At least, Dr. Norman Page has none. After a 2012 prediction that was disastrously wrong, the Doctor is back with yet another prediction of cooling. It's always just around the corner, you know.
He has apparently learned nothing.
Rule #1 of PhD's: Be wary -- very wary -- of people who include "Dr." in front of their name. Page's doctorate is in geology, but using "Dr" is just a cheap attempt to grasp a smidgen of legitimacy before he has earned it. So you got a doctorate Norman. Big deal.
Page's 2012 prediction was that "the earth is entering a cooling phase which is likely to last about 30 years and possibly longer." That was about as wrong as could possibly be:
Did the doctor learn anything at all from his disastrous prediction? Clearly not, because now he's back making essentially the same prediction again:
What is the basis for the doctor's new prediction? First there is this interlude:
The modelling approach is inherently of no value for predicting future temperature with any calculable certainty because of the difficulty of specifying the initial conditions of a sufficiently fine grained spatio-temporal grid of a large number of variables with sufficient precision prior to multiple iterations.
That's actually correct! Climate models, which solve partial differential equations that describe the physics of climate, aren't initialized to an exact, existing set of initial conditions -- because no one knows the inital conditions, which would require detailed information on atmospheric and deep ocean currents, which we do not have (especially the latter).
This is exactly why modelers say their models make projections (not predictions) for the equilibrium state, after all forcing and feedbacks have played out -- which takes centuries, at least. Trying to interpret their results after 10 or 20 years is meaningless, because these models were never built to make such predictions.
So Page has that part right, though naturally it goes right over the heads of WUWT readers. But this hardly means that we expect cooling by 2020, or 2050 pr 2100, or pick whatever future point you want. More atmospheric CO2 warms the planet, whether you know the initial conditions or not.
Page wants to rely on supposed long-term climate cycles of 60 years and 1000 years. But he doesn't realize that anthropogenic forcings have changed everything.
60 years? OK, we have the PDO and AMO cycles, with something like this period for a full cycle. But these cycles don't create long-term warming -- just look at how much warming has happened since 60 years ago.
A thousand-year cycle? Based on what? Page doesn't say. Instead he is stuck in the past, not realizing that manmade forcings have changed everything.
But the climate forcings we're creating are far faster than any 1000-yr cycles, any Milankovitch cycles, or fuzzy millenium solar cycles.
This shouldn't be hard to understand, especially for a "Dr." But Page misses this completely. Perhaps he is angling for the vacant Bob Carter climate denial chair -- Carter was also a geologist.
If so, he might fit right in -- and be just as wrong as Carter was. And just as wrong as he was last time.
Norman Page didn't even fool us once. And already he's back trying to fool us again? Doctor, have a little pride, would you?
26 comments:
I agree with David's criticism of this article. It's weak in many, many respects.
I would dispute David's comment, "naturally it goes right over the heads of WUWT readers." In fact, the commenters at WUWT are not terribly supportive of the article. One commenter supports its publication, but calls it "speculation" and "interesting". That's pretty weak support.
Cheers
So let me get this straight. The climate models foresee a level of future warming which may take centuries, at least, so they should not be judged after 10 or 20 years. Currently, their record has been as follows :-
a) 1988 James Hansen based his predictions on three scenarios :-
Scenario A predicted an increase of 0.34-0.44 C/decade
Scenario B predicted an increase of 0.18-0.30 C/decade
Scenario C predicted an increase of 0.19-0.29 C/decade
Actual increase from 1988 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.13 C/decade
b) 1990 IPCC First Assessment Report Working Group 1 predicted 0.2-0.5 C/decade
Actual increase from 1990 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.12 C/decade
c) 1995 IPCC Second Assessment Report Working Group 1 predicted 0.1-0.35 C/decade
Actual increase from 1995 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.14 C/decade
d) 2001 IPCC Third Assessment Report Working Group 1 predicted 0.13-0.43 C/decade
Actual increase from 2001 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.11 C/decade
e) 2007 IPCC Assessment Report 4 Working Group 1 predicted 0.11-0.64 C/decade
Actual increase from 2007 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.12 C/decade
The increase from 2007 is only provisional since we have not completed a full decade yet.
In testimony before the US House Committe on Science and Technology on 2nd. Feb. 2016 Dr. John Christy of the University of Alabama at Huntsville showed that the mid-troposphere temperature trends since 1979 (the start of satellite observations) were as follows :-
Average of 102 IPCC CMIP5 climate model runs (used in IPCC AR5) +2.14 C per century.
Average of three satellite data sets UAH, RSS, NOAA +0.91 C per century.
Average of 4 balloon data sets NOAA, UKMet, RICH, RAOBCORE +0.79 C per century.
The mid-troposphere (surface to 50,000 feet) was chosen because it is the most responsive to the effects of greenhouse gases.
Are you saying that we are wrong to hold Hansen and the IPCC to account for these disrepancies because we need to wait 'centuries at least' ?
Addendum the First Fig referred to should be Fig 5.
Welcome, Dr. Page. It's great that you've joined the discussion of your paper here.
I am dubioius about the accuracy of predictions based on long, model-free cycles, for several reasons:
1. Very old data is unreliable.
2. One seldom has enough past repetitions of the cycle to have great confidence in it precise continuation.
3. If one doesn't know the cause, one doesn't know whether that cause is continuing to operate in exactly the same was as in the past. Even if a cycle generally continues, its timing might be different in the future.
4. There's a human tendency to see patterns where none exist. A famous example is that people tend to see patterns in a long-term random walk, where we know that any apparent pattern is just a coincidence.
David in Cal
Just check on the evidence for the millennial cycle in the previous comment.
.David in Cal
Just check on the evidence for the millennial cycle in the previous comment.
Dr. Page, it's hard for me to address your specific exhibits, because I'm afraid I don't fully understand them. Here are some observations. Please correct me if I'm wrong.
On Exhibit 2, I take it that the blue line is some measure of temperature. How was temperature measured? How accurate are the temperatures and how accurate are the dates?
On the blue line on Exhibit 2, I see 5 peaks, which are sort-of cyclical. But, the times between peaks aren't the same. The earliest pair of peaks are around 90,000 years apart, while the most recent pair are more like 120,000 years apart. A total of 5 peaks with differing times between them is suggestive, but hardly conclusive IMHO.
In Exhibit 2.3, the pairs of peaks also have different times between them. But, in this exhibit, the more recent peaks are closer together in time than the older peaks. Then, there's an earlier period where temperatures look more random than cyclic to me.
Cheers
Richard Mallett said...
a) 1988 James Hansen based his predictions on three scenarios :-
Scenario A predicted an increase of 0.34-0.44 C/decade
Scenario B predicted an increase of 0.18-0.30 C/decade
Scenario C predicted an increase of 0.19-0.29 C/decade
Actual increase from 1988 to 2014 (mean of HadCRUT4, NCDC and GISS) = 0.13 C/decade
-------------------------------------------------
Because I am sometimes curious as to whether people are actually correct in what they claim, I took the time to look at the data myself. Richard doesn't say what interval he used for the Hansen calculation, so I'm going to assume that he used 1988 to 2014 as was given for the mean of the measured temperatures. There is reason to go extend this back a bit to 1984 or so because the paper only got published in 1988, but let's stick with the basic premise, start at 1988 but include 2015 as well. My interval is the 2-sigma range:
Scenario A: 0.26 - 0.33 C/decade
Scenario B: 0.25 - 0.34
Scenario C: 0.12 - 0.19
My numbers don't seem to agree all that well with Richard's, especially Scenario C.
Now let's do the different data sets:
BEST: 0.12 - 0.23
NOAA: 0.11 - 0.22
GISTEMP: 0.11 - 0.23
HADCRUT4: 0.10 - 0.22
The global temperatures don't seem to agree with Richard's either.
In his 2006 paper, Hansen writes:
"Temperature change from climate models, including that reported in 1988 (12), usually refers to temperature of surface air over both land and ocean. Surface air temperature change in a warming climate is slightly larger than the SST change (4), especially in regions of sea ice. Therefore, the best temperature observation for
comparison with climate models probably falls between the meteorological station surface air analysis and the land–ocean temperature index."
Since the temperature range for the 4 data sets I just presented use SST, rather than air temperature over the ocean, let's do what Hansen says and see what land only data gives us. We expect this to be an overestimate:
BEST: 0.14 - 0.39
NOAA: 0.18 - 0.36
What do we get from this? We know that Hansen in his 1988 had a rather high climate sensitivity of 4.2 C per doubling of CO2. He also overestimated the influence of methane and CFCs. However, just on the basis of a raw comparison of his simulations with the data, at the very least there is good agreement between the global data and Scenario C. If we take into account that his model uses air temperature rather than SST, then the comparable temperature data are somewhere above Scenario C and somewhat below Scenarios A & B. Not bad at all.
The great Dr. Norman Page wrote:
"You are being ,shall we charitably say, disingenuous ,in presenting the GISS data as a proof that my 2012 forecast was wrong."
Your forecast was completely wrong no matter what dataset one chooses.
DOCTOR Norman Page wrote:
"Even the IPCC itself has now given up on estimating CS – the AR5 SPM says ( hidden away in a footnote)
“No best estimate for equilibrium climate sensitivity can now be given because of a lack of agreement on values across assessed lines of evidence and studies”"
A bald-faced lie.
The IPCC says there is no best value between 1.5-4.5 C. That certainly does *not* say that they've "given up" on estimating CS."
--
Norman, do you not realize how lying about these simple things makes you look like an untrustworthy ass?
David in Cal wrote:
"2. One seldom has enough past repetitions of the cycle to have great confidence in it precise continuation."
Is that supposed to be a joke?
If you lack sufficient data on the "continuation" of a cycle, then you can't conclude it's a cycle to begin with.
So stop using the word.
David in Cal: calling people you like (Norman Page) "doctor" and not using that for people you don't like (me, Joe T) just makes you look biased and desperate.
Titles don't matter, except to people who don't have them.
David A. -- I didn't mean any disrespect to you or Joe T. I called Norman Page "Dr." because he prefers to use that title.
Regarding the word "cycle", what you wrote is what I meant. Sorry if my wording was clumsy. I used the word "cycle" because Page used that word. My comment, like yours, was meant to dispute whether what he sees as a cycle really is a cycle.
BTW, David, how did you get the idea that I like Norman Page? I don't admire his work. On the contrary, I said his paper was weak in many, many respects and then I listed a bunch of problems as regards just one aspect of his paper.
Cheers
Ok, David. Thanks.
Does Norman ever provide a function for the waves that inform his predictions? That would be very interesting to see.
Reply to Joe T :-
For Hansen's 1988 projerctions I used the figures from http://www.realclimate.org/index.php/archives/2007/05/hansens-1988-projections/ which says :-
"For the model simulations, the trends are for
Scenario A: 0.39+/-0.05 deg C/decade,
Scenario B: 0.24+/- 0.06 deg C/decade and
Scenario C: 0.24 +/- 0.05 deg C/decade."
For the years 1988-2014 I used the 2014 data from GISS, HadCRUT4 and NCDC as it was published at the start of 2015 and the linear regression in Excel which gives (for the years 1988-2014) :-
NASA GISS 0.151 deg C/decade
HadCRUT4 0.150 deg C/decade
NOAA NCDC 0.140 deg C/decade
which are all below Hansen's figures that I copied from the article on Real Climate.
Hi Richard,
Thanks for clearing that up. So what you did was you took the numbers that Gavin Schmidt used based on Hansen's data from *1984* - 2006 (note that Gavin did not start in 1988) and you compared that to the observational data in the years 1988 - 2014. That's not a fair comparison.
In the article Gavin compared the GISS data for the same 1984 - 2006 interval and said "From 1984 to 2006, the trends in the two observational datasets are 0.24+/- 0.07 and 0.21 +/- 0.06 deg C/decade, where the error bars (2 sigma ) are the derived from the linear fit." The two data sets are the met station and the land/ocean. That actually agreed fairly well with Scenario B and C. Why didn't you report that result?
If you wanted to extend Gavin's analysis, why didn't you load up the Hansen data for 1988 to 2014 at least and see how that changed from when Schmidt did the calculation back in 2007? That's what I did.
Also you claimed above that the average of HADCRUT4, NCDC and GISS was 0.13 C/decade. How is the average of 0.151, 0.150 and 0.14 equal to 0.13? Why didn't you give the 2 sigma spread?
Do you think you made a fair comparison of Hansen's simulations to the actual real temperature data sets?
"2. One seldom has enough past repetitions of the cycle to have great confidence in it precise continuation."
You mean Eli should not expect night and day, spring summer fall and winter? Drat
Page, of course, does not understand that when cycles crash into physical limits, the cyclist gets badly hurt. Of course others have tried this and not prospered. As Carl Sagan pointed out we also laugh at Bozo the clown
http://davidappell.blogspot.com/2012/04/roy-spencers-entertaining-polynomial.html
Reply to Joe T :-
I assumed that Hansen's forecasts when he made them in 1988 were as reported in Real Climate, so I was comparing 3 forecasts from 1988 (to some unspecified date in the future, implying that that rate would continue as a linear trend, ignoring natural variation, at least until 2014) with actual temperature data from 1988 to 2014.
What Hansen did between 1988 and 2014 is answering a different question than
'How accurate was Hansen's forecast in 1988 ?
How quickly has the temperature increased since 1988 ?'
If Norman Page (I neither know nor care whether or not he's a real doctor) had made a forecast back in 1988, we could have looked at the temperature change from 1988 to 2014, and see how accurate was his forecast.
Yes, I originally (back in January 2015) calculated the average trend to be 0.13 C per decade; but, when I recalculated it today, it came out to 0.147 C per decade. That is still considerably less than
Scenario A: 0.39+/-0.05 deg C/decade,
Scenario B: 0.24+/- 0.06 deg C/decade and
Scenario C: 0.24 +/- 0.05 deg C/decade
which were the three forecasts made in 1988.
Richard,
In short, your assumption was wrong. When you quote Schmidt's temperature trend and say those were the three forecasts made in 1988 it simply isn't true. The temperature trend for each Scenario that you got from Gavin Schmidt was specifically for the interval 1984-2006. As I showed that was not the same as the trend for 1988-2014 or 1988-2015. As Gavin showed, the temperature trend from 1984 to 2006 agreed very well with Scenario B and C. Why didn't you report that?
You could have easily looked at the Hansen data yourself since it was linked to in the Schmidt article that you referenced in the first place. You could have calculated what Hansen's trend was for any given interval. You chose not to.
What you're missing here especially is that these are not forecasts, the term you use multiple times. They are scenarios. They are not predictions. They are meant to say that if such and such is true, then this will happen. Each scenario corresponds to a different forcing profile. That's why the trend for a given scenario that Hansen gave depends critically on the time interval that you are considering.
If anything, the actual forcing scenario that comes closest to reality is Scenario C. As I showed the actual temperature trend for the various data sets came very close to what Hansen produced in his simulations.
Reply to Joe T :-
Many thanks for pointing me to the Hansen et al 1988 paper 'Global Changes as forecast [there's that word again] by Goddard Institute for Space Studies Three Dimensional model'
Figure 3 was what I remembered from last year, This shows the three scenarios A, B and C - I didn't use that because I could not find forecast anomalies for the individual years from 1988 onwards, ehich is what I was interested in. Please let me know how to find values of the 1988 forecast anomalies for each year, from which I could calculate the forecast warming rates over any period. Many thanks in advance.
It does say in the text that :-
"even with the drastic, and probably unrealistic, reductions of greenhouse forcings in Scenario C, a warming of 0.5°C is attained within the next 15 years."
and :-
"The 1°C level of warming is exceeded in the next few decades in both scenarios A and B; in scenario A that level of warming is reached in less than 20 years and in scenario B it is reached within the next 25 years."
[Just a reminder here of the three scenarios :-
"Scenario A assumes that growth rates of trace gas emissions typical of the 1970s and 1980s will continue indefinitely"
"Scenario B has decreasing trace gas growth rates, such that the annual increase of the greenhouse climate forcing remains approximately constant at the present level"
"Scenario C drastically reduces trace gas growth between 1990 and 2000 such that the greenhouse climate forcing ceases to increase after 2000."
Which scenario do you belive that we are following now ?]
From the context it appears that Hansen is talking about increases from 1987 and not to some pre-industrial level.
So let's go with that for now, in the absence of numerical values for the three scenarios for each year.
Forecast for 1987-2007 scenario A = 1.0°C
Forecast for 1987-2012 scenario B - 1.0°C
Forecast for 1987-2002 scenario C = 0.5°C
Actual anomalies for 1987 :-
NASA GISS 0.28°C
HadCRUT4 0.19°C
NOAA NCDC 0.33°C
Actual values for 2002 :-
NASA GISS 0.60°C
HadCRUT4 0.50°C
NOAA NCDC 0.61°C
Actual values for 2007 :-
NASA GISS 0.62°C
HadCRUT4 0.49°C
NOAA NCDC 0.59°C
Actual values for 2012 :-
NASA GISS 0.57°C
HadCRUT4 0.47°C
NOAA NCDC 0.57°C
1987-2007 increase (scenario A forecast 1.0°C) :-
NASA GISS 0.34°C
HadCRUT4 0.30°C
NOAA NCDC 0.26°C
1987-2012 increase (scenario B forecast 1.0°C) :-
NASA GISS 0.29°C
HadCRUT4 0.28°C
NOAA NCDC 0.24°C
1987-2002 increase (scenario C forecast 0.5°C) :-
NASA GISS 0.32°C
HadCRUT4 0.31°C
NOAA NCDC 0.28°C
Note that Hansen does point to the "drastic, and probably unrealistic, reductions in greenhouse forcings in scenario C"
Richard,
To answer your question as to where the data is for Hansen's 1988 paper you can find the temperature anomalies here and the and the forcings here . I like using forcings in W/m2 rather than the temperatures that Hansen shows in his figure 2.
I prefer to work this problem backwards rather than pick and choose what are the anomaly differences between certain years. The latter approach not only ignores the 2-sigma uncertainty in the trend but it also ignores natural variability. We've already seen Gavin Schmidt's Realclimate post which showed excellent agreement between the projections (sorry, they are not forecasts even if Hansen uses that term) and the real data.
Let's start here: If you look at the forcing differences between 1987 and 2010 for scenarios A, B and C you get 1.7, 1.1 and 0.4 W/m2 respectively (I chose 1987 to get away from the El Chichon volcano eruption in 1982). Now if you go the Forcings in GISS climate models page you can find two recent set of forcings. The CMIP5 forcings here and Hansen's 2011 forcings here. The Hansen data set only goes up to 2010, which is the reason why I picked that as the end date for the comparison.
If you compare 1987 to 2010 you find there was an increase in the forcing of 1.16 for CMIP5 and 0.62 W/m2 for Hansen. They differ roughly by a factor of 2 because Hansen has a larger cooling tropospheric aerosol contribution to get the net forcing closer to the 0.5 - 1.0 W/m2 that is estimated from the ocean heat content.
I asked Gavin Schmidt about the difference between the two data sets and he replied, "Different kinds of calculation. The Hansen et al 2011 values are from an inverse calculation, while the values used in CMIP5 simulations are from forward models driven by more up-to-date emissions data. The uncertainties in both are relatively large. – gavin" You can read it over at realclimate.org.
Given the factor of 2 difference in the forcings and what Gavin calls the large uncertainty in both, it's probably safe to say that the actual forcing is somewhere between Scenarios B & C. So what does the data say?
Well, as it turn out Nick Stokes has a great little graphic with radio buttons so you can switch in various data sets and compare them directly to the Hansen projections. You can find that here. Again emphasizing the fact that the Hansen projection was for air temperature and not land+ocean SST, we expect the real answer to be somewhere between the global data and the met stations. And that's exactly what we find from Nick Stokes.
OK, we still seem to be talking about (at least) two different things, so I will leve it there.
Post a Comment