Green World Trust

ian · Posted: 15 Oct 2008 01:53 am Post subject:

Hey Lucy...like yourself I have had to take on other approaches also. My brain doesn't operate in a way that makes the sciences/mathematics come naturally. As such I have had to rely on my intuition, my understanding of human nature, societal paradigms, politics and history.

I started to become skeptical after reading a little notice on our towns communal notice board directing the reader to www.iceagenow.com. My approach could be summed up as 'what are these pesky skeptics on about' and in that light I checked it out. To cut a long story short it eventually led me here (through a post of yours @ the CO2 Skeptic website). Through you I had some conversations with Peter Taylor and then came across David Noble and Denis Rancourt, to name a few, who are not of the neo-liberal crowd. It feels quite odd to be a member of the Greens who questions AGW.

It continues to be quite the journey and I suspect I've been helped along by that 'force' (Great Spirit) that quantum mechanics can't yet fathom!

best wishes, ian

Lucy Skywalker · Posted: 16 Oct 2008 05:38 pm Post subject:

Ian - glad you've been in touch with Peter Taylor, funny how things come full circle since he was one of my own awakeners, and is another True Green...

Lance Endersbee has basically vallidated the stats of his graph by doing another graph using the "correct" compatibility - 21-year moving averages for both axes. In the end it makes no difference - the correlation is still superbly high. This still stands therefore as powerful evidence that sea surface temperature is the real determinant of CO2 levels, and that our emissions simply get absorbed by, and lost in, the huge biological and physical annual flux. I've been altering my pages accordingly now...

I've asked Lance Endersbee if he can do another graph along the same lines, this time subtracting human emissions from each Mauna Loa reading - ah, but annual or cumulative? and do we allow for part of it to be absorbed? The CO2 levels should parallel the cumulative effect plus kinks and changes in slope, IF there is any human influence. Which it looks already like they don't. But I'd like to see the lowered correlation I expect.

After all, the biosphere has already proved itself capable of absorbing the continually-emerging fresh CO2 from the earth's interior, over millions of years... ah yes, and THAT is why CO2 started so high and has been steadily falling ever since, of course!

Lucy Skywalker · Posted: 17 Oct 2008 10:24 pm Post subject:

Ferdinand noted in a message to me

Questioning_Climate · Joined: 10 Oct 2008 Posts: 28 Location: Leics. UK

FerdiEgb,

Do you think it might be appropriate to go back to fundamental science and work through to a discussion of the possible mechanisms that might explain the observed changes in species concentrations within the firn, at and after close-off? It is my impression that an approach of this kind would benefit the rest of the forum and may be draw some conclusions.

Assuming you don't mind participating, where should we start? Here are a couple of suggestions:
1) Basic ice science.
2) Criteria for trapping and maintaining accurate atmospheric gas samples.

Do you have any thoughts?
_________________
"Correlation is NOT Causation"

Lucy Skywalker · Posted: 19 Oct 2008 09:39 pm Post subject:

I want to lock this thread now, for several linked reasons, none of which are negative.

Stimulated by Ferdinand et al, I have researched the ocean and CO2 factors a lot and I now feel clear about it. I've written it all into the primer in the section "CO2 is innocent". I've written quite a lot. This was my original purpose in opening this thread.

Ferdinand, although I've ended up not agreeing with you over human contributions to the CO2 increase, I cannot tell you how grateful I am to you for your challenges. They really made me think hard, taught me a lot, improved my understanding out of all recognition, and prepared my mind to recognize the missing science when I found it.

The missing factor was the impact of Henry's Law, which actually says that CO2 will tend towards the state whereby CO2 is in equilibrium between oceans (50 parts) and air (1 part) (leaving the biosphere aside for the moment). That means 49/50ths of our emissions SHOULD vanish. The fact that 49/50ths DON'T vanish should have alerted scientists to the presence of an invisible elephant.

I want to use this forum to actually distill and advance the real science. Regarding the title of this thread, this has happened, thanks to everyone's help and challenges. Now I've found adequate clear understanding of the CO2 flux, and actually grasped Segalstad's material, I want to let the CO2 science / ocean science lie fallow and settle down. I've cracked the issues on which clever AGW scientists could still have floored me, and I want to move on to the next pressing issue, without leaving material hanging here.

Like Steve McIntyre, I want to try, where reasonably possible, to keep at least my own threads on target! But this is something that I can only gradually gain the skill to do, as key issues emerge out of the primordial soup, and gradually take shape and clarity. It doesn't happen overnight. I haven't locked the thread yet as I cannot yet see the best way to handle the very interesting ice issues here that are actually OT but deserve space on open threads for further consideration. I've not mastered the art of splitting threads - yet - I need to try to sort this out. If I cannot split the thread, you could perhaps extract the ice issues and add them to the Arctic thread I'm going to start - or start it yourself if you beat me to it.

I'm going to start an Arctic issues thread, since Arctic issues have just "flowered" for me today into a fresh page that I consider to be another "workshop page". Beautiful serendipity again! I haven't even read the ice posts carefully yet and only really "saw" them tonight. Arctic issues impinging from all directions! Clearly it's the next topic to crack.

Derek · Posted: 19 Oct 2008 10:17 pm Post subject:

Volcanoes..
_________________

An induced feeling of guilt is not sufficient reason
to convict, or punish the human race.

Questioning_Climate · Joined: 10 Oct 2008 Posts: 28 Location: Leics. UK

Lucy,

FerdiEgb does have a point with regard the smoothing used by Endersbee. Normally, the smoothing would be matched on both axes with a similar filters. In general, the variable with the highest intrinsic frequency content would be filtered to match, as closely as possible, the upper frequency limit of the other parameter.

Once this is done, there may be a phase shift between the two which can be noted and corrected before determining the ultimate correlation. It is important to realise that any phase (time) shift could be the result of a physical process and thus holds key information.

The thermal lag due to the huge thermal capacity of the oceans may well mean that changing the smoothing it will have little effect on the result.

My biggest concern is that he states a 21 year moving average was used, and shows close to the same data period. It is certain that criticism will be aimed here because the filter length is too long for the data set.

Whilst correcting the smoothing will probably make the correlation lower, it is still likely to be high, particularly when any time shifts are taken into account.
_________________
"Correlation is NOT Causation"

Lucy Skywalker · Posted: 19 Oct 2008 10:46 pm Post subject:

QC
very good, looks like the topic is not finished!

I'm a novice in stats. Endersbee did supply me with an updated graph with both axes showing 21-year correlations, as I requested; and as it made such apparently negligible difference to an already extremely high correlation, I didn't trouble to even change the graph in the Primer. But I have put a new page behind it (click on graph etc) which does talk about the issue. Perhaps I should put the new graph in the primer.

Now can you explain what this "phase shift" means?

I'm puzzled by your "thermal lag" since the graph is not time-oriented but shows the link between SST and CO2, in my understanding the oceans outgas like fizzy, in absolutely VAST overall quantities for tiny overall temp rises. CO2 mixes well enough to use MLO; the satellite readings give a good enough SST mean value, as I understand it.

I tend to agree with your concern about the 21-year filter length being too long for the data set. But it concerns me that this correlation is important to demonstrate in order to close the AGW case, and if we can factor in the possible errors owing to this unavoidable situation, we should surely still have useable results. We need to find the legit way to do this, rather than say "shelve it" because it is not legit enough. It is, at the very least, suggestive enough to warrant pursuing further until we can legitimize it to withstand AGW sniping. IMHO!

Oh yes, I want to ask Ferdinand - what is your take on Segalstad and Henry's Law and all the rest Segalstad says eg on the carbon isotopes... I think his arguments outrun yours but I came in knowing a lot less then either of you.

Where I can, I'm aiming to slice through the Gordian knots here... Rolling Eyes

Derek · Posted: 20 Oct 2008 06:42 am Post subject:

"thermal lag" - is this possibly the variations / cycles / phases of the oceanic currents adding up to the THC.
The THC appears to have some cycles upto a 1,000 years let alone the 20 to 30 odd year phases of the PDO.
_________________

An induced feeling of guilt is not sufficient reason
to convict, or punish the human race.

Questioning_Climate · Joined: 10 Oct 2008 Posts: 28 Location: Leics. UK

Lucy,

I will try to explain as simply as possible. Phase shift for a sine wave is essentially the same as a time shift, just a different way of expressing it.
Normally when you create a plot of CO2 versus temperature each point on the x-axis is matched to a point on the y-axis. Whilst they are not plotted against time, they are linked by reference to time. The time dimension is the factor that twins the points. In other words, they a re linked by time.

Imagine two sine waves then time shift one of them relative to the other. Now, when you look at the signals there will be periods when one is moving positive and the other negative. They are no longer in phase. When these are plotted together it will no longer give R^2 = 1, but something lower.

Obviously we do not have sine waves in this case and as presented there is little evidence of anything but a two relatively smooth featureless curves. As Ferdinand pointed out the correlation might not be so strong with less smoothing on the temperature.

There is a small amount of character on the MLO signal. If the SST had similar and correlated well then this might be more credible than what we have at present. If there is something to work with and there is a time shift, or none, this would tell us more about the process involved.

The thermal lag statement was not as I meant it. The heat capacity will tend to smooth the SST compared to the air, but the SST is also faster reacting than the ocean as a whole. So what is an appropriate smoothing for SST with regard CO2 flux? Should it have anything other than a 1 year moving average to remove annual variations in the same manner as CO2? Surely if El Niño and La Niña were modulating the SST then CO2 should have followed? Or were they in fact sufficiently local as to not have much effect on the CO2 at MLO? Maybe the wind at MLO was blowing in the "wrong" direction? Confused

Sorry to play devil’s advocate here, but it is important to get this as correct as possible. IMO it would be quite reasonable to lose some correlation significance by using a 1 year filter. If the correlation is as strong as it seems then R^2 around 0.8 is still possible.

Maybe Endersbee would be so kind as to provide this?

I think we also require references to the data sources.
_________________
"Correlation is NOT Causation"

Lucy Skywalker · Posted: 23 Oct 2008 04:29 pm Post subject:

Hi RC
Thanks for explanations, I think I understand.
My feeling from reading Segalstad and Anthoni is that at the surface, the CO2 fizzes so quickly that there is very little lag to worry about. However, of course, we are also dealing with the deep slow currents. But it still seems to me that such a high correlation as we have with Endersbee is highly suggestive that we've reached the point of causation - and perhaps we just don't need to worry about other lags. Perhaps we can just consider the "surface effect" since globally that is what it is, even though cross-section-wise it is not. El Nino I would expect to make changes but it is still somewhat local and CO2 whizzes round the world pretty quick as far as I can see, at least latitudinally. Perhaps EN relates to the obvious notch points?

I asked Endersbee if he would do a second graph plotting CO2 MINUS our emissions - but he has not replied so far (I wouldn't have done either in his shoes!) and on reflection this is an increasingly muddy measure. Do we measure total accumulated human CO2 or do we measure annual CO2 or something in between? I'm still not feeling confident with Engelbeen's position but also have to admit that neither do I understand it clearly enough to refute. So I've noted on the "workshop" page that the debate is still open.

I'm now thinking about asking Jeff Id to have a look at these stats. More work but will happen as soon as I can. It would be good to see him applying his clear, incisive mind to another statistical issue than HS.

FerdiEgb · Posted: 23 Oct 2008 07:18 pm Post subject:

FerdiEgb · Posted: 23 Oct 2008 08:58 pm Post subject:

Lucy Skywalker · Posted: 24 Oct 2008 10:57 pm Post subject:

Ferdinand first to say I haven't forgotten your challenge, it is a good one, but I'm still chasing my tail (done a lot of revision on the primer etc). I asked Endersbee for another graph relating MLO / SST but including our emissions; haven't had any reply so far. I've found not exactly what I requested, but MLO / our emissions, and it looks like an incredibly high fit at the rate of 57% of emissions, I'll post it on the workshop page asap. Guess I need to send THAT to Endersbee.

JUST HAD A DEJA VU

So the science is still not clear to me. I promise to work on it this weekend. Very Happy

Great meeting you BTW. I hope you enjoyed the trip as much as I did and I think Tony did. Very Happy

FerdiEgb · Posted: 25 Oct 2008 08:50 am Post subject:

Lucy,

The graph in post #2 here contains the accumulated CO2 emissions and the accumulation in the atmosphere over the full last 100 years. Contrary to the temperature - atmospheric CO2 plot over the full period, there is an excellent correlation between emissions and increase, even without the use of extra smoothing.

Smoothing only occured in the ice core CO2 measurements (average about 10 years) and the emissions figures for the first period are less accurate than in more recent decades. From 1960 on, both emissions and CO2 measurements are based on far more accurate calculations (from fuel use inventories) and CO2 measurements (Mauna Loa and other stations). These are all yearly emissions and yearly CO2 measurement averages. The slope line is:

Compare that to the temperature - CO2 increase plot:

Again, for yearly averages.
What is interesting is that relative large changes in temperature in a short time period (1992 Pinatubo - 0.4 °C; 1998 El Niño +0.4 °C) have little effect on total CO2 levels, while the overall increase in temperature of about 0.8 °C should be the cause of +70 ppmv CO2. That is not very probable...

From the Pinatubo and El Niño events it is possible to calculate the short-term influence of temperature on CO2. That is about 3 ppmv/°C. The variation in global CO2 levels vs. global temperatures over the seasons is of the same order: about 5 ppmv/°C and the very long term influence of temperature on CO2 (including changes in ocean currents, ice/vegetation cover,...) over the past near million years (Vostok, Epica) is about 8 ppmv/°C.

Thus while there is a short time influence of temperature on CO2 increase rate, the bulk of the total increase is caused by the emissions...