Arguments
What you need to know about the NOAA global warming faux pause paper
Posted on 8 June 2015 by dana1981
Last week, a paper out of NOAA concluded that contrary to the popular myth, there’s been no pause in global warming. The study made headlines across the world, including widely-read Guardian stories by John Abraham and Karl Mathiesen. In fact, there may have been information overload associated with the paper, but the key points are relatively straightforward and important.
1. Rapid Global Warming Continues
Arguments about short-term temperature changes only deal with the Earth’s surface temperatures, which account for just 1–2% of the overall warming of the planet. More than 90% of that heat goes into the oceans, and as my colleagues and I noted in a paper published 3 years ago, if anything that warming is accelerating, building up heat at a rate faster than 4 atomic bomb detonations per second.
If you carefully cherry pick start and end dates, you can find a period around 1998–2012 during which the warming of surface temperatures slowed a bit due to temporary natural cooling factors (like more La Niñas), just like it sped up a bit during the 1990s due to temporary natural warming factors (like more El Niños). But these are just wiggles on top of the long-term human-caused global warming trend. As Michael Mann put it,
there never was any “pause” or “hiatus” in global warming. There is evidence, however, for a modest, temporary slowdown in surface warming through the early part of this decade.
2. The Surface Warming Slowdown is Probably Over
This is a tough pill to swallow for those who have misused the short-term slowdown in global surface warming to argue against climate policies, but it’s likely over. 2014 was the hottest year on record, and 2015 looks likely to break the record again.
These slowdown-based anti-policy arguments have been made by everyone fromRepublican presidential candidates to political think tanks to science-denying blogs. It’s a simple argument – if we pretend the surface warming slowdown can continue indefinitely, then global warming is less of a concern and we don’t need policies to stop it.
Since we’ve always known the slowdown was temporary, these were never credible arguments, but they nevertheless helped to delay efforts to curb global warming. After the publication of this NOAA paper, and possibly two consecutive record hot years, it seems unlikely that these arguments will be considered credible any longer.
3. The Most Common Denial Response: Conspiracy Theories
Social science research has shown that conspiracy theorists are more likely to reject scientists’ conclusions about climate change. It’s a logical connection – given that 97% of climate scientists and their research agree on human-caused global warming, the easiest way to deny that reality is to accuse all those scientists of being part of a vast conspiracy. Otherwise it’s difficult to justify rejecting the conclusions of 97% of experts.
Thus it’s not surprising that those in denial are accusing the NOAA scientists of conspiring to fudge the data to make the slowdown disappear. For example,Anthony Watts, who runs a climate science denial blog, wrote an email to one of the authors of the NOAA paper, telling him,
In my last telephone conversation with you, I stated (paraphrasing) that “I believe you folks aren’t doing anything fraudulent, but you are doing what you feel is correct science in what you believe is a correct way”.
After seeing the desperate tricks pulled in Karl 2015 to erase “the pause” via data manipulation, I no longer hold that opinion. You needed it to go away, so you prostituted yourselves, perhaps at the direction of higher ups.
In fact, accusing the NOAA scientists of fraudulently ‘manipulating data’ for the benefit of the Obama administration was a common theme in the climate science-denying blogosphere. But there’s a glaring flaw in this particular conspiracy theory.
4. The Adjustments Reduce Global Warming!
This is clear from the bottom frame in this figure in the NOAA paper.
NOAA global surface temperature changes with new analysis, old analysis, and with and without time-dependent bias corrections. (A) The new analysis (solid black) compared to the old analysis (red). (B) The new analysis (solid black) versus no corrections for time-dependent biases (cyan). Source: Science; Karl et al. (2015) Source: Science; Karl et al. (2015)
According to the raw, unadjusted data, global surface temperatures warmed about 0.9°C from 1880 to 2014. According to the new NOAA analysis, they warmed about 0.8°C during that time. That’s a bit more than in the previous version of NOAA’s data set (0.75°C), but the net effect of these adjustments is toreduce the overall estimated warming as compared to the raw data!
5. The Adjustments are Important
Contrary to the conspiracy theories, climate scientists process the raw temperature data for an important reason – to remove biases that don’t represent real temperature changes. The big one in the new NOAA analysis deals with changes in the ways ocean temperatures have been measured. They’ve been measured from water samples in insulated buckets, uninsulated buckets, from valves in ships’ hulls, and from instruments on buoys. As Zeke Hausfather explains,
A number of studies have found that buoys tend to measure temperatures that are about 0.12°C colder than is found by ships at the same time and same location. As the number of automated buoy instruments has dramatically expanded in the past two decades, failing to account for the fact that buoys read colder temperatures ended up adding a negative bias in the resulting ocean record. This change is by far the largest single factor responsible for changing global temperatures in the past 17 years compared to temperatures found in the prior NOAA record.
While this adjustment happens to have reduced estimates of the 1998–2012 surface warming slowdown, it’s important to account for changes in the way ocean temperatures have been measured. It’s not a conspiracy, it’s science.
John Christy states "“The six bulk-atmospheric temperature datasets agree that essentially nothing [no warming] has happened since the big El Niño of 97-98,” Christy said in an email. “If Karl’s work holds up…, this will only add to the puzzle of diverging surface and atmospheric temperatures which stands in contrast to model expectations". (http://tinyurl.com/qa747cs).
Why the divergence? Which data set is the more correct?
ryland - Christy has a history of, shall we say, interesting graphic decisions, including non-standard and short term baselining that just happens to maximize divergence, averaging UAH and RSS data with 3-sigma differences in trend, not showing state of the art in the difficult to calibrate radiosonde data, testifying in Congress with un-normalized or completely irrelevant data implying conclusions that aren't in the data, ignoring the fact that satellite based mid-tropospheric data has much higher variance (and therefore uncertainty) than surface temperatures, on and on and on. Some of his Congressional testimonies read like the SkS denier myth list - I've highlighted a few of them here.
Having investigated a few of his graphs and opinions (here and here, for example), I take his comments with a wee grain of salt. Your mileage, of course, may vary.
On the other hand, current work such as Sherwood and Nishant 2015 indicates that tropospheric warming is indeed occurring much as predicted. If that work holds up I would expect two primary results:
KR Are you suggesting that Dr Christy is deliberately makig unsupportable claims regarding the RSS data? As for your final point "That ongoing work on the satellite records will turn up additional corrections to the many that have already been made." That is equally applicable to the many corrections made to temperature measurements. Such corrections are central to this current discussion a point, of course, that will not have escaped your attention. Should these corrections also be taken "with a wee grain of salt"? If not, why not?
ryland - Dr. Christy has made any number of unsupportable claims, see here and here for documentation of his Congressional testimonies alone. I won't claim knowledge of what he's thinking, but rather go by his _quite_ public record in this regard.
Yes, corrections and improvements in calibration and consistency continue to be made in all the temperature records. But UAH (Christy and Spencers work) is the only one where the corrections were so large as to reverse recent trends from negative to positive. Quite a distinction. I suggest you look at the comparative scale of those corrections.
The corrections discussed in Karl et al are really quite small in their effects on recent trends. And yet the so-called 'hiatus' vanishes, demonstrating that it wasn't a statistically robust element of the data - rather it appears to be part of natural climate variation. And, I'll point out, entirely consistent with the rest of the evidence, such as ocean heat content and changes in the cryosphere.
---
When you look at the entirety of the evidence, warming and AGW is quite clear. But one of the hallmarks of climate denial is selective evidence, in this context of which temperature record (or which poorly measured portion of the atmosphere) to highlight. It was UAH, then (after they corrected it) HadCRUT3, then UAH again, now RSS, cycling back and forth between tropospheric and surface - whichever temperature record shows the least trend at that moment. That's not skepticism, not a proper consideration of evidence - that's denial.
In similar vein there have been a number of explanations from climate scientists for the so called hiatus as I am sufre you recall. Use of the term denier invariably taints what one would hope would be a civilised discussion of the science. As I expect you know there have been comments from other climate scientists noting that there are flaws in the data and some cherry picking has occurred (http://mashable.com/2015/06/04/global-warming-hiatus-study/)
[PS] Fixed link. This discussion would be greatly improved if members would acknowledge which points raised they agree with and concede, and which they still dispute and why. Avoiding an issue and sliding to another topic smacks of rhetoric and high-school debating tactics rather an investigation of the truth. Ryland, you will notice that KR's original points were about dubious presentation techniques by Christie, not about the development of the UAH dataset.
Ryland, you asked a rather popinted question to KR about Dr Christy making unsupported claims. Fair enough, it is a rather serious matter.
So I am asking you a couple of questions in turn: KR provided documentation that Christy indeed has repeatedly made unsuported statements, most notably to the representatives of the people of the U.S.. Your last post seems to be changing the subject of your previous one so let's get back on track. Do you acknowledge that Christy has repeatedly made unsupported statements?
You said something about cherry-picking. Do you acknowledge that the non-existent pause depended exclusively on cherry-picking the year 1998 has a starting point, despite that year being as far off an outlier as can possibly be? If you disagree with that point, could you calculate trends from the years 1997 and/or 1999 to support your argument?
I find it interesting that the coolest year of the 21st century, 2008, was still warmer than all of the years of the 20th century except for everyone's favorite outlier year, 1998.
Consider the implications; in the 21st century, when supposedly we didn't have any warming, what counts for a "cold year" is actually in the 99th percentile when measured against the warmest years of the 20th century.
Does that sound to anybody like there hasn't been warming in the 21st century?
Just to add some extra information regrding whether John Christy has been selective:
"The six bulk-atmospheric temperature datasets agree that essentially nothing [no warming] has happened since the big El Niño of 97-98"
Firstly, there are seven datasets. Christy is using the mid-troposphere satellite data usually referred to as TMT. But he only uses UAH and RSS, ignoring the third series calculated by Zou et al at STAR/NESDIS. This shows a significantly higher temperature trend than UAH or RSS.
Next he doesn't mention the signficant differences between the radiosonde datasets with some known to have a cool bias due to.
And by simply averaging datasets together he hides the range of results they produce.
Well Glenn, at least he didnt use the upper atmosphere satellite record (which of course shows substantial cooling because of GHGs). What was his rationale for using TMT instead of TLT?
Scaddenp - Yes, I found that a rather curious choice as well, given the _known_ mixing of cooling stratosphere signal in TMT. But consistent with all the other issues with those graphs that IMO add up to a misrepresentation of the data.
As part of the Denial 101 MOOC course I had to select a myth to write an article about. I picked a Christopher Monckton article on WUWT about the pause.
wattsupwiththat.com/2015/06/03/el-nino-strengthens-the-pause-lengthens/
It wasn't a problem debunking the myth - I mainly used info from the Kevin Cowtan video shown on SKS in Dana's post of 3 June. However I was surprised at one aspect of the RSS satellite data shown by Monckton. While 2014 is recognised as the warmest year so far an eyeball of the RSS data shows it quite low, much lower than 1998 and 2010 and probably down to about number 5 or 6. I looked on an RSS site but couldn't find much of use. Can anyone explain this please? Maybe he's using the TMT data too.
BTW the MOOC course was great!
RSS and UAH attempt to measure lower troposphere temps (the lower 4km approx of atmosphere) rather than surface temperature. These temperature parameters shows a much stronger response to ENSO than surface temperature does (colder in La Nina, hotter in El Nino). 1998 and 2010 were El Nino years and 1998 was a monster. You would not expect a record in RSS or UAH unless El Nino is present (there is a few months of lag). El Nino is building. I would expect a new record at a smaller ONI value than 1998. 2015/2016 might be that year.
SkS has a useful temperature trend tool for just this sort of thing. Using Christy's own UAH data, and cherry picking 1998 as the start date, there's still a postive trend (most likely 0.072°C/decade). If we instead choose 1999 as the start date, the trend is twice as big, at 0.145°C/decade.
So, all I can say is that Christy is spouting nonsense (that's about the nicest way to put it), and as one of the keepers of the UAH data set, he should know that it's nonsense.
To compare satellites, radiosondes and models you need an apples with apples comparison.
I believe he reversed applied the weighting functions used for the satellite channels to convert model output to 'what a satellite would have seen'. Hopefully he did the same with the radiosonde data. If he had tried to match to the lower troposphere channel (TLT) he would have also had to reverse apply the off-nadir readings and differencing algorithm used to synthesis TLT from the TMT channel - TLT isn't a direct measurement but one derived from extra processing of the TMT channel.
BC @11, one of the features of satellite (and radiosonde) data is that it responds far more strongly to ENSO effects. It also responds more strongly to volcanic effects, and may also respond more strongly to other short term fluctuations but I have no data on that. The upshot is that satellite and radiosonde data is far more noisy, resulting in a greater standard deviation of the data. That in turn requires a larger underlying trend for it to be considered statistically significant over short periods, or (alternatively put), for given trend, it requires a longer period of data before it is statistically significant. Given that the denier test for "no warming" is "the trend is not statistically significant from zero" regardless of whether it is statistically significant from IPCC projected warming, it becomes almost a gimme for satelite data to show "no warming" in the deceitful terminology of deniers.
All that is basic. Less frequently commented on is something very interesting that has been happening with the Southern Oscillation Index (SOI). The SOI measures the pressure difference between Darwin and Tahitti. Because winds flow with pressure differences, that makes it a good proxy of the strength of the Walker Circulation, which drives the El Nino Southern Oscillation (ENSO). Essentially, if the pressure is greater in Tahitti than in Darwin, the Walker circulation blows surface Pacific waters towards the Pacific Warm Pool in waters near Indonesia. If they are only slightly stronger, the Walker Circulation is weak and you have normal conditions. If the pressure in Tahiti is much greater, you have an unusually strong Walker Circulation and a strongly positive SOI, and a La Nina results. In contrast, if the pressure falls in Tahiti relative to Darwin, the Walker circulation stops and you have strongly negative SOI values, along with an El Nino.
The interesting thing that is happening in the SOI is that it has been systematically increasing since the early 1990s. Because the SOI sign is the reverse of the global temperature effect, and because the peak temperature effect is felt about six months after the peak SOI effect, I have graphed the inverted, seven month lagged SOI below. (Seven month lagged rather than six as that allows me to include a full 2015 value.) The result should show the sign and relative magnitude of ENSO effects on global temperature over the years since 1950:
Another way of putting that is that if ENSO was the only factor affecting the Global Mean Surface Temperature (GMST), a scaled version of the above graph would show the actual fluctuations in GMST.
Obviously ENSO is not the only thing effecting GMST. For instance, in 1982 the El Chichon volcanoe erupted in Mexico, the effect of which reduced 1983 to an average year rather than the warmest on record (to that time). Likewise the Pinatubo eruption in 1992 cooled what was otherwise going to be a very warm year in 1992. Pinatubo's effect had largely dissipated by 1995, however, which had the 13th strongest ENSO warming effect since 1877, and the 4th warmest in the satellite era. Notably, every year in the 21st century has had a lower ENSO warming than 1995, and there is a consistent downward trend in that warming from 1998.
Absent global warming and volcanic effects, the various years in the satellite era would have ranked as follows for warmth:
Year SOI rank (inverted)
1979 21
1980 13
1981 15
1982 26
1983 1
1984 22
1985 24
1986 17
1987 7
1988 11
1989 35
1990 20
1991 12
1992 3
1993 8
1994 5
1995 4
1996 25
1997 23
1998 2
1999 34
2000 32
2001 29
2002 16
2003 9
2004 18
2005 6
2006 27
2007 10
2008 31
2009 33
2010 14
2011 36
2012 30
2013 19
2014 28
You will notice that 2014 (a tied record year in GMST) would have only ranked 28th absent the effect of global warming and volcanoes. 2010 ranked 14th on ENSO warming, and just four years earlier, was also a statistical tie with 2014. Both years were warmer than the denier poster year, 1998 (ranked 2nd on ENSO effect, or first if we include volcanic effects as well).
Finally, the point of all this is that satellite record are (as stated at the beginning) far more effected by ENSO warming than are the surface records. Because of the radical difference between actual warmth as measured by GMST and ENSO only rankings, we therefore expect a significant disparity in ranking between surface and satellite records. In particular, early strong El Nino years will rank far higher relative to the later weak El Nino's that have been breaking records due to global warming.
dana @13, I believe the SkS tool still uses v5.6 of the UAH TLT pseudo channel. As you know Spencer and Christy are in the process of developing v6, which in beta release far more closely matches RSS. As presumably Spencer and Christy would claim that v6 is superior to v5.6, I don't think we should be using v5.6 to argue what "Christy's own UAH data" shows.
I'm curious to know if Karl's et al corrected data will be available as an option on the SkS trend calculator?
Thanks for the explanation (Scaddenp @12 and Tom Curtis @ 15) for outlining this difference between surface temps and satellite readings and why 2014 wasn't a record year according to satellites.
Also, because satellite readings are more sensitive to the El Nino and volcano contributions they'll show up more of a hiatus than the surface readings, which would be why Monckton is using them.
The general tenor of the comments about Dr Christy, UAH and RSS is hardly complimentary and in my opinion some, in particular that from Dana, come close to accusations of malpractice. Why therefore is Dr Christy and for that matter Dr Roy Spencer, still employed at UAH in a fairly high profile and senior position if he is as scietifically unsound as has been claimed? (http://www.skepticalscience.com/examining-christys-skepticism.html). Does UAH have particularly low standards for their climate scientists? It also seems surprising that Dr Christy is still able to publish in the peer reviewed literature if his reputation is as poor as is implied here.
[PS] This site exists to debunk misinformation by fake skeptics. You have been pointed to the exactly what has been claimed and what the science actually says. What is your opinion?
ryland: Christy has tenure, so he can do and say nearly anything without any effects on his employment. Both Christy and Spencer do some good work. In particular, what they write in their peer-reviewed publications tends to be reasonable. It's what they write and say outside of those peer-reviewed publications that is the main problem. Contrast their behavior with that of RSS's Carl Mears.
Ryland...Tenure is there to protect free speech from exactly this kind of threat. You want free speech so there is a battle of ideas that eventually leads to the truth. There will always be contrarians in any field — they serve a useful purpose by keeping ideas alive that may still be useful or injecting new ideas. To silence such voices would have a chilling effect on discourse over all.
For me, the problem is not Spencer and Christy per se, but rather the apparatus that exists to amplify their more contrarian views beyond their actual influence within the climate science community. That amplification brings in a lot of voices that do not have perspective and context to understand information, and who consequently interpret information to match their own fears, predipositions and narrow vested interests and political aims.
It ends up distorting the debate entirely. It ends up being about who has the power rather than who has the most rational argument. Look at attempts to cut Geosciences funding within NSF with the current House and Senate budgets, for example.
Thanks for your comments Tom Dayton and Simon Baines. I too was a tenured University professor at an Australian univesity where similar rules apply but not always as there have been professors notably Murray Salby who have been sacked for views not thought suitably concordant with mainstream climate science. Tom Dayton I was surprised at your statement "Both Christy and Spencer do some good work. In particular, what they write in their peer-reviewed publications tends to be reasonable" as most pieces commenting on these scientists rarely say any such thing. And Stephen Baines I thought your statement "For me, the problem is not Spencer and Christy per se, but rather the apparatus that exists to amplify their more contrarian views beyond their actual influence within the climate science community" showed far more insight than that shown bny many others who comment on Christy and Spencer.
ryland, the problems with Spencer's and Christy's aggressive public statements include very vocal, horribly biased, unjustified and even outright wacky claims about what their own peer-reviewed publications say.
ryland, #22,
Professor Murray Salby was dismissed by MacQuarie University for "not fulfilling his contract obligations", and for repeated overseas travel when he had been asked to stay and teach. Salby would still be at the University, climate views and all, if he had played ball with the administration.
A Salby defender steps back
FWIW, it subsequently emerged that in 2005, the US National Science Foundation opened an investigation into Salby's federal funding arrangements and found that he had displayed "a pattern of deception [and] a lack of integrity" in his handling of federal grant money.
Wikipedia on Murray Salby
Tom Curtis @16 - you're right, I'd forgotten that UAH are working on a new version of their data set, and that may be what Christy is referring to. Though it if does show no lower troposphere warming since 1998, I would seriously question its reliability.
As for Christy in general, as others have said, he's published some good (and some bad) research papers. The problem lies in his public comments, which often aren't supported by research, aren't accurate, and misinform a lot of peple. And in the media that amplifies that misinformation, and in the members of Congress who keep inviting him to testify despite (or because of) his repeatedly misleading testimony.
dana1981 - It's my understanding that V6 of the UAH data uses a completely new method for calculating lower tropospheric temperatures, resulting in lower LT trends. Dr. Spencer describes the differences as follows:
There are also differences in how they calculate diurnal drift adjustments for the satellites, particularly NOAA-15.
The previous method for LT temperatures used multiple angles with individual weighting functions, the new method uses a combination of multiple channels at a single location to derive LT temperatures - with a pre-processing step of averaging all view angles.
As such, I believe that V6 LT temperature estimates cannot be directly compared to V5.6 or earlier, as they are not the same computation at all. The LT code previously released is now outdated, and it remains to be seen when/if the current code will be released.
Given past actions by the UAH team (including the Spencer/Braswell 2011 low sensitivity debacle leading to editor resignations) I find myself a bit suspicious of the reduced LT trend; we'll have to wait and see how it stands up to examination. In the meantime it would seem imprudent to be making policy decisions on a new and untested methodology.
dana, KR, for what it is worth, here is a comparison of UAH v6 to v5.6 and to RSS for the TLT pseudo channel:
As you can see, the major effective difference between v5.6 and v6 is that v6 much more closely matches RSS. That means it is not possible reject v6 as inadequate without also rejecting RSS as inadequate, ie, without effectively rejecting satellite temperature data altogether. While the appropriate caveates about satellite data are often ignored by deniers, I do not think rejecting it altogether is a tenable position.
The change in method of calculating the psueudo channel mentioned by KR changes the pseudo weighting profile, making it higher in altitude to that used in v 5.6 and by RSS:
This would have its greatest effect on measured temperatures in the tropics and least in polar regions due to differences in lapse rates between the two regions. Despite that, the largest difference in trends between UAH v5.6 and v6 is in the Arctic:
That combined with the greater sensitivity to ENSO displayed by v6 makes me suspect that primary difference is a reduced coverage of the Arctic to match that by RSS, but I have been unable to confirm that (gridded data being rather intractable on spreadsheets). It may be just a result of the change of diurnal drift corrections to effectively match those of RSS. Spencer's introduction of v6 takes no effort to explain the differences between measured temperatures between the two products which I consider very poor scientific practise. Hopefully the reviewers and editors of the paper introducing the changes will expect more of him.
Ryland I note that you have not answered any of my questions. The SkS contributors are familiar with Spencer & Christy's published work. There is no reason to be surprised by Tom Dayton's statement. The work has passed peer-review and is of some interest to the overall understanding.
That work does not show, however, anything that deviates significantly from the rest of the body of litteature on similar sujects. The poblem that I, KR, and others (including Tom Dayton, I believe) have is with the vast abyss that separates their published work from their public statements, especially those made to Congress. Nothing in S&C publications allow to support the statements linked above. So I ask again, do you acknowledge that Christy made in congressional testimonies statements that are not supported by the weight of the scientific evidence, including his own research? If not, why not?
I did not see a response either to my other question about cherry-picking the year 1998, so I ask again: what are the "trends" starting from 1997 or 1999, or, for that matter 1996 and even 2000? Do they indicate the presence of a "hiatus"? How different are they from the "trend" that starts in 1998? What objective reasons based on statistics justify to select 1998?
PhilippeChantreau@28 Sorry to have made you rather peevish but I'm on holiday in the Baltics at the moment and access to the internet is intermittent plus don't consider accessing SkS every day as obligatory. My answer to your question on Christy's comments to congress is I do not know the detail of every statement Dr Christy has made to congress. That said Nature editorialised that extreme weather should not be linked to Climate Change and this was echoed by Dr Christy in his comment "Extreme events, like the recent U.S. drought, will continue to occur, with or without human causation’ This does not seem insupportable. Neither does his comment "‘As a result what passes for science includes, opinion, arguments-from-authority, dramatic press releases, and fuzzy notions of consensus generated by preselected groups. This is not science’ or "Climatologist Dr. John Christy: ‘Oil & other carbon-based energies are simply the affordable means by which we satisfy our true addictions – long life, good health, plentiful food…’‘…internet services, freedom of mobility, comfortable homes with heating, cooling, lighting and even colossal entertainment systems, and so on. Carbon energy has made these possible’
With regard to "cherry picking starrt dates using 1997 as a start the UK Met Office stated "The linear trend from August 1997 (in the middle of an exceptionally strong El Nino) to August 2012 (coming at the tail end of a double-dip La Nina) is about 0.03°C/decade, amounting to a temperature increase of 0.05°C over that period, but equally we could calculate the linear trend from 1999, during the subsequent La Nina, and show a more substantial warming" (WUWT http://tinyurl.com/8p5zs9o).
From 1999 "Over this interval 1999-2010 the warming trend is actually larger than the long-term trend of 0.175 °C per decade. Yet it is not statistically significant" (From RealClimate http://tinyurl.com/pzvxlhm)
I doubt I shall be returning to this thread for some time as I am heading for France and intend to spend days sampling the food and wines of various regions so my apologies for my less than perfect attempts to allay your peevishness.
[JH] Your snarky rhetoric is neither appreciated nor welcome on this website.
ryland - I believe I pointed you to documentation on just how Christy mislead Congress in earlier messages; you can find the details there. His comments on extreme events were 'documented' by national averages, a rhetorical trick considering that increases in both high and low precipitation and other events with strong regional effects (such as the California drought) are rather hidden by averaging them. His unsupportable comment regarding the quality of climate science marks him as a Dismissive, and his last comment presents a false dichotomy fallacy implying that only fossil fuels can be beneficial, which is utterly incorrect - renewable sources can provide energy without the associated costs of climate change.
Overall, that spans a great deal of Christys behavior, climate change denialist behavior I described in my first comment on this thread. You've presented nothing contradictory, rather, your references support my comments.
On a completely separate note, please enjoy your vacation. French cuisine is delightful.
Thanks for your good wishes KR they are much appreciated.
I note that you still did not really answer the question about reasons to choose 1998 as a starting point. I also note that data are available until 2014 and there is no reason to stop in 2010 or 2012, so trends should be computed to the end of the data. We should also examine how statistically significant any "trend" starting less than 30 years ago actually is. If you are not aware of the what Chrsity said, then you should get acquainted with it so as to be able to form a truly informed opinion. Links were graciously provided and what you quoted is not exactly what others and myself have referred to.
Saying that no extreme events can be linked to climate change is a phony rethoric tool. No individual event can be linked to climate change, but it is well known that climate change increases the probability of extreme events and their severity, especially those that have to do with sea level or heat waves.
I read my post again and they were rather courteous and well formulated, I'm not sure what you perceive as peevish.
[JH] Ryland has been warned that his snarky rhetoric will not be tolerated.
At the risk of further aggravating the situation which I have absolutely no desire to do, to answer your question you wrote"Ryland I note that you have not answered any of my questions." and "I did not see a response either to my other question about cherry-picking the year 1998, so I ask again: what are the "trends" starting....". In retrospect I should not have used the term peevish and apologise for that. I should have written you seemed rather exasperated with my lack of response which perhaps has less unfortunate connotations than peevish even though both words mean the same thing namely "annoyed".
I am not exasperated, Ryland. I am probing because I want to know if the right answer can be acknowledged. Not everything is a matter of opinion. It is possible to narrow down a given question so that the right answer can be clearly established.
In the case of 1998, it is abundantly clear that it is the worst year to choose as the beginning of a sample because it is such an outlier, due to the massive El-Nino that happened across 1997-1998. What exasperates me is that there has been so much noise made around the "pause" that even serious people give credence to an idea that has in fact no basis in reality. If the year is placed in context, it is painfully obvious that one must not use it as the beginning of any sample. Furthermore, no sample too small to identify a truly statistically significant trend should be used at all. These are objective truths. If one wants to look at even something of as little value as the "most recent trend", he should start from the lastest data (2014) and go back far enough to have a sample that will yield a valid trend. How big a sample that is is not a matter of opinion either. It can be determined by statistical means. These are facts. If one fails to do that, he just reveals how fake his "skepticism" is.
I am interested to see if even one whose ideological leanings appear to be the kind for which climate change is a very inconvenient problem can acknowledge these objective facts, i.e. think rationally. Can you?
The same applies to Christy's declarations. One can agree wholeheartedly with the ideological convictions that compel him to participate in certain organizations. However, the distance between what the weight of the evidence shows (including the one he himself produced) and what he said in Congress is also an objective fact. One can decide to not think rationally and intentionally overlook this fact, of course. However, it remains plain to be seen for those who do think rationally. And it is not a valid argument to point attention on to others who supposedly do similar or worse things. The question at issue is clearly posed, other issues are other issues. On this specific question, can you think rationally?
Doing so inevitably forces one to accept a great numbers of things about climate change. What you do with them within the framework of your favored ideology is up to you, but if the ideology can not accomodate the reality, then the ideology must be amended.
In the old Soviet Union, they pretended that, in their camp, airplanes did not crash. In a certain current of thought nowadays, some pretend that temperatures are not increasing, that see level is not rising, that misleading arguments are honest, etc. What's wrong with this picture?
Phillippe @34
The most eregious use of "the pause" I have come across is by UK contrarian Peter Taylor here
Another five years of no warming will nail it. Unless a major Nino or volcano muddies the water!
Taylor clearly believes that is OK to start a trend in an El nino year, but not to end with one.
That is not so clear from the quote, in fact, since he does mention a volcano as well. At first glance, the quote suggests that either event would make the data more difficult to interpret, which is understandable. I'm not sure what his meaning exacly is, and I don't have the time or inclination to explore it. There is plenty of interesting reading with which one can use his time.
Phil @ 35
Ahhh! Peter Taylor!
My Garbage In - Garbage Out filter had managed to delete all memory of him until you triggered some residual linkage. About 5 years ago, I borrowed the book he casually mentions from the local library - and virtually pissed myself laughing.
It comes highly recommended if you want a good giggle. As for rigourous, objective intellectual content, it's right up there with Erich von Daniken and Immanuel Velikovsky.
Why did they decide that ship-based measurements are more accurate than buoy measurements? I assume the buoy system was thoroughly tested, but when the measurements were .12C colder than ship-based measurements, they chose to correct the buoy measurements. How did they decide that the buoys are wrong?
mwsmith12, the researchers' adjustments removed the systematic differences between ship-based measurements and buoy measurements. The effect on the temperature trend over time is the same regardless of whether you change the buoy measurements to make them consistent with the ship measurements, or change the ship measurements to make them consistent with the buoy measurements. That's because for the purpose of determining temperature trend, the absolute temperatures are irrelevant; only the relative temperatures matter.
Tom Dayton, I understand why trend and relative temperatures are what is important and that the effect on trend is te same regardless of which way the systematic difference is removed, but if they knew, hypothetically, that the buoy measurements are more accurate, would they still remove the systematic difference by adjusting the data that are more correct? Did they adjust the buoy data because they know it is wrong and, if so, how do they know it is wrong? But if they don't know it's wrong, and if they don't know which data are wrong, why remove the systematic difference at all?
mwsmith12, temperatures are used for trend analysis by determining a temperature for each geographic grid square. In a given square at a given time, sometimes the only temperature will be from a single buoy, sometimes from a single ship, sometimes from combinations of buoys and ships. For that square, the temperature from one time to the next time would be biased if at time 1 the temperature was from a buoy and at time 2 from a ship, or vice versa, or any other different combinations of sources of the measurements varying across time. Adjusting the buoy and ship measurements to remove their systematic difference, removes that bias.
mwsmith12,
Even if they know that the bouy data is more accurate it makes less of a change in the entire record to change the bouy data. All the data has been previously adjusted to a single baseline with the ship data. If they adjust the bouy data they only need to change the last 15 or so years of data. If they accept the bouy data they have to adjust the rest of the record down so it looks like a bigger change.
Since they are only interested in the trend and not the absolute temperature it makes more sense to adjust the bouy data to match the old data. In general, the record is not used to measure the absolute temperature of the Earth, only the change.
mwsmith12 @38, global temperatures are presented as anomalies (ie, variations from the mean value) of a thirty year period. From the zero value they appear to have used a 1971-2000 anomaly, although 1951-1980 and 1981-2010 anomalies are also commonly used. Had Karl et al taken the buoy data as accurate and adjusted all ship based data accordingly, that would make no difference to their having done the opposite because the anomaly adjustment would have eliminated any difference. Indeed I see no evidence that that is not what they did.
Everyone:
Please note that there is a user "Martin Smith" asking much the same questions over at RealClimate. This is the answer I posted over there (currently awaiting moderation).
But it's because the thermometers move from grid to grid, and the record for a particular grid can include both types of measurment. If buoys were used in the Pacific, and ships were sed in the Atlantic, there would be no interference so they wouldn't have to be adjusted.
This is important because it illustrates how contrarians leverage these details into doubt and suspicion.
This is a quote from post by Patrick Michaels, Richard Lindzen, and Paul Knappenberger over at WUWT. I don't read that site, but posts from there get used everywhere, so I had to deal with this one:
"In addition, the authors’ treatment of buoy sea-surface temperature (SST) data was guaranteed to create a warming trend. The data were adjusted upward by 0.12°C to make them “homogeneous” with the longer-running temperature records taken from engine intake channels in marine vessels.
"As has been acknowledged by numerous scientists, the engine intake data are clearly contaminated by heat conduction from the structure, and as such, never intended for scientific use. On the other hand, environmental monitoring is the specific purpose of the buoys. Adjusting good data upward to match bad data seems questionable, and the fact that the buoy network becomes increasingly dense in the last two decades means that this adjustment must put a warming trend in the data."
http://wattsupwiththat.com/2015/06/04/noaas-new-paper-is-there-no-global-warming-hiatus-after-all/
The authors are trying to convince me that homogonization is cheating. They put it in quotes. Well, I already know why homogenization is important and how it improves data qaulity, but even so, why add 0.12C to all the buoys when the buoys were designed for this and should be more accurate than measurements from ships that are just trying to get from A to B as fast as possible?
And despite my being a career software engineer who at least reads a lot of climate science reportage and the occassional paper, and who understands the scientific method and the concepts of statistics etc, I was still left with some doubt because it isn't clear to me why data that seem to me to be more accurate are adjusted up to be compatible with data that seem to me to be less accurate.
Having now understood the explanation, I can't believe the three authors of the WUWT post don't know it already. And if so, their objection is disingenuous at best. But it would be very useful to the general public if each scientific paper could have an accompanying link to a page on which these explanations are provided, together with the perhaps bogus objections that require them.
Maybe a section here at Skeptical Science, where these papers are catalogued toghether with all the contributed explanations for questions like: Why did we add 0.12C to the buoy data, and why is that the right thing to do?
mwsmith,
The data need to be adjusted because the temperature of the ocean used to be measured using ship intake tubes and now is measured with buoys and ships. If all the data was currently collected using buoys the data would need to be adjusted because the older data is higher than it should be due to calibration issues. Whenever there is a change in the method of collection adjustments have to be made.
Much larger adjustments were made when canvas buckets were switched to wooden buckets and when the change was made from wooden buckets to engine intakes. Those adjustments were made by adding temperature to the older measurements. This lowers the measured temperature trend, The overall adjustments lower the measured trend for the entire globe about .1C, but they will never say that at WUWT. Imagine how many adjustments have to be made in the satalite record where the calibrations change with each new satalite, as each satalite orbit decays and as the time of day changes.
You are correct that the scientists you report at WUWT know that they are disingenious at best. I see no way that all data can be reported in such a way that people who want to lie cannot find a way to decieve others. This site is dedicated to countering those lies by providing correct explainations.
In the end, the record temperatures for 2014 and 2015 will convince many people that AGW is real. The record floods in Texas that ended their record drought will convince others (since AGW predicts both more drought and more floods).
michael sweet,
The authors of the WUWT post claim that adding .12 to the buoy data it must put a warming trend in the data:
"Adjusting good data upward to match bad data seems questionable, and the fact that the buoy network becomes increasingly dense in the last two decades means that this adjustment must put a warming trend in the data."
http://wattsupwiththat.com/2015/06/04/noaas-new-paper-is-there-no-global-warming-hiatus-after-all/
That's right, isn't it? As more buoys are added, more artificially high readings are added.
mwsmith12 @48:
1) Just rehashing, because the temperature series is an anomaly series, it made absolutely no difference to the end result whether Karl et al added 0.12 to the buoy data, or subtacted it from the ship data. For simplicity of illustration, assume there was no buoy data till 2001. Supose then we subtracted the bias correction from the ship data. It follows that the temperatures durring the anomaly period (1971-2000) would have been reduced by 0.12 C. However, recalculating the anomaly would reset the average over that period to 0 C, thereby adding 0.12 C to all values to obtain the anomaly value. For shipboard measurements, the two adjustments would cancel, but for buoy measurements there would be a net 0.12 C adjustment regardless of whether you initially adjusted shipboard or bouy data. The only difference is whether you include that change to the buoy data as an explicit adjustment or implicitly in caclulating the anomaly period.
You already get this, and so should everybody else including the authors of the WUWT post because (excluding the sheer number of values involved) this is simple arithmetic. Put another way, if Michaels, Lindzen and Knappenberger are not being completely disingenuous, they are demonstrating that fourth grade arithmetic is above their payscale.
2) Yes, adding more buoys will increase the temperature but only because it removes an artifact produces by the increase in the number of buoys while incorrectly assuming that shipbourne and bouy temperature measurements were not biased with respect to each other.
3) From the paper:
(My emphasis).
And from the supplementary material:
So, contrary to the Michaels, Lindzen and Knappenberger's argument, the buoy data was given greater weight because it was more accurate. Once we ignore Michaels, Lindzen and Knappenberger's smoke and mirrors trick (ie their failing grade in arithmetic), the paper in result in fact gives more weight buoy data which increases the measured trend.
wmsmith:
Karl et al have measured that all the temperatures measured before the buoys were 0.12C too high (presuming that the buoy data is correct). In order to be consistent they have to either:
1) Subtract 0.12 from the entire record previous to the buoy data or
2) Add 0.12 to the buoy data.
Changing the buoy data means that it is easier to compare the current record to papers published in say 1995 or 2005. It also means that less of the record is affected.
The old record has already corrected all the old data so that it lines up with the ship data. This means all of it has to be changed, not just the past 50 years or so.
The people at WUWT will complain no matter what is done.
Note again: if the raw data is used the global increase in temperature is 0.1C greater than using the corrected data, but WUWT will not tell you that.