edit]
[Response: I don’t know why this thread seems to have brought out the most pedantic and tedious aspects of conversation, but it is tiresome. Please focus on substance rather than on who said what when. – gavin]
Gavin:
“what happened to the GISTEMP or NCDC?”
Regarding GISTEMP, in the last decade it is diverging from RSS, UAH, and HadCrut3 at a rate of about .13 C per decade. The divergence is more than half the supposed .2 C decadal trend due to AGW. It fails to cover much of Africa, Northern Canada, and Greenland, and the extrapolations at the poles would seem to be dubious in light of the Antarctic cooling. That along with the failure of the network to follow it’s own quality control standards, hinged adjustments that adjust rural sites from the past down, weighting urban station adjustments to rural stations such that stations from 500 to 1000 km use the same range as those from 250 to 500 km, etc, give me little faith in GISTEMP.
But, I’m willing to accept that there is a chance that the others are wrong and that GISTEMP is more correct, and so I will make my statement another way. If the HadCrut3, RSS, and UAH trend lines are the correct ones, and if ENSO adjustments to RSS and UAH lead to similar results as the adjustments to HadCrut3, then we do not seem to have an understood culprit among the elements of natural variation on whom we can blame the absent warming.
Fair enough?
[Response: No. First off you ignored NCDC completely, second HadCRUT3v uses basically the same input data as the other two. GISTEMP does not control any of the station issues you think are important, and you completely ignore the large structural differences in the satellite records. I’m not claiming that I know which (if any) approaches are correct, but picking just the ones that agree with a pre-determined idea of what you want to see is not ‘fair enough’, it is cherry picking. You have to use all the relevant data. – gavin]
If something as simple as a volcano can cut temperature so severely isnt human replication of this possible?
Adam M (227), In theory you are right but the unintended consequences (such as drought) would likely be severe. There is also the problem that the effects of stratospheric SO2 injection only last a couple of years while the effects of CO2 last hundreds of times that long. These issues have been discussed here several times. Search the site for: “geoengineering”.
Re 219 Thanks david. That is a great chart. Now 1t looks to me that since 1900 you have one 10 year peroid with a
downward trend. And that ten year peroid just happens to
be the peroid that is going to be adjusted upward when
the bucket/inlet problem gets fixed.
So go back 100 years, excluding those peroids where volanoes spiked the tempature down, and excluding those peroids where we have data that is still being corrected
( the bucket peroid) can we find some 10 year peroid that is flat or down. I dont know. I’ll have to look.
And when we find one can we tie to an actual weather cycle rather than weather armwaving. again, open question. I dont know. might be fun to look.
Please note it won’t make AGW false, it will just be an interesting thing to understand.
David B. Benson #219
Layman’s question – I’ve looked at your chart (and 1000 others). What is it that explains the warming from ~1920 – ~1940 that does not explain warming since mid century?
John Reisman – “1950 to 1978′ (1942 to 1978) temps were likely due to aerosol pollution.” OK, then what portion of warming since 1978 is due to the reduction of aerosols, soot & pollution rather than AGW?
Thank you both.
Is there any way to get RAW data before everyone does their tricks with the data, like GISTemp data WITH the outliers kept in, or HadCRUT before data made up data from the artic and other areas get added. It really would be interesting to see just the RAW DATA plotted out and then compare the plots to what keeps being shown.
[Response: Yes. Download the GHCN data. – gavin]
“Layman’s question – I’ve looked at your chart (and 1000 others). What is it that explains the warming from ~1920 – ~1940 that does not explain warming since mid century?”
Have you looked at this?
http://en.wikipedia.org/wiki/Image:Climate_Change_Attribution.png
Between 1900-1940 there is notable increasing in solar activity that isn’t present any time after that. There is also a prolonged period of relatively little volcanic activity. (Volcanic activity has a cooling effect) There is also a drop in aerosol cooling through the 30’s which could play a role as well.
“ OK, then what portion of warming since 1978 is due to the reduction of aerosols, soot & pollution rather than AGW?”
Warming caused by a reduction in aerosols can still be AGW. All aerosols do is hide warming for a time, and most of the warming they are currently hiding is almost certainly anthropogenic in nature. Aerosols have been increasing but are not doing so as fast as CO2 levels, so they balanced CO2 emissions for a time but in the mid 70’s CO2 starts to dominate.
#231 T Siefferman:
When you speculate that data is “made up”, it’s axiomatic that someone (me, in this case) is going to ask why you believe it’s made up? Can you show how this is so? Next, how is it made up? Are numbers drawn from a hat? Darts thrown at a dartboard?
I believe if you think about it carefully, you’ll realize you’ve been told that data is “made up” and you’ve accepted what you’ve been told without further consideration.
Further, it’s likely that if you picture to yourself a scientist throwing darts at a dartboard, carefully recording the output and then proceeding to stick his/her neck on a chopping block by publishing the results, you’ll come to understand what an absurd idea has been planted in your noggin.
L. Miller:
Thanks. I’ve seen the chart and your commentary about solar activity 1900-1940 helped me see that.
The chart is not so great; It ends in 1995. And it shows aerosols continuing to force temperature down.
That is a problem because despite your thinking the contrary, aerosols have not been increasing, they are “at historic lows” for some time (~2000). So if you’re going to support the aerosol cooling masking AGW theory, it makes it more difficult to explain the lack of warming that coincides with these “historic lows” does it not?
[Response: Aerosols are only at ‘historic lows’ in the US and Europe. They are at historic highs in Asia – the net effect relatively to previous decades is currently uncertain. – gavin]
#232 L Miller:
True, but I would make the reservation that our knowledge of this solar increase is highly uncertain. It was adjusted downward substantially in AR4.
Furthermore see Figure 9.5a in the AR4 WG1 report. You see that the peak around 1940 pushes the top edge of the ensemble of model simulations. This suggests that a substantial part of the 1920-1940 trend is natural variability in the one realization that is the measured temp curve. Note that for later years, this black curve lies much closer to the ensenble average.
Finally note the accuracy of the measurements-based curve, which degrades before 1940 due to lower station density, among other things. See Brahan et al. 2006.
Gavin:
“You have to use all the relevant data.”
Gavin:
“HadCRUT3v uses basically the same input data as the other two.”
I am using all of the data. I’m simply not using all of the different adjustments of the same data. And I see no reason why I should. I suppose that you could theorize that by using different peoples adjustements that the errors will cancel each other. But of course that’s not verifiable. I also don’t see you making much use of UAH or RSS. So I think that the accusation of cherry picking is unfounded.
And lastly, you have shifted the topic from one that you do not want to deal with, mainly how to explain the decadal flat trend, to one where you can bicker about data set choices.
[Response: You miss the point entirely. The other two datasets don’t have a ‘flat’ trend. Thus your flat trend is not a robust result, and so asking me to explain something that might not even be true is silly. However, the variability in the system is undeniable, and it’s expression the global mean temperature clear. As for my not using the satellite data, that is untrue. I’ve used it when it was relevant. Despite the current fashion for thinking them equivalent to the surface data, I do not confuse the lower troposphere with the surface. They might be related, but as many papers have shown (including Santer et al (2005)) showed clearly, they are not the same. Should I use the satellite data, I would use both records (and maybe Vinnikov and Grody as well) because there are clear structural uncertainties in those products that should not be brushed aside. – gavin]
Reality check.
Once again we suffer from people drawing conclusions (in this case, “flat trend”) based on visual inspection of a graph with no numerical analysis. It’s particularly problematic in this case because not only is the noise not white, it isn’t even AR1 (the usual model used to correct for autocorrelation); the AR1 correction to trend analysis underestimates the uncertainty because the autocorrelation coefficients higher than the 1st decay more slowly than that.
So I computed the trend rate from 1998 to the present (a wee tiny bit more than the last decade), for both ENSO-corrected data series, using the complete formula for autocorrelation correction to linear regression, modeling the AR coefficients as rho_j = rho_1 * alpha^(j-1) (which gives a much better fit to the AR coefficients than the AR1 model rho_j = (rho_1)^j). The results: for ENSO-corrected HadCRU: +0.0014 +/- 0.0172 deg.C/yr (2-sigma); GISS: +0.0134 +/- 0.0180 deg.C/yr. The 95% confidence intervals are, for HadCRU: -0.0158 to +0.0186; GISS: -0.0046 to +0.0314. Both ranges include the oft-quoted modern trend rate +0.018 deg.C/yr.
Of course my model for the autocorrelation coefficients is only an approximation and the numerical estimates of its values are uncertain. But that only ADDS to the uncertainty in the trend estimate, making the error ranges even larger.
The upshot is that when you take autocorrelation into account in a more rigorous way than is usually done, the ENSO-corrected data for the last decade (for *both* GISS and HadCRU) are perfectly consistent with an uninterrupted continuation of the 30-year trend. Yes, they’re both consistent with a trend rate of zero as well — but given the huge uncertainties, this merely underscores the utter folly of making pronouncements about the trend based on a mere 10 years of data.
It REALLY IS a mistake to draw conclusions about trends (or reversals of same) based on small samples, especially when the signal-to-noise ratio is small and the autocorrelation is both strong and complicated. “Your flat trend is not a robust result” is an understatement.
Re #234 Radar
Here’s a more up to date illustration of the forcings driving Earth’s temperature changes, including the estimated trend in aerosolic cooling contributions:
http://pubs.giss.nasa.gov/docs/2005/2005_Hansen_etal_1.pdf (see Figure 1)
http://pubs.giss.nasa.gov/docs/2007/2007_Hansen_etal_3_small.pdf (see Figure 5)
If you can get hold of this article, it also illustrates the forcings of aerosolic components and greenhouse gases.
V. Ramanathan and G. Carmichael (2008) “Global and regional climate changes due to black carbon” Nature Geoscience 1, 221 – 227
http://www.nature.com/ngeo/journal/v1/n4/abs/ngeo156.html
If you can’t access that paper you can read some of the relevant data in Ramanathan’s testimonial to the Wegman hearing:
i.e.: http://oversight.house.gov/documents/20071018110734.pdf
In their Table 2, Ramanathan/Carmichael diagram the contributions from various man-made greenhouse gas (GHG) and man-made aerosols, considering the effect on both the atmosphere or surface:
all GHG’s (CO2, methane, N20, halons, ozone):
atmosphere +1.4
surface +1.6
total +3.0 W/m2
CO2:
atmosphere +1.0
surface +0.6
total +1.6
black carbon (BC):
atmosphere +2.6
surface -1.7
total +0.9
non BC man-made aerosols:
atmosphere +0.4
surface -2.7
total -2.3
There are other bits and pieces that address your question in part. Focussing on the Arctic, a recent paper examines the solar irradiation at the surface (“global dimming”/”brightening”) presumably as a result of atmospheric aerosols. Measured in stations in the high northern latitudes in N. Europe/Arctic, the “surface solar irradiation” averages around 115 W/m2 in 1960 (note that fewer stations were monitored during this period) and decreases progressively to reach a low near 103 W/m2 around 1987 from which it recovers a bit to reach a level near 106/107 W/m2 now, still well below the levels of the early 1960’s. (see their Figure 3).
http://folk.uio.no/jegill/papers/Stjern_etal_2008_IJClim.pdf
and a couple of other papers (I’m sure there must be more!) address changes in surface solar irradiation in Arctic regions associated with changes in atmospheric aerosol content.
S. T. Weston et al (2007) Interannual solar and net radiation trends in the Canadian Arctic Journal Of Geophysical Research 112, D10105
Stanhill G (1995) Solar Irradiance, Air-Pollution And Temperature-Changes In The Arctic, Philosophical Transactions Of The Royal Society Of London Series A- 352 247-258
abstracts/extracts:
[Weston et al (2007)]
Although the trends are not explicitly linear, both data from Alert and Resolute Bay show an overall decrease in K↓ over the past half-century. Data from Alert shows a decrease of 2.25% of the daily mean CI, and Resolute Bay shows a decrease of 2.50% of the daily mean CI per decade. Although further data are needed to tell conclusively, it also appears that both sites show a recent recovery over the past decade. As is speculated by other authors [e.g., Stanhill, 1995; Lohmann et al., 2004; Che et al., 2005], it is most likely that changes in atmospheric constituents (aerosols and/or greenhouse gases) are the major cause. Further study in this area is definitely warranted.
(K↓ is solar radiation; CI is daily “clearness index”)
[Stanhill (1995)]
“A highly significant decrease in the annual sums of global irradiance reaching the surface of the Arctic, averaging 0.36 W m(-2) per year, was derived from an analysis of 389 complete years of measurement, beginning in 1950, at 22 pyranometer stations within the Arctic Circle. The smaller data base of radiation balance measurements available showed a much smaller and statistically non-significant change.
Reductions in global irradiance were most frequent in the early spring months and in the western sectors of the Arctic, coinciding with the seasonal and spatial distribution of the incursions of polluted air which give rise to the Arctic Haze.
I am still trying to work my around the different facets of climate science and have a few questions I would like answers to:
As the net energy levels increase due to AGW (less going out than coming in) is this causing a global mean increase in wind speed? If this is the case would this actually result in a decrease in SST due to increased evaporation?
The last question, and probably impossible to answer, is how much of the heat energy is under the oceans and is it possible to extrapolate a mean ocean temperature and if so the warming trend?
Gavin,
A few years ago (January 2004) I posted a comment that the temperatures seemed to be warming more than the statistics indicate. In summary, your reply was that the statistics are likely correct and that the impression was the result of some unusually warm weather. I accepted your reassurance at the time.
However at the risk of seeming to nag you, this winter where I live (on the Monaro Table land in New South Wales, Australia) is again fueling my suspicion that the warming is stronger than the statistics are saying.
The dam on my property for example. Prior to 4 years ago it froze over more than 5 times per winter. Three years ago it froze over 3 times, last year twice, this year none so far. Prior to 8 years ago snow falls down to 700 meters were not unusual in this area, last year it fell down to about 900 meters for one day only, this year it was down to 1000 meters for 1/2 a day.
My observations also seem to correlate in nearby regions, see: http://www.smh.com.au/news/national/snowman-would-be-stretching-it/2008/07/09/1215282928012.html
Local old time residents to this area tell me that a solid week of snow could be expected on the ground where I live and the local mountain would be snow capped for a month. In addition, strong cold fronts could bring snow right up until the end of December. They recall three Christmas day snow falls (over 50 years). There is no way that would happen now.
It is as though the Southern Ocean has warmed enough to take the bite out of the cold prior to the wind reaching Australia.
The Statistics however are saying that it is only about 1 degree C warmer. That seems like rubbish, 5 degrees C warmer seems to be the more realistic number (over the 30 year average).
My specific question to you is: is it possible that the temperature measurements are somehow under-estimating the temperature increase?
Cheers…
Please advise if the latent heat of ice cooling can come into this on a planetary scale. If ice is melting, shouldn’t the increase in mean global temperature slow or stop while ice is melting, due to the energy absorbed by the melting process?
So, if la nina truly decreases the observed temperatures, does the heat dissolved in the deep ocean reach the treacherous methane clathrates? Well I guess no one knows for certain.
During the holidays, as one reader to another, may I suggest those coming here with questions for the first time
— click the Start Here link at the top of the page
— click the first link under Science in the sidebar
— remember — it’s rare someone asks a question utterly new.
You can look most of these things up.
Use the Search box at the top of the page, limiting it to ‘realclimate’ search for extra creditable information.
Rob (241), I wondered the same thing, but Gavin did the math once in a response somewhere…It doesn’t come to much at all in the big picture.
I’m not sure exactly what your question is, but La Nina only decreases the temperatures observed at the surface . Heat content measurements show it has almost no effect on total ocean heat content.
The treacherous methane clathrates are mostly far from the Indo-Pacific warm pool, which is where the heat is stored during a La Nina. For methane clathrates, I believe the primary concern is the rapid warming of the Arctic Ocean (thus all the concern about the unexpected shrinkage of sea-ice cover in the summers of 2005 and 2007 (2008 reached a sea-ice coverage minimum below 2005 but above 2007)).