Arguments
The evidence for climate change WITHOUT computer models or the IPCC
Posted on 11 May 2013 by dana1981
This is Peter Hadfield's (potholer's) latest climate video, summarizing the evidence for human-caused climate change, excluding climate models and the IPCC.
More great work from Potholer.
No 'skeptic' appears to be doing well in taking up his challenge in the comments below the video. Perhaps one of the locals will do better? Unlikely...
On observation of this video, at 3:45 - 3:53 we are told that "a doubling of Co2 from pre-industrial levels will lead to a rise in average global temperatures of about 1 degree Centrigrade". And that at 5:30 - 5:39 - "for a doubling of CO2 concentrations we should see a rise in global temperatures of between 2 and 4.5 degrees" This might cause some confusion.
Dear Dana,
Great summary! One thing got me frowning, though. Do you know why there were coral reefs at the poles (at 8:05) although CO2 was 25 times higher than it is today? Did the ocean have a higher pH or were they a different type of coral or what was different when compared to todays coral reefs dissolving in highly diluted carbonic acid?
Regards
The oceans only become corrosive when CO2 is added to the atmosphere in a geologically-rapid manner such as today. This is because the chemical weathering of rock is able to supply alkalinity back to the oceans over tens to hundreds of thousands of years.
There is an upcoming rebuttal on this - it's one oversight that needs to be fixed.
Also note that the equatorial oceans would have been extremely warm during the "Greenhouse Periods" on ancient Earth. Too hot for reef-building coral similar to modern-day to have survived.
Oriolus Traillii @3, I think Potholer may have made a mistake in this case. He cites a solar luminosity of about 0.94 of the current value at the time of the purported near polar coral reefs. That places the event at around 700 million years before the present, or over 100 million years before the evolution of the first corals in the Cambrian. In the precambrian, the most probable reefs to be formed would have been stromatolite reefs, although there were other forms of precambrian reefs.
You should really contact Potholer for the source of that claim, and the basis for claiming the reef as evidence of warm water. I am sure he will be able to provide you with a suitable scientific referrence that will get the facts straight (and possibly prove me wrong).
If that's the case Tom, it would place the timing hundreds of millions of years before the evolution of Scleractinian (reef-building) coral.
Potholer has released a sources video for this one - it's currently his most recent. I'm sure he'd be happy to have any errors pointed out.
We all know CO2 is a green house gas, we all know atmospheric CO2 increases alone won't force significant warming by it self, we all know forcing by H2O in the atmosphere will multiply the CO2 effect. What isn't certain yet is the value of the forcing by H2O. because the actual measured results - (-snip-)? Maybe we need to go back to the basics with a fresh look at this, (-snip-).
[DB] Sloganeering and inflammatory tone snipped.
Warren #9
No one should really take seriously statements likes no statistical warming for nearly 20 years despite increase of CO2 (or 17 years, or 15 years). That has been oft-refuted here and on other blogs.
www.skepticalscience.com/short-term-variability-vs-long-term-trends-mclean.html
www.skepticalscience.com/watanabe-et-al-2013-another-piece-of-the-puzzle.html
Warren - Global warming has accelerated. Only by excluding all the evidence for warming and focusing on the (roughly) 2% of global warming that goes into heating the atmosphere, and only over a short period of time, does it appear otherwise.
I'm sure that you'll agree that looking only a one small piece of the evidence is not a valid scientific approach.
Warren "no statistical warming" does not mean "no warming", it just means that the observations do not effectively rule out the possibility that there has been no warming over the period in question.
If you want a statistical test to show there has been no warming, you need to show that the observations are inconsistent with the extistence of a warming trend of the magnitude suggestsed by the IPCC. If you do this you will find that the evidence for no warming isn't statistically significant either. Curously the "skeptics" never actually do this ;o)
Warren Hindmarsh @9, the global climate warmed at a rate of 0.071 C per decade durring the twentieth century (1901-2000). In contrast, it warmed at 0.108 C per decade from Oct 1994 to the present, the longest period of statistically insignificant warming (for Gistemp). That is, in this supposed warming pause, the Earth has been warming more than 50% faster than the record breaking Twentieth century rate of warming. Only in the wierd world of climate change denial can the rate of warming be significantly above average but it be called a warming pause solely because the chosen time interval is too short to contain enough measurements for statistical significance.
Now, can you explain to me without the deceptions why a warming rate greater than the twentieth century average, and very close to the predicted rate once known confounding factors (ENSO etc) are accounted for requires us to fundamentally reconsider the theory?
@mcooke #2
at 3:45 - 3:53 we are told that "a doubling of Co2 from pre-industrial levels will lead to a rise in average global temperatures of about 1 degree Centrigrade". And that at 5:30 - 5:39 - "for a doubling of CO2 concentrations we should see a rise in global temperatures of between 2 and 4.5 degrees" This might cause some confusion
...except the video content between the two points is actually an explanation of how one gets from one figure to the other - namely positive feedback due to water vapour.
Warren @9: This might interest you:
http://diyclimate.x10.mx/responsemodel/nbox.html
It's a toy 'empirical' climate model, which works by determining the climate sensitivity which best explains the observed temperature record from the forcings. It's got a lot of limitations, but it can provide a good cross-check on your intuition.
Go to the page, scroll down and set 'Years to fit' to 1880 - 1997. Now click Calculate. Underneath you'll see a figure for TCR (a measure of climate sensitivity over a human lifetime). You should get a number like 1.58C.
Now change 1997 to 2011, to include the most recent data available to the model and hit calculate. Note that including the last decade and a bit increases the estimate of climate sensitivity.
Your intuition may rebel against this, but think about it for a bit. If you can't work out why, come back tomorrow and I'll explain it.
[Note: I'm not claiming validity for the model, only that intuition is unreliable concerning the interpretation of recent temperature trends.]
Warren: Did you try the model I gave you?
Kevin come back tomorrow and I will explain it
... on a post featuring a video consisting almost entirely of Peter Hadfield's dry, measured delivery.
Oriolus Trailli and Rob Painting, following up from a hint by 64jcl at Potholer's channel, I have found a reference to fossil coral on Devon Island, Nunavat (in the Canadian Arctic). The fossil formation spans an interval from 600 million years ago (ie, in the Pre-Cambrian and approximately the time of the last snow ball Earth episode) to 39 million years ago. The fossil corals, however, appear in the late Ordovician, or approximately 450 million years ago. If this is indeed the reference Potholer had in mind, he has confused the timing of the corals with the begining of the formation.
To place the timing of the coral fossils in context, here is the broad temperature history of the Earth:
The fossil corals appeared in the late Ordivician just prior to the partial glaciation indicated by the thin blue line. The chart has a resolution of 10 million years, so fluctuation of CO2 concentration by up to 2000 ppmv (with corresponding fluctuations in temperature) may have occured at finer resoltuions and not show up on the chart, so global temperatures at the time of the corals may have been several degrees above those shown in the chart. However, it also cannot be excluded that the fossils are of cold water corals.
Kevin C @15
I tried your model but with 1997-2011 there is a pop-up error message ("Insufficient data to fit").
Warren didn't come back but I wondered if you wouldn't mind explaining the counter-intuitive result anyway. I have always had a blind spot with statistics!
Bob @20, when I use the model, it only allows for data up to 2010. I also notice that when truncating the data from 2010 down to 1997, the Transient Climate Response (TCR) falls from 1.675 to 1.58, while the Coefficient of Determination (r2) falls from 0.925 to 0.86. The later surprises me. I also note that when the duration is reduced, the graphed observations are reduced, but the graphed model is not. I am not sure whether that is a bug or a feature.
The increase in TCR for the full 1880-2010 period is likely due to an acceleration of the underlying warmng rate due to anthropogenic factors. That is, the underlying warming is increasing faster than would be expected just from the increase in GHGs and aerosols, etc. This has been masked by a reduction in energy recieved from the Sun, along with transient effects such as ENSO and volcanoes.
The formula for calculating the change in the Sun's Luinosity over time is
L(t) = [1+ 2/5(1-T/To)]-1 Lo
where the T is time from the birth of the sun. This will give you the luminosity change between any two points in time. The subscript 'o' is always the later of the two times. This is cited in Gough 1981, page 28
So the value you get depends on what age you use for the Earth. The standard value is 4.7 billion years. So 94% gives us 750 million years ago, 95%, 618 million years ago.
This is an increase of 40% since the birth of the Sun up to the present.
Peter does say however that it was 'around 6%'.
Glenn Tamblyn @22, I used 4.57 billion years as the current age of the Sun, following Feulner 2012. That dates "about 6% weaker", ie, 94% of current luminosity, to around 0.73 billion years ago (Gya) compared to your 0.75 Gya. If Potholer was indeed referring to the Devon Island corals, the correct figure would have been about 3.8% weaker than today, but of course, those corals are not associated with the termination of a snoball Earth event.
Bob: You really need about 70 years of data for the model to have any validity (i.e. change the second number but leave the first at 1880). Ideally those 70 years need a volcano and a significant change in a slower forcing. And then there are big caveats due to the uncertainty in the forcings - try double black carbon as per recent research and see what happens to TCR. So don't overinterpret this model.
However I still maintain (along with Isaac Held) that it can be a useful check on your intuition. The interesting case is comparing 1880-1997 with 1880-2011. Including the extra years increases TCR in this model. Why?
The years 1997-2011 have two features:
1. They are all much hotter than the average over the whole period.
2. There is comparatively little trend within that period.
Somewhat simplified: TCR is related to changes over a period of 70 years or so. So the trend within a 15 year period has rather little 'leverage' to affect TCR. On the other hand, the huge difference between the last 15 years and the period 70 years before, so adding more hot years at the end has a lot of 'leverge' to affect TCR.
In other words, adding more extreme hot years in a clump at the end tells us a lot more then the small variations within that clump.
Tom: When you reduce the fitting period, the model is only fitted to observations in the fitting period (the thick part of the obs line). The R2 is based only on these observations. However having determined the parameters of the model I go back and recalculate the fitted temps using the forcings for the whole period, thus the model is extended to cover the whole period, predicting values which were not used in fitting the model. That means you can do a visual 'hold-out' test, to determine the skill of the model in predicting temperatures it has never seen.
Strictly you should also re-optimise the time coefficients of the exponentials when you do change the fit period. One day I may get round to automating this in the Javascript version.
Kevin C @20
Thanks for the clarification. I was being dense and had misread what you had suggested @15 for the modified date range.