Bishop Hill Cool exchange
Apr 12, 2012 Climate: Sceptics Climate: other A couple of weeks ago Tamsin Edwards discussed what I think might be a better way forward for those who are interested in understanding the climate debate.
I think a large part of the audience who visit this blog (thank you) contradict these findings. Your trust in the science increases the more I talk about uncertainty! And I think you place greater importance in “calculative” rather than “relational” trust. In other words, you use the past behaviour of the scientist as a measure of trust, not similarity in values. I’ve found that whenever I talk about limitations of modelling, or challenge statements about climate science and impacts that I believe are not robust, my “trust points” go up because it demonstrates transparency and honesty. (See previous post for squandering of some of those points…). Using a warm, polite tone helps a lot, which supports Hebba’s findings. But I would wager that the degree of similarity to my audience is much less important than my ability to demonstrate trustworthiness.
Reader Comments (166)
I am reminded, possibly unfairly, of those times when one is asked to admire the artistic output of a friend's child. No matter how good they are 'for their age' they really don't amount to a picture you'd put on your fridge alongside the brilliant offerings of your own sprog. These models are like Sam Johnson's dog walking on its hind legs. All very clever, you wouldn't want to squash it with your contempt, but you don't have to take it seriously for a minute. So in this case, let's pat it on the head and get on with something serious, while commending Richard et al for their valiant attempts and hoping they'll crack it some day.
(Retrospective forecasting is NOT cracking it.)
Michael,
Assuming you are in the UK, could you share name and address of a "denier bar" or two. SWMBO and I are planning our annual visit to the Island to savour civilisation, and would love to visit one.
Having read the comments on this thread, in my opinion all have completely missed the fundamental point - what Richard, Tamsin and others are trying to do is MATHEMATICALLY IMPOSSIBLE.
Unverified and unvalidated computer models can never provide useful information on the state of a system. Numerical simulations containing a large and uncertain number of dependant and independant vairiables, acting within a non-linear and sometimes chaotic system, can have no predictive power AT ALL!
I am sorry to use capitals (shouting) but to me this is an absolute and fundamental truth. If anybody thinks that I am wrong I beg you to tell me why. However, if I am right, Richard, Tamsin and many others are completely wasting their time and our money. Moreover, the consequences of ignorant people in positions of power paying heed to their work are potentially catastrophic.
I have made this simple point several times on this site and it has never been refuted. I now ask that Richard and/or Tamsin come here and address it directly.
"Unverified and unvalidated computer models can never provide useful information"
So would you accept that a verified OR validated model can?
On a similar note, just because you've failed to persuade others to address your concerns, failings, bad manners, etc. DOES NOT (ALWAYS?) MAKE YOU RIGHT.
Climate model
Apr 16, 2012 at 10:21 AM | Michael Saunby:
Validation (software meets its intended puropose) implies verification (the correct use of algorithms). Climate models fail on both counts.
You are correct that I am not always right - that is why I begged anybody to tell me why I am wrong. I am sorry that you consider my post to be "bad manners" - how exactly?
Steve Easterbook is your man for questions about validation of climate model. He has an interesting blog at http://www.easterbrook.ca/steve/?p=2734
"I now ask that Richard and/or Tamsin come here and address it directly." Is RUDE. If you don't know why it's rude ask your mum.
Apr 16, 2012 at 11:58 AM | Michael Saunby:
The link you gave does not address verification or validation.
Is it rude to ask why you, Tamsin, Richard and a host of others are spending my money on something that, in my opinion and for the reasons stated, can never give useful answers? You are government scientists who should be accountable for your actions.
I didn't claim it did, only the Steve has given such matters some thought. Indeed I believe he has published papers on the subject. I've given you a pretty decent hint as to where to start looking. If you really believe such models cannot ever work then how the heck do you imagine the weather is ever forecast. Or do you imagine that little fairies whisper in the forecaster's ear?
Yes it is rude - whatever you believe. If you have a GENUINE concern about how taxes are spent there are proper, tax payer funded, processes for addressing such concerns. Expecting anyone you chose to add comments to a blog to feed your personal desire for little tidbits of easily digested knowledge really would be a waste of time and money - wouldn't it?
Enough of this feigned outrage to change the subject (as usual). The key words are "mathematically impossible". Any competent mathematician can tell you why. Which words do you not understand?
Weather forecasts work over short periods because the timescale of the prediction is less than the timescale given by the Lyapunov exponent of the chaotic dynamics.
It is good to have another person from the Met Office commenting here (Michael Saunby maintains the excellent Met Office web page of HADCRUT-related data).
Michael I do think the concern here is with the public face of the MeT Office, press office leaders say, rather than what the scientists think.. and the scientists then get caught up in the politics.
May I suggest you change your tone a little, Tamsdin and Richard do well here, because they think some of us, are not climate change deniers..
May I ask what the proper channels are.. because they seem to be not listening...
"mathematically impossible". Any competent mathematician can tell you why.
Which, as you know is a stupid lie. Why would you lie about such a thing? Do you imagine there are people reading this that are smart enough to read but so stupid as to believe such a daft claim?
Try to control your temper Dr Saunby. Accusing people of lying does your reputation no favours.
(Note the thread title!)
Michael, Roger
Calm down please.
Oh dear, poor old Michael has lost the plot.....
May I RESPECTFULLY request that somebody from the MET office, with an understanding of mathematics, come here to address the matter? (Hint, Paul Matthews has already given you part of the answer).
Yes "weather forecasting works" using models that Prof Longstaff claims are "mathematically impossible" and "can never provide useful information".
So which of us is lying? We can't both be telling the truth can we now?
I guess what the honourable Prof is failing to say is that he believes models can tell us nothing about the distant (climate) future. That would surely require some rather careful study from both those arguing for and against such applications, and clearly doesn't fall into that nice simple category of "impossible". To suggest, as he does, that the no competent mathematicians have contributed to the science and any who might critique the models would dismiss them as useless is as close to insane as I can imagine.
Michael,
Nobody is accusing you of lying - please calm down.
"weather forecasting works" for a few days (sometimes), however Met office long range forecasts were hopeless (barbeque summer, etc.) and had to be stopped. It has been stated that "climate forecasts" use the same models.
My point is that models that use incomplete or inaccurate algorithms (and/or data) produce errors that propogate (sometimes exponentially) in a numerical integration. As all admit that algorithms to simulate albedo (to take just one example) are at best incomplete, it follows logically that the longer the integration the worse the prediction becomes. Indeed, "worse" is the wrong word - they are just plain wrong.
I am not trying to propogate a lie. As I stated at the outset - if I am wrong, please tell me why.
Judith Curry (Climate Etc) has a reviewers comments of the Steve Easterbrooke paper, and it is not that complimentary
Michael have you seen that, any thoughts?
Am I seriously expected to believe that weather forecasts derive from models with no regard to what ever is a day or two upwind at the moment? That success in short-term forecasting is any indicator that a model informed by satellites and thermometers and other equipment will hold out for more than a day or two? I do not so believe.
I got to thinking that a weather forecast that said, say, tomorrow's weather will be much like today's would be right up to two-thirds of the time right here in the UK. Then I got to thinking that if I were to substitute days for years, decades or centuries I wouldn't be too far out either. Weeks or months probably not. How would the MO model compare with some such simplistic model?
Rhoda,
Met Office forecasting accuracy is currently the top subject for discussion at WUWT:
http://wattsupwiththat.com/2012/04/16/the-met-office-coping-response/#comment-957622
Rhoda, absolutely not. In my book nobody should be expected to believe anything. Indeed regardless of opinion I would absolutely defended anyone and everyone's right to believe nothing!
Your simplistic model is what we would call "climate" and deviations from such is either variability or change (to adopt another simplistic model) - separating one from t'other is the hard part. So you are correct in that with no model at all you could reasonably state the average temperature for the night of 16th April at any location of your choice, say Nottingham. Of course to do this you would first need to collect measurements for many years. Fortunately this is another of the tasks that weather services carry out and the results are something like this Nottingham statistics No belief required, these are the facts. However collecting data for many years might reveal other things that some folks are uncomfortable with - perhaps because of their beliefs.
The performance of models is of course relative to climatology. However stupid some may think us to be we wouldn't do what we do if guessing gave a better result. We are not astrologers.
There's some introductory material about Met Office models here - Unified Model
My simplistic mdel is not guesswork. If I were to take it up, I'd constantly review it's performance against reality. I don't need years of weather data, that is a construction you have put on it. It would be reviewed against results and modifications to the original model proposed. All without knowing what weather is a couple of days upwind, without having a weather map or isobars or anything measured save the actual in my target area. It's like knowledge-based software. However, it doesn't matter how it works, the point is the possibility that constant refinement might make it perform in the short-range area. I don't know how well the met model performs. Now it is a piece of software for sale, I will not trust the marketer to tell me either.
Answer me a question that isn't in the link you posted. How does the CO2 proportion for today get into the model? Or is it not in the short-term variant but in the longer-term ones? At what timescale is it included? Does it work from first principles, or is the 'sensitivity' assumed using IPPC-like figures?
Note, I do not think you are all stupid and I have some magic ability to tear away the curtain. I do think the met office is agenda-driven, what bureaucracy is not?
Rhoda,
"don't need years of weather data, that is a construction you have put on it" Don't you? So how does your scheme for days/weeks/months/years/decades/centuries actually work? If you have a "magic" machine that can tell you what this week's temperature is, never mind "this year's temperature" they please do tell.
Also how do you know what the weather is "a couple of days upstream"?
Whilst it's true that 150 years ago early attempts by Fitzroy and other relied on coastal observations and later weather ships were employed to observe weather out at sea before it reached our coasts, it would take someone with an outrageously retro view of science and technology to seriously propose such a system today. So I get that what you are conducting is some sort of thought experiment and kind of gewt how you're thinking - hell I can actually hear the cogs grinding.
Where do you want to go? Somehow prove that at bureaucracies would hide the blinding success of a primitive forecasting system in order to dupe the entire globe into adopting numerical forecasting?
As for the CO2 question. First show you have the SLIGHTEST grasp of what you are discussing.
What I'm getting at is that such a primitive forecasting system might just rival the model in short-term forecast accuracy depending on how you measure that. Two schemes , in fact. The knowing what's upwind, which is how the met worked totally or in part in the pre-computer age, simplistic at first and then informed by experienced experts. I think that would be pretty good. The other is just a thought experiment indeed. It is that some simple formula, such as today will be like yesterday, unless you've had three days the same, in which case expect a change. Maybe a little more complex than that. NOT April 16th is always showery. No, April 16th last year or a hundred years ago is of no relevance, and I don't need the records to work the formula. (Nor will it ever tell me of a surprise event). I think the met is completely invested in technology and would not consider such schemes for a moment, nor should it. But for an interested observer to see just how close you can get to short-term accuracy by these methods or any other would be interesting. Who validates the met predictions against actual outurn? The met? Anybody independent? Note that I have made no aspersions on that accuracy, I don't know how good it is. I don't know what criteria are used. I would think that forecasting surprise events, storms, snow, disruptive weather, is of far more value than telling you today is like yesterday (an it is, today, so far).
On the CO2, I'm asking questions, and you are mocking my ignorance. Arrogant bastard. When you expand from a short-term forecasting model to a climate model for decades or centuries, where does the change in CO2 level enter the model?
Good for you Rhoda. So much for the Met's "public outreach". For all I can tell Richard Betts may likewise really think we're mostly a bunch of idiots, but at least he's polite enough not to show it.
I wonder if Mr. Saunby of the Met Office realises that I could take legal action against him for branding me a liar on Apr 16, 2012 at 2:07 PM? A court case could be very interesting because he would have to justify his claim.
I give him this opportunity to apologise.
No no, Roger, he is engaging with the sceptics to improve communication. It's constructive.
Rhoda,
I use my real name here and Saunby attacked my reputation - potentially damaging my business.
If he apologises like a gentlemen we will say no more about it.
Sorry Roger, missed the /sarc from my last. You must do as you see fit, but it will be a shame if all the met folks leave because of it. Although I have not yet seen anything useful coming from the 'constructive exchange' I don't eliminate the possibility.
Thanks Rhoda. If Saunby apologises that will be an end to the matter, and we can agree to disagree.
Re - "I wonder if Mr. Saunby of the Met Office realises that I could take legal action against him for branding me a liar on Apr 16, 2012 at 2:07 PM? A court case could be very interesting because he would have to justify his claim."
Roger, indeed there are laws regarding such matters. To save unnecessary expense may I suggest, without prejudice, that in the first instance you take this matter up with the publisher.
Kind Regards,
Michael Saunby
"take this matter up with the publisher."
Your Grace, I think he means you.
Mr. Saunby,
Without prejudice,
May I ask who owns the computer that you sent the libelous communication from on Apr 16, 2012 at 2:07 PM? May I also ask if you were in paid employment at that time, and if so by whom?
Sincerely,
Roger Longstaff
Apr 18, 2012 at 8:00 AM | Rhoda
Hi Rhoda
Thanks for asking a perfectly sensible question!
When the model is used for weather forecasting, the atmospheric CO2 concentration is prescribed at the present-day value.
When it's used for simulating climate change under a GHG emissions scenario, the concentrations of CO2 and other GHGs (e.g.: methane) are either updated periodically using numbers taken from a scenario of CO2 rise provided from elsewhere, or the change in CO2 is simulated within the model itself on the basis of a scenario of emissions and also simulated sinks and sources of carbon in land ecosystems and the oceans.
Does that help? If not, let me know.
Cheers
Richard
Great 'outreach' guys. keep up the good work. Heh.
"If you have a GENUINE concern about how taxes are spent there are proper, tax payer funded, processes for addressing such concerns. "
Mr/Dr Saunby
I don't pay taxes in your country but I can get in line like a peasant ask you a question?
Please look at this diagram. It is the IPCC projections for different scenarios. I draw your attention, specifically, to the *range* of possible outcomes occupied by the projections. It spans approximately one quarter of the total range of all possible outcomes going forward. The yellow line is virtually horizontal, and the red line is ~45 degrees. Among the possible outcomes, the ones in the shaded areas are highly unlikely - corresponding to a greater-than-already-observed rate of rise, or fall in temperatures being sustained over multiple decades. The remaining possible valid prediction space corresponds to the remaning two unshaded quadrants. Of these, the IPCC has simply occupied one whole qauadrant of available prediction space.
That's is like buying half the lottery tickets ever printed for the next 100 years and expecting to win.
Where is the predictive skill in such an exercise?
And yes, for a model to to produce mathematical proof - it has to predict the absolutely exact outcome - and nothing else. This means, for a model to be mathematically possible - it has to predict the exact temperature of every single point in the earth system, every second, every day going forward. Since we all know that is not how climate models work, we have to gravitate toward the known conclusion - they are not mathematical models.
Which is what Roger's point was.
http://nigguraths.wordpress.com/?attachment_id=2126 Link for the diagram above.
Mr/Dr Saunby - I too would very kindly appreciate an answer to Shub's question. (and for my sins, I am a taxpayer in the UK).
Apr 19, 2012 at 3:48 AM | Shub
Hi Shub
For predictive skill, the early versions of these models were used in the early 1970s to project a warming of 0.6 degrees C warming by the year 2000 so 0.2 degrees C per decade. For comparison, HadCRUT4 shows that the rate of global warming from 1979 to 2010 was about 0.17 degrees C per decade.
Your lottery analogy is wrong because that implies it's all based on past statistics, whereas it's actually based on physical reasoning (expressed in a numerical model).
The projected rate of future warming depends to a large extent on the future trajectory of emissions - we project faster warming for higher emission scenarios, and even for a scenario where the emissions keep increasing at a constant rate, the build-up of CO2 in the atmosphere is simulated to lead to a faster rate of warming.
We don't know which emissions scenario is more likely - too early tell (see discussion on pages 71-72 here, which I think is still valid now).
We're pretty confident, however, that the yellow line is highly unlikely as that assumed CO2 concentrations being kept constant at the year 2000 levels, which has already not happened, and it's looking like it will be a while before emissions peak....
Cheers
Richard
Hi again Shub
BTW regarding "outreach", anyone from the Met Office who posts here (including Michael Saunby, myself and anyone else) is doing it completely off their own bat, and is expressing their own views and not those of our employer.
Cheers
Richard
I was going to post this yesterday but waited to ask Richard if it was OK to quote his email.
Chris M, I agree that it's impossible to prove, but I can vouch for Richard not thinking you are mostly a bunch of idiots :) Quite recently he was emailing Keith Kloor about his motivations for all this online stuff, the Bishop's visit to the Met Office and so on, and copied me in:
"So, my aim is not really to try to convince Andrew [Montford] (or indeed anybody else) that global warming is a problem - that is a personal judgement, and anyway there's plenty of other people trying to do that kind of thing! Rather, my aim is to let him/them see that the evidence informing that judgement comes from real people with credible expertise and whose motivations are merely to seek and communicate the scientific truth. That will hopefully then form the basis for a more useful discussion, in which the evidence can then be discussed in a more informed (and less suspicious) manner."
I think that sums it up well.
Shub, I'm not disputing that the uncertainties are large, but can I just second the point that the yellow line on your graph is not intended to be a projection for a "plausible future", unlike the others. It's more of a control experiment to see what the model predictions are with no change in forcing (e.g. taking into account the lag in the system from pre-2000 forcing).
Tamsin
Richard, Tamsin,
If I understand it correctly, there has been no statistically significant warming for 15 years, despite a continual rise in the concentration of CO2 in the atmosphere. Is this correct? Do you have any Met Office published results from around 1997 that give predicions for 15 years, and if so, how do these compare with the reality?
I would also welcome your comments on a previous statement, that rather got lost in a nasty row:
"weather forecasting works" for a few days (sometimes), however Met office long range forecasts were hopeless (barbeque summer, etc.) and had to be stopped. It has been stated that "climate forecasts" use the same models.
My point is that models that use incomplete or inaccurate algorithms (and/or data) produce errors that propogate (sometimes exponentially) in a numerical integration. As all admit that algorithms to simulate albedo (to take just one example) are at best incomplete, it follows logically that the longer the integration the worse the prediction becomes. Indeed, "worse" is the wrong word - they are just plain wrong.
These are genuine questions. If you have the time to answer I will be grateful, if not I will understand.
Roger - I have to do new results for, and prepare, two conference posters today so am about to go offline :)
A very off-the-cuff, not double-checked answer is: (a) I think the warming is statistically significant, (b) even if it isn't, or only barely so, the effects of El Nino and La Nina etc on annual temperatures are so big they can alter the trend of 10-15 years so you have to look at multi-decadal trends to be sure (c) yes, seasonal predictions use climate models but they are trying to hit the exact trajectory of weather - i.e. the year-to-year variations, the exact bumps and wiggles - rather than the long-term statistical properties, i.e. the probability distribution of bumps and wiggles, which is that long-term climate projections (conditional on particular emissions scenarios) are attempting to do. Error propagation is indeed important if you are trying to predict a trajectory (like weather and seasonal prediction) but in a forced problem (a problem of the "second kind" -- Lorenz) it doesn't have the same importance because the system evolution is dominated by the boundary conditions. There's a paper by (Ed) Hawkins and (Rowan) Sutton that estimates the transition (in time) between the two types of problem.
Of course, you're right that if processes aren't represented well, or at all, then the output will be wrong. (*cough* allmodelsarewrong.com *cough*). The carbon cycle is a big one - see Ben Booth (same guy as aerosols/AMO being discussed elsewhere)'s paper just accepted http://stacks.iop.org/1748-9326/7/024002 They vary the uncertain parameters of the carbon cycle part of the Hadley Centre model and find: "the range of CO2 concentrations arising from our single emissions scenario is greater than that previously estimated across the full range of IPCC SRES emissions scenarios with carbon cycle uncertainties ignored".
Sorry this explanation is a jargon-y brain dump! If I had more time I would explain it better. But perhaps you are familiar with many of these terms anyway.
OK, don't let me come back here until my posters are done.
Hi Roger
It all comes down the fact that there is both internal variability and long-term forcing. The former is certainly more important in the short term (years to a decade or two), while the latter is confidently expected to more important in the long term (many decades), see Ed Hawkins' seminal paper.
The internal variability is very, very difficult to predict due to the chaotic nature of the atmosphere. We are only just beginning to make some progress in that, and it certainly wasn't possible in 1997. Internal variability was, however, simulated reasonably well in a statistical sense, ie: the models got the general character of the variability OK, but it was not possible to get it right for individual years or individual decades. Ben Santer's paper is relevant here.
The longer-term increase in radiative forcing expected by GHG increases is what gives us the chance to project changes in the longer-term character of the climate. We still won't get individual years right, that's simply not possible, but we can reasonably confident that, for example, the average global temperature over the period 2050-2080 will be warmer than now, by an amount which depends on both the emissions of GHGs and feedbacks in the atmosphere (and some component of internal variability too).
Seasonal forecasting (BBQ summer) tries to forecast the internal variability. The BBQ summer disaster was essentially a failure of communication. The UK seasonal forecasts were always known to have only 60% skill, but (and this is the bit where I prove I'm speaking as an individual and not corporately!!!) the folks in charge of communicating this (who were not climate scientists!!!) did not fully appreciate this point. After a few years of reasonably successful experimental seasonal forecasts, it was thought OK to release press releases about them and attract attention. Despite an attempt at careful wording ("Odds-on for a BBQ summer") this was clearly a massive error of judgement! Those of us who knew about this stuff were immediately worried about such a statement being made, but it was too late..... and the rest is history :-)
Hope this helps!
BTW Tamsin is right that I don't think people here are mostly a bunch of idiots! (Thanks Tamsin!)
Can't stick around today I'm afraid - IPCC AR5 won't write itself..... :-)
Richard, Tamsin,
I asked the question "Do you have any Met Office published results from around 1997 that give predicions for 15 years, and if so, how do these compare with the reality?"
A post at WUWT summarised historical Met Office predictions at:
http://wattsupwiththat.com/2012/04/16/the-met-office-coping-response/
Do you think that the WUWT analysis was fair, or did it distort the truth? If it was fair, why do you think that the predictions were so far from reality?
Tamsin,
Thank you very much for your kind words - I appreciate them and I imagine others will as well. I was originally happy to accept Richard's efforts at engagement at face value, and saw it as a very positive development, but then my view became a little jaundiced when it became clear that our perceptions of reality are very different, perhaps immutably so.
The only way forward that I can see is to keep talking to each other, kindly, sincerely, respectfully, in the hope of finding some common ground. To be more specific I believe it would be very helpful for the Met Office and similar institutions to move away from implied support for CO2 mitigation (which cannot possibly work due to the essential energy needs of China, India and developing countries) towards a concept of adaptation if (a big if) this proves necessary. I understand that this is Judith Curry's position. It is not enough to imply, as Richard's colleagues did in their reply to Prof. Lindzen, that mitigation is the only option that governments need to consider. (All of that assumes, as I don't, that rising CO2 is a problem.)
I believe it is a grave mistake to think that climate science can be kept separate from politics after all that has happened. Genuine climate scientists need to recognize that in their position of influence they have a social responsibility to make things better, not worse. I would like to think that genuine people like you and Richard can be part of the solution to a policy issue that has long since transgressed the bounds of rationality, imho.
Richard,
Thank you very much for your reply (our last posts crossed).
You state that "The longer-term increase in radiative forcing expected by GHG increases is what gives us the chance to project changes in the longer-term character of the climate".
I am trying to put this as respectfully and politely as I can, but there is no other way to say it - in my opinion thereby lies the fatal flaw in your logic. In my opinion it is circular reasoning. If you are looking for a defining signal in noise and errors that are propogating forward in a numerical simulation that is all that you can hope to find, for all of the reasons that I, and others, have expressed.
This is my personal opinion after 30 years of involvement with computer models, in a wide variety of applications (and I have the scars to prove it). I hope that I am not misrepresenting the views of others, but it seems that just about everybody who visits this site (many of them seemingly with a wealth of experise) agree. I am sure that they will correct me if I am wrong.
Roger Longstaff:
"I am sure that they will correct me if I am wrong."
At the moment, only the spelling of "propagate", Roger :)
For my own clarity of thinking I always remind myself of the questions that arise which I've posted on BH in reply to the open letter by iwannabeasceptic:
http://www.bishop-hill.net/discussion/post/1729708
What was predicted [with error bars/uncertainty]?
When was that specifically predicted?
Was it a significant prediction?
Did it happen? [i.e. the data, with error bars/uncertainty]
David Lilley seems to understand this quite well, I'm glad to say.
Hi Tamsin,
Thank you very much for your reply (again, posts crossed).
My very genuine problem is I simply cannot understand the logic in your statement: "the effects of El Nino and La Nina etc on annual temperatures are so big they can alter the trend of 10-15 years so you have to look at multi-decadal trends to be sure".
I do not understand how you can be "sure" about multi-decadal trends when models have such a poor track record over decadal timeframes, for whatever reasons (please see my "circular reasoning" reply to Richard, above). If anybody can explain how this logic can be correct, without prejudicing the results in relation to GHGs, I would be genuinely grateful (this is not a "trick" or "points scoring" question).
I understand that both you and Richard are very busy, and if you eventually get time to reply I will be grateful.