Articles on this Page
- 01/05/19--09:55: _ Comment on Rec...
- 01/05/19--09:59: _ Comment on Rec...
- 01/05/19--10:15: _ Comment on Rec...
- 01/05/19--10:23: _ Comment on Rec...
- 01/05/19--10:53: _ Comment on Sea...
- 01/05/19--10:57: _ Comment on Rec...
- 01/05/19--10:58: _ Comment on Rec...
- 01/05/19--11:15: _ Comment on Rec...
- 01/05/19--11:16: _ Comment on Rec...
- 01/05/19--11:40: _ Comment on Rec...
- 01/05/19--11:42: _ Comment on Rec...
- 01/05/19--11:59: _ Comment on Sea...
- 01/05/19--13:56: _ Comment on Nat...
- 01/05/19--14:32: _ Comment on Rec...
- 01/05/19--14:54: _ Comment on Sea...
- 01/05/19--15:00: _ Comment on Sea...
- 01/05/19--15:05: _ Comment on Sea...
- 01/05/19--15:06: _ Comment on Sea...
- 01/05/19--15:36: _ Comment on Rec...
- 01/05/19--16:12: _ Comment on Sea...
Paul, your 1st attempt was: "There seems to be an assumption here that the “multi-decadal signal” represents internal variability. I would suggest you try applying this procedure to the CMIP5 mean. You’ll find the same pattern."
The (multi) decadal signal was the difference between the NA and the SO as it was stated in the main post, NOT ENSO influence, solar- and volcano forcing (tamino filter) , which is not (multi) decadal. Therefore you should admit, that the "multi decadal signal" is not a product of the forcing as you estimated but indeed internal variability as I showed it in the figure above. Thanks for beeing patient.
Frank, Have you done the same thing using 1980-present?
dpy: I tried it but I'm afraid that the data quality is too limited ( especially for -50...-65N; 70...-70E) to applicate the procedure before 1950. The HadSST datasets have big gaps and I'm not quite sure how ERSSv5 manages the set back to 1850. And this (multi) decadal signal is the core of the post, therefore I would not recommend to overlook the big, big uncertainty before 1950.
I meant 1980 to present NOT 1880.
One of the other things that gives me cause to doubt this article is that it concentrates largely on atmospheric CO2 and the biomass. Patrick Moore provides context, and to my mind opens up different ways of thinking:
"Let’s look at where all the carbon is in the world, and how it is moving around.
Today, at just over 400 ppm, there are 850 billion tons of carbon as CO2 in the atmosphere. By comparison, when modern life-forms evolved over 500 million years ago there was nearly 15,000 billion tons of carbon in the atmosphere, 17 times today’s level. Plants and soils combined contain more than 2,000 billion tons of carbon, more that twice as much as the entire global atmosphere. The oceans contain 38,000 billion tons of carbon, as dissolved CO2, 45 times as much as in the atmosphere. Fossil fuels, which are made from plants that pulled CO2 from the atmosphere account for 5,000 – 10,000 billion tons of carbon, 6 – 12 times as much carbon as is in the atmosphere.
But the truly stunning number is the amount of carbon that has been sequestered from the atmosphere and turned into carbonaceous rocks. 100,000,000 billion tons, that’s one quadrillion tons of carbon, have been turned into stone by marine species that learned to make armour-plating for themselves by combining calcium and carbon into calcium carbonate. Limestone, chalk, and marble are all of life origin and amount to 99.9% of all the carbon ever present in the global atmosphere. The white cliffs of Dover are made of the calcium carbonate skeletons of coccolithophores, tiny marine phytoplankton.
The vast majority of the carbon dioxide that originated in the atmosphere has been sequestered and stored quite permanently in carbonaceous rocks where it cannot be used as food by plants."
dpy: sorry for missreading! I did it in between and when one estimates the TCR from the regression of the GMST (C&W) vs. the anthropogenic forcings for 1980....2016 one gets one gets TCR=1.77. If one applies the method described in the mean post one gets the best coincidence between the raw data and the reconstructed time series with the mentioned equation (1). The TCR is not 1.77 but 1.3 with the consideration of the internal variabiliy. Thanks for pointing to this.
The first mention of multi-decadal signal (in fact, the only use of the word 'signal') in your post was this:
'The residuals of this regression contain a (multi) decadal signal'.
The cause of multi-decadal variability in the sub-polar North Atlantic (and I believe it is basically the North Atlantic doing the work in that NA-SO equation) is probably a complex topic. Obviously there will be some internal variability in this region, but the particulars of the multi-decadal pattern people tend to talk about show strong correlation with timing of volcanically-active and volcanically-quiet periods. And there's been a lot of work in the past few years on the link between volcanic events (plus anthropogenic forcing) and North Atlantic variability.
As you point out, climate models are at best equivocal about a strong volcanically-forced signal in the sub-polar North Atlantic, but some do show a similar pattern. It's quite plausible there is some regional dynamic response to volanic forcing that the models are missing.
I would think events of the past decade should give pause to those who believe sub-polar North Atlantic internal variability is really driving much at the global scale (at least in the direction usually implied) given that surface temperatures there have plummeted at the same time that global average temperatures have risen rapidly.
dpy: btw, this is what fig. 4 of the mean post shows :) however I tested it in a different way which is good.
Thanks Frank, that confirms a comment at SOD that you may have been the author of. It makes sense that longer time frames are more immune to internal variability.
Paul, your statement:" I would think events of the past decade should give pause to those who believe sub-polar North Atlantic internal variability is really driving much at the global scale (at least in the direction usually implied) given that surface temperatures there have plummeted at the same time that global average temperatures have risen rapidly." shows some believing in short term noise ( "past decade"). I showed that this (multi) decadal signal is at work on longer timescales ( after 1950 at least) to mute/amplify the anthropogenic signal and also the last decade ( after removing ENSO) follows this line. However, thanks for this inspiring dialogue which seems to be wishfull for bridging the "blog frontiers".
How that evolves over time
An incredibly important question and I believe, a largely ignored question. Thinking about it more, maybe one of the most interesting questions since there might be some unknown unknowns lurking out there.
Sorry, I'm new to this, where did you explain the above?
I avoided the radiated heat budget - far too complicated.
My job is to think about stuff that normally doesn't get thought about, so thanks for the kind comment - it is what I aim at.
A very large input of CO2 and methane and - in particular - water vapour. All feedback to the orbital warming.
The ecology of the flooded areas in the Holocene looks like being a hot topic for the coming decades
Dr. Curry - well done. Vitally important piece.
With regard to your two requests for comments:
1) "So scientific overconfidence seems to be a victimless crime, with the only ‘victim’ being science itself and then the public who has to live with inappropriate decisions based on this overconfident information".
The public having to live with inappropriate decisions based on this overconfident information is the most serious of crimes in and of itself. Here I think of "scientists" like Trofim Lysenko. So, too is the critical matter of reliance on sound scientific method to highly-advanced industrial societies. Both are victims, and at a level of importance it would be impossible to overstate.
2) There is no shortage of analogues of unjustified, overconfident scientific conclusions:
Ehlich/Holdren & resource depletion
Club of Rome & population/resource "sustainability"
Every IPCC report's Summary for Policymakers (even if AR5 less so)
USGS and others - "peak oil"
The Catholic Church vs. Copernicus & Galileo
Your intellectual honesty, courage, and scientific integrity shall be noted by future generations of science historians in relation to climate science. Billions of current and future human residents of planet earth owe you a debt of gratitude you shall never receive from them, directly or indirectly, for your scientific integrity. As one of them, I hereby thank you.
Paulski wrote: "I would suggest you try applying this procedure to the CMIP5 mean. You’ll find the same pattern."
What make you think any AOGCM has the ability to realistically reproduce internal variability? They all struggle or fail with known phenomena. Unless something has changed recently, we still can't predict the next El Nino more than 9 months in the future with even the most sophisticated models designed to deal exclusively with this problem!
The multi-model mean is even worse. If a subset of models produce some form of AMO with a period of a 65 years, unless you properly initialized them in phase, they would simply smear the signal out. Now complicate matters by assuming they deliver periods ranging from 50-80 years and start even slightly out of phase.
Alan wrote: "Water vapor is the most potent greenhouse gas and changes in sea level form a feedback process: when warming is taking place the ice caps melt and new areas of warm tropical seas are added that will increase water vapor in the air through evaporation ..."
I believe this part of your analysis is incorrect. What convection can only remove heat from the surface as fast as radiation transfers it to space from the upper troposphere! This is the rate-limited step in cooling the planet. You can't have net evaporation if the upper atmosphere is too warm for an unstable lapse rate to develop somewhere.
The rate of evaporation is determined by two factors: wind speed and undersaturation ((1-RH)*saturation vapor pressure). If the upper troposphere is too hot to permit convection, relative humidity rises or wind speed falls. This is how climate models reduce the rate of increase in precipitation from the expected 7%/K to around 2%/K.
The situation is somewhat different over land (and you are discussing more land). But the principle is the same: heat must leave the upper troposphere before convection can develop.
Alan wrote: "During the stable glacial maximums, the present day coastlines were 120 m above sea level. Thus the drop in air pressure over present land expressed in equivalent present day height above sea level would be."
This may be incorrect. Surface air pressure is due to the weight of the atmosphere above the surface. With the exception of a slightly stronger pull of gravity from being 120 m closer to the center of the Earth, the weight of the atmosphere is proportional to its mass. Unless you add mass to the atmosphere, surface air pressure can't rise.
Thank you for your reply. Regarding the cause of the polar warmth during the Eocene Thermal Maximum (ETM), Wikipedia mentions several possible causes, and says that CO2 concentrations cannot explain it.
<b>Early Eocene and the equable climate problem</b> https://en.wikipedia.org/wiki/Eocene
The Early Triassic (250 Ma) was the hottest period. Most of the time from 542--365 Ma (Cambrian to Late Devonian) and 100--80 Ma (Mid Cretaceous) was hotter than the ETM. See Scotese (2018) chart on p3 here: https://www.researchgate.net/publication/324017003_Phanerozoic_Temperatures_Tropical_Mean_Annual_Temperature_TMAT_Polar_Mean_Annual_Temperature_PMAT_and_Global_Mean_Annual_Temperature_GMAT_for_the_last_540_million_years
These charts are updated from Scotese 2016 which explains the methodology, but the update uses what is considered better data for the tropical temperatures.
I'll reply to your comment about the carbon budget later.
Scotese (2016): https://www.researchgate.net/publication/275277369_Some_Thoughts_on_Global_Climate_Change_The_Transition_for_Icehouse_to_Hothouse_Conditions
The assumption that C&W version of a GMST is correct is a big one. It ignores the great likelihood of errors in the adjustment of temperature data. See chapter 9 of the audit of the HadCRUT4 dataset for details of the likely errors.
It's difficult to find a coastline map from about 10,000 BC. Then I recalled this one:
One of my points is ocean circulations.