From the “watch the pea under the thimble” department, the IPCC appears to have secretly changed the definition of what constitutes ‘climate’ by mixing existing and non-existing data
By Dr. David Whitehouse, The GWPF
The definition of ‘climate’ adopted by the World Meteorological Organisation is the average of a particular weather parameter over 30 years. It was introduced at the 1934 Wiesbaden conference of the International Meteorological Organisation (WMO’s precursor) because data sets were only held to be reliable after 1900, so 1901 – 1930 was used as an initial basis for assessing climate. It has a certain arbitrariness, it could have been 25 years.
For its recent 1.5°C report the IPCC has changed the definition of climate to what has been loosely called “the climate we are in.” It still uses 30 years for its estimate of global warming and hence climate – but now it is the 30 years centred on the present.
There are some obvious problems with this hidden change of goalposts. We have observational temperature data for the past 15 years but, of course, none for the next 15 years. However, never let it be said that the absence of data is a problem for inventive climate scientists.
Global warming is now defined by the IPCC as a speculative 30-year global average temperature that is based, on one hand, on the observed global temperature data from the past 15 years and, on the other hand, on assumed global temperatures for the next 15 years. This proposition was put before the recent IPCC meeting at Incheon, in the Republic of Korea and agreed as a reasonable thing to do to better communicate climate trends. Astonishingly, this new IPCC definition mixes real and empirical data with non-exiting and speculative data and simply assumes that a short-term 15-year trend won’t change for another 15 years in the future.
However, this new definition of climate and global warming is not only philosophically unsound, it is also open to speculation and manipulation. It is one thing to speculate what the future climate might be; but for the IPCC to define climate based on data that doesn’t yet exist and is based on expectations of what might happen in the future is fraught with danger.
This strategy places a double emphasis on the temperature of the past 15 years which was not an extrapolation of the previous 15 years, and was not predicted to happen as it did. Since around the year 2000, nature has taught us a lesson the IPCC has still not learned.
With this new definition of climate all data prior to 15 years ago is irrelevant as they are part of the previous climate. Let’s look at the past 15 years using Hadcrut4. The first figure shows 2003-2017.
It’s a well-known graph that shows no warming trend – except when you add the El Nino at the end, which of course is a weather event and not climate. The effect of the El Nino on the trend is significant. With it the trend for the past 15 years is about 0.15° C per decade, close to the 0.2 per decade usually quoted as the recent decadal trend. Before the El Nino event, however, the warming trend is a negligible 0.02° C per decade and statistically insignificant.
The second graph shows the 15 years before the recent El Nino, i.e. 2000-2014. The trend over this period is influenced by the start point which is a deep La Nina year. Without it the trend is 0.03 °C per decade – statistically insignificant. Note that there are minor El Ninos and La Ninas during this period but they tend to have a small net effect.
So which does one choose? The El Nino version that leads to 0.6° C warming over the 30 years centred on the present, or the non-El Nino version that suggests no significant warming? The latter of course, because the trend should be as free from contamination of short-term weather evens — in the same way as they are free from decreases caused by aerosols from volcanoes blocking out the sun and causing global cooling for a while.
The same problem can be seen in the IPCC’s 1.5C report when it analyses the decade 2006-2015 which it does extensively. In this specific decade 2015 is significantly warmer than the other years, by about 0.2°C. NOAA said, “The global temperatures in 2015 were strongly influenced by strong El Nino conditions that developed during the year.” The temperature trend including the El Nino year of 2015 is 0.2°C, that future again. Without the El Nino the trend is statistically insignificant.
To see the future temperature and climate the IPCC envisage in their report consider their Summary for Policy Makers figure 1, (click on image to enlarge.)
The IPCC’s attempt to move the goalposts is highly questionable. Non-existing data extrapolated for assumed temperature trends over the next 15 years should not be part of a formal definition of what constitutes climate.
UPDATE: From the IPCC Summary for Policymakers report, page 4, there is this footnote that defines the IPCC’s erroneous thinking. – Anthony
Someone needs to point out to these so-called scientists that by changing the time period these temperature measurements are averaged over they have made ALL the previous data not relevant to the previous data and whatever conclusions they may or may not have tried to discern in those series. If the most recent data is now ‘smoothed over 15 years it must NOT be compared to any data smoothed over 30 yrs. In other words, to be useful for comparison all the data must be smoothed over the same time frame.
Oxygen Isotope ratios have clearly shown that it has been much warmer in the past. On top of that CO2 ability to create heat is Logarithmic.
You can have as much as you like.
The more we get the greener the planet gets.
Worryingly: We are heading into Global Cooling….Solar Physicists have been predicting it for sevetal years. It started in 2013, the Jet Stream is going south….Sun Spots disappearing.
A good example of creative accounting by the IPCC
See also their carbon cycle balance assuming exclusive impact by humans
https://tambonthongchai.com/2018/05/31/the-carbon-cycle-measurement-problem/
And the energy balance assuming exclusive impact of human CO2 although past climate change shows large geological impacts on both carbon and heat balance.
https://tambonthongchai.com/2018/10/28/petm/
30-year trends (K/century) with 95% confidence intervals (corrected for auto-correlation):
1.76 +/- 0.37 1/1988 to 1/2018 Starts in El Nino year
1.71 +/- 0.34 1/1986 to 1/2016 Ends in El Nino year
1.73 +/- 0.32 1/1983 to 1/2013 Starts in El Nino year
1.86 +/- 0.32 1/1978 to 1/2008
1.87 +/- 0.38 1/1973 to 1/2003
30-Year trend are only modestly effected by presence of El Nino year. There 95% confidence interval is about 20% of trend.
15-year trends (K/century) with 95% confidence intervals (corrected for auto-correlation):
1.69 +/- 1.06 1/2003 to 1/2018
0.52+/- 0.77 1/1998 to 1/2013 Starts in El Nino year. Warming not statistically significant.
2.28 +/- 0.85 1/1993 to 1/2008
2.12 +/- 0.99 1/1988 to 1/2003
1.71 +/- 0.96 1/1983 to 1/1998 Starts and ends in El Nino year
1.38 +/- 0.90 1/1978 to 1/1993
1.69 +/- 1.16 1/1973 to 1/1988 Ends in El Nino year
-0.30 +/- 0.97 1/2002 to 1/2013 Cooling or absence of warming not statistically significant.
15-Year trend are dramatically effected by presence of El Nino year. 95% confidence interval is about 50% of trend. In other words, these trends are highly uncertain. If one cherry-picks the right starting and ending years from dozens of possibilities, the 95% confidence interval no longer means that roughly 19 out of 20 experiments will fall within this range. Cherry picking makes confidence intervals meaningless.
”Cherry picking makes confidence intervals meaningless.”
Exactly!
And it’s almost impossible not to cherry pick!
Mike: If you noticed, I started with the 15-year period chosen by the IPCC (1/2013-1/2018) and systematically stepped it backwards five years at a time. Then I chose 30-year periods and did the same thing covering the same period. I looked up the trends, ONLY after designing the experiment. No cherry-picking.
Then I added a 30-year period ending with the recent, El Nino (1986-2016) a situation not explored above, to further validate my conclusion that El Ninos didn’t distort 30-year trends. I should have put that entry in a separate category.
Then I cherry-picked an 11-year period I already knew showed with slight cooling to illustrate how misleading that could be.
To be completely honest, I’d need to admit I cherry-picked how far back in time I went (1973). When I went back to 1968, the 30-year trend began to fall because the 1945-1975 Pause becomes part of the data. However, if you look at the rate of increase in radiative forcing linked below, you’ll see a dramatic increase in the rate of radiative forcing beginning about 1970 from about 0.7 W/m2/century to 4 W/m2/century. A post-hoc rational, which is always dangerous.
If a post-1970 forcing increase of 4 W/m2/century produced the post 1970’s warming rate of 1.8 K/ century, then the pre-1970’s forcing increase of 0.7 W/m2/century would be expected to produce a warming rate of about 0.3 K/century in the decades that preceded 1970 or about +0.1 K of warming over the 30-year pause from 1945-1975. So there isn’t a lot of warming missing from that Pause and there was “excess” warming in the two previous decades. (Real hand waving now.)
http://www.ipcc.ch/report/graphics/images/Assessment%20Reports/AR5%20-%20WG1/Chapter%2008/Fig8-18.jpg
AR5 WG1 Figure 8-18
Cherry picked up — in fact all these discussions relate to trend, then what about the cyclic part of 60-year cycle that the Sine Curve varies between -0.3 oC and +0.3 oC — for the period 1880 to 2010, the trend is 0.60 oC per Century or 01.34 oC for 1880 to 21000 or 0.91 oC for 1951 to 2100 with around half is global warming with adjusted data series. Your selected period needs correction accordingly.
Dr. S. Jeevananda Reddy
Dr Reddy asks: “What about the cyclic part of 60-year cycle that the Sine Curve varies between -0.3 oC and +0.3 oC — for the period 1880 to 2010?”
On a planet with chaotic ocean currents, two cycles of data is inadequate to characterize a phenomena as following a sine curve, which I understood to have a peak to trough amplitude of about 0.25 K, not 0.6 K. (Perhaps I’m wrong.) If you look at CET temperatures going back centuries, there are sizable oscillations, but they are very irregular.
Since this cycle could be responsible for the 1940’s warmth and 1950-1970 Pause, a 60 year cycle would put the next Pause at 2010-2030. Temperature since the 15/16 El Nino haven’t returned anywhere near the Plateau of 2001-2013 (being up at about 0.2 K) with suggestion of a weak El Nino this winter. It certainly isn’t obvious to me that we are halfway through another 1950-1970 period.
If current conditions were part of a 60-year cycle, we would expect: a trough (no warming), followed 15 years later by maximum warming, followed 15 years later by a plateau (no warming) and 15 years later by maximum cooling. Even with 30 year averages, one might expect to see part of this pattern since I calculated the trend for four 30-year intervals starting 5 years apart. One of them should have had a near zero 30-year trend, such as those seen beginning in 1945.
Dr. Reddy is certainly correct in suggesting that internal variability, changes that aren’t “forced”, can always be responsible for climate change and that we have no way except the Holocene proxy record for temperature to estimate how much change might be attributed to unforced variability. 45-years of steady (on the 30 year time scale) warming totaling 0.8 K is unusual for the Holocene.
“A harmless and beneficial rise in CO2 level is now the leading cause of outrageous politicization and corruption of science, media and public debate by the United Nations.” – UN IPCC.
My comment will be wasted as usually, but I do think it is disingenious to say the last El Niño period was weather and imply it should not be included in a climatic trend. I hate that oft-repeated idea.
All it says in fact is, that you need a long period, more than 15 years to get the existing warming trend visible. When one stops cherry-picking short periods, a clear 30 year trend is seen. At a decadal scale, obviously both positive and zero trends may show up.
Accepting this as a fact would increase one’s credibility while talking about confirmation bias that runs so rampant in climate science let alone its reporting in media.
” … My comment will be wasted as usually, but I do think it is disingenious to say the last El Niño period was weather and imply it should not be included in a climatic trend. I hate that oft-repeated idea. …’
—
Yes, it very much was wasted.
El-Nino is without question a weather-cycle transient, i.e. noise.
Noise is not signal.
If the climate fear industry was right about CO2 driving climate , and they aren’t , then why have no science based organizations announced and provided proof that humans can set the earths temperature ?
Yes C02 has some minor effect , thankfully mostly positive when increasing ,but the notion that natural occurring variables are being usurped is utter rubbish .
That is why science by direction as in the IPCC is such a complete joke . Science pretenders used to validate a scam . Sad but the public ain’t buying the inconvenient goof any longer .
Isn’t this just a cynical way to “take control” of the climate. They effectively invent it and of course we all know the direction they will invent in. They are setting up the IPCC to be a permanent blight on Mankind and the West in particular. It goes without saying this means jobs for life, in fact they can create dynastic employment. We will now be treated to a saw tooth temperature profile, half “real” if you trust their manipulated gathering techniques and half, well what ever sounds sexy to keep the believers in the required frenzy.
The climate is a system, and the system has NEVER changed.
David Whitehouse is right in that, within the past 30 full years of global temperature data, you can get 15 year periods such as 2000-2014 with very low warming trends (0.02 C/dec) or you can get 15 year periods like 2003-2017 with statistically significant warming trends (0.17 C/dec).
All this tells us is that within the WMO’s defined 30-year period of ‘climatology’ you can get shorter periods of slower or faster warming. But we all knew that already, right?
So how does the IPCC, or any group charged with extrapolating future global temperatures, decide which 15 year trend to go for, the slower one or the faster one? The answer of course is that they don’t use periods as short as 15 years to extrapolate future temperatures. They use periods of 30 years, as the WMO advises.
Perhaps a better question is, what is the trend in HadCRUT4 over the full 30 year period from 1988 to 2017? Is it closer the the 2000-2014 rate or closer to 2003-2017? The 30 year trend in HadCRUT4 from 1988 to 2017 is +0.19 C/decade. That’s even faster than 2003-2017! The linked chart below shows HadCRUT4 per month from 1988-2017 with the full 30-year trend and the two 15 year trends described by David Whitehouse. Whichj of the two 15 year trends do you think is more representative of the full 30 year trend?
http://www.woodfortrees.org/graph/hadcrut4gl/from:2003/to:2018/trend/plot/hadcrut4gl/from:1988/to:2018/plot/hadcrut4gl/from:1988/to:2018/trend/plot/hadcrut4gl/from:2000/to:2014/trend
The IPCC isn’t basing it’s 2017 + 15 year estimate on the El Nino influenced 2003-2017 trend, it’s basing it on the trend seen in the 30 years running up to 2017. 30 years of data, not any arbitrary 15.
Correction: the rate of warming in HadCRUT4 from 1988 to 2017 is 0.18C/dec, not 0.19 as stated. But that’s still warmer than the 15 year period ended 2017.
30 years is arbitrary. Why not 50 or 100? I would think that longer would be better if you are going to engage in such an exercise in the first place.
I agree up to a point that the longer the period used the better. 30 years is certainly better than 15, yet David Whitehouse and others seem to prefer the shorter period.
30 years typically covers around 6-8 full La Nina/El Nino cycles and typically includes the influences of 3 solar cycles; so it covers most of the known natural variables on climate.
If you use 50 years instead of 30 the warming results are little changed. 50 years from 2017 centres at 1992 and starts at 1968. The rate of warming in HadCRUT4 over that period was still +0.17 C/dec; just fractionally slower than the 30 year period ending 2017 and the same as the 2003-2017 trend. It’s within the IPCC ~ 0.2 C/dec estimate. Extrapolate the 50 year trend out and you still pass +2.0 C above pre-industrial by the mid 2070s.
Using trends for periods much over 50 years you start to drown out possible real recent influences; but then that may be some folks’ preference.
This linked chart compares the 50 and 30 year trends in HadCRUT4 to 2017:
http://www.woodfortrees.org/graph/hadcrut4gl/from:1950/plot/hadcrut4gl/from:1968/to:2018/trend/plot/hadcrut4gl/from:1988/to:2018/trend
Ice extent changes with temperature changes.
100% of all scientists know this.
97% believe temperature changes cause ice extent changes.
3% believe ice extent causes temperature changes.
This is what 100% of the ice core proxy data shows.
It snows more in warmer times when more oceans are thawed and it does get colder after that.
It snows less in colder times when more oceans are frozen and it does get warmer after that.
umm, correct me if I’m wrong but isn’t assuming an el nino causes warming a mistake? All an these types of events do is redistribute heat, in the nino case from where we don’t have sensors to where we do. Shallow waters don’t absorb more solar energy than deeper waters, they just concentrate it more – and that wind energy expressing as heat in water? came from heat – so, again, just moving it around.
There’s also the matter of the La Nina phase of ENSO, which removes surface heat temporarily. Lot’s of folks here highlight ENSO’s warming effect but ignore its counterbalancing cooling effect. Check out the NOAA ENSO index: over the longer term, it’s in near perfect balance with regards to its influence on tropical sea surface temperatures: http://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ONI_v5.php
It’s revealing how many people are eager to condemn the IPCC for what they’ve done here. It’s “devious manipulations by scheming minds working to an agenda”…”By hijacking language this way, climate alarmists insidiously manipulate words in their favor, by being deniers themselves of the original definitions of words that the general public understood those words to be. ” Etc. Any little excuse to condemn the IPCC will result in similar rants.
It certainly doesn’t help that the issue is framed as it is, starting with leading post title. In what way is this a “secret”? Besides, as others have pointed out, this is not a new definition of climate, but of global warming. The two are not synonymous.
Everyone should just take a deep breath and think about this. The trend in present global warming MUST be estimate. It doesn’t exist mathematically: you can’t calculate it as the average change over a given time period when the future is unknown. There is no slope to the trendline when you are at the endpoint. As an estimate of the current trend, one necessarily has to extrapolate. Otherwise, what are your options?
1) You make an exception and use the previous 30 years for the current trend. Then you’d have the same trend 15 years ago that you have now.
2) You always calculate the trend for any time based of the previous 30 years. But that, of course, would not be the current trend, it would be the past trend.
You can ONLY calculate the past trend.
As long as it’s made clear that is what the IPCC is doing, and that it is just an estimate with associated uncertainty, why is it a problem if they defines it like this? It’s not a magic trick, it’s a definition of something that isn’t otherwise clear if one takes the time to think about it. Is it even a big talking point in the report what the present rate global warming is?
From the post:
“It’s a well-known graph that shows no warming trend – except when you add the El Nino at the end, which of course is a weather event and not climate. ”
Since when is weather not a part of climate? It would only make sense to exclude this El Nino if you also excluded all other El Ninos and El Ninas – which is not done…
“The second graph shows the 15 years before the recent El Nino, i.e. 2000-2014. The trend over this period is influenced by the start point which is a deep La Nina year. Without it the trend is 0.03 °C per decade – statistically insignificant. Note that there are minor El Ninos and La Ninas during this period but they tend to have a small net effect.”
What kind of logic is this? It’s a smaller effect, so it’s not a “weather event”? Why is this sort of dubious science overlooked in favor of bashing the IPCC for giving their definition of a phrase?
I see Dr. Whitehouse’s rationale, but it’s still not reasonable to simply exclude inconvenient “weather” when looking at either climate or global warming.
Years ago when I was a UK ‘A level’ student (aged 17) in 1954, I recall doing something similar to this disputed analysis and not surprisingly, got ‘taken apart’ by the physics teacher of the day who carefully explained – just as others here have done – that there was a strong risk of using the manipulated data as if it was real. His further response was interesting in the light of this discussion. He said there is no problem showing such a trend if an arguments needs it but every graph showing this or based on this, must be annotated to explain what is real data and what is not. It would have been better if the IPCC followed this concept here.
Patrick Powers,
I agree with your professor, but I don’t understand why you seem to think that the IPCC has not annotated its graphs properly. Nor do I know what you mean by “manipulated” data.
There are reasons to criticize the IPCC report without resorting to illusory faults. It’s goals are simply not realistic. It’s possible that they justify setting high goals just to spur more effort to meet them, but it ends up making the whole thing less credible. But that’s another issue, not relevant to the post.
Kristi,
The logic of your argument applies to a gambler beside a Roulette wheel guessing about future states, but your logic is not appropriate for use in scientific advancement.
Here is an analogy. We start within the Fukushima nuclear plant hit by a strong tsunami. Plausibly, a lot of work had gone into safe operation, but there was damage compounded from the unsafe placement of emergency generators. There would have been a complex model of plant vulnerabilities, but they could not synthesise a full scale tsunami test. So they guessed at some tsunami values and ended up with a safety model and high subjective confidence in it. And some guesses.
Now, to climate models, which are clearly unable to have full scale live tests, so they have some guesses. The climate state over the next 15/30 years is a guess. But note, there is a non-zero probability that something perverse will happen to upset the climate models with their levels of confidence. And guesses.
This is all non-contentious black swan material until you address how to treat guesses. My assertion and plausibly that of many here is that one must not treat guesses as data and do typical things like taking standard deviations and forming distributions and blending them with measurements.
Where there are guesses, they correctly should carry big red flags. Maybe with a slogan like “We Guessed About the Fukushima Tsunami and We Lost.” Geoff
Kristi,
You asked, “Otherwise, what are your options?” Your options are to use standard practice in analyzing scientific data. If you wish to smooth it with a moving window, and center the results, then you truncate half the width of the window at the beginning and end of the time series. That gives you the smoothed behavior up to a 1/2-window width from the most current measurements. If you don’t want to lose the most recent data, then don’t smooth it! Either way, if you want to predict what the future holds, then select a method of extrapolation, defend the choice, and present the results as your best estimate of what the future holds. But, there is NO justification for defining the current climate as being based on your risky extrapolation. Climate is history up to today. Anything else is speculation!
This stats lesson from 2006 is rather pertinent today.
http://wmbriggs.com/post/735/
Do NOT smooth time series before computing forecast skill
Geoff
From the IPCC – a new word for a new form of data – ‘Unobtanium’ (courtesy of the SciFi Avatar movie) = ‘data we don’t yet have
Wonderful method. I’m going to do this with my taxes.