How would you guys respond to the following statement?

"The principles of a climate model are very similar to other fluid dynamics simulators. Now one of the great shortcomings of these simulators is that the cell size is generally too coarse to get reliable results as there simply isn't sufficient computing power available to run simulators with smaller grid sizes. Generally these simulators run on a much smaller scale than world climate simulators, and even then can't cope very well with turbulent flow or mixed phase flow. Now for someone to tell me I should trust a climate model with an incredibly coarse grid scale and far more complex fluid interactions well, understandably I have to suspend belief."

---

personal response?

-Don't tell me, he's the senior scientific advisor to George Bush.

No doubt someone else will provide the fluid dynamic point of view.

Marj

---

_________________________________

Nothing like that :) Its from a guy whos expertise is in numerical analysis and who have doubts about the 'warnings' coming from these coarse climate models. I know enough science to be dangerous, but on this point I am clueless. So I would like to know how best to respond to a statement like that. Is there anyone that can give the question a NewScientist styled answer? :)

---

> Nothing like that :) Just a guy whos expertise is in numerical analysis and
> who have doubts about the 'warnings' coming from these coarse climate models.
> I know enough science to be dangerous, but on this point I am clueless. So I
> would like to know how best to respond to a statement like that. Is there
> anyone that can give the question a NewScientist styled answer? :)
>
I guess I'd like to know what the person asking the question means by "reliable".

Not sure if you have seen this

http://www.climateprediction.net/science/model-intro.php

introductory writeup on climate models, what they can and can't do. And this writeup at realclimate.org on whether climate modelling is a science:

http://www.realclimate.org/index.php?p=100

> I guess I'd like to know what the person asking the question means by
> "reliable".
>
> Not sure if you have seen this
>
> http://www.climateprediction.net/science/model-intro.php
>
> introductory writeup on climate models, what they can and can't do. And this
> writeup at realclimate.org on whether climate modelling is a science:
>
> http://www.realclimate.org/index.php?p=100

Cheers for the realclimate article link. Hadn't read that piece eventhough I go there often enough. I read the model introduction article before, but just didn't put two and two together.

So in essence, the large grid sizes means that any theoretically important subgrid phenomena is modelled with a parameterised value, and its these values and how sensitive the model's are to the choice of these values that CPDN is testing. I think the function of the 'parameterised values' and how they make coarse grids ok for global climate modelling is the key point for this guy to understand.

Let me know if my understanding is whacked...

Cheers,
Hasse

---

Fluid dynamics simulators have been used for many years to design aircraft wings, true they can't model turbulent flow very well, but they are calibrated against wind tunnel tests and do a pretty good job of predicting real world wing performance.

Climate (weather) simumlators have been used for many years to model climate and similar codes at higher resolutions for weather forcasting, true there are some aspects of climate they cannot model very well, but they are calibrated against past climate (and weather) and do a pretty good job of predicting real world future climate (and weather).

---

____________________________<br>
<a href="http://www.boincforum.info/boinc/">boinc forum</a> and <a href="http://www.uk4cp.co.uk/">United Kindom</a> team, my climate change <a href="http://www.livejournal.com/users/mike_atkinson/">blog</a>.

&gt; How would you guys respond to the following statement?
&gt;
&gt; "The principles of a climate model are very similar to other fluid dynamics
&gt; simulators. Now one of the great shortcomings of these simulators is that the
&gt; cell size is generally too coarse to get reliable results as there simply
&gt; isn't sufficient computing power available to run simulators with smaller grid
&gt; sizes. Generally these simulators run on a much smaller scale than world
&gt; climate simulators, and even then can't cope very well with turbulent flow or
&gt; mixed phase flow. Now for someone to tell me I should trust a climate model
&gt; with an incredibly coarse grid scale and far more complex fluid interactions
&gt; well, understandably I have to suspend belief."

It's an interesting position. In many ways it is a completely fair criticism: we know that small-scale processes can manifest themselves on large scales in finite time, and that the chaos thus induced means things like weather forecasts lose all skill on timescales longer than about two weeks. However, as long as the aggregate features of the sub-scale processes are adequately parameterised, there's no reason (in principle) why this should affect the statistics of the output (ie the climate of the model). In some cases the models and parameterisations do a good job of simulating recent climate and climate change, in some cases they do a worse job (kind of a large part of what this experiment is about).

So I accept that climate models are very far from perfect tools, and their predictive skill and ability to represent a system may not satisfy someone used to dealing with problems in alanytic chemistry or spectroscopy or whatever. But that's because they're looking at much simpler systems. In those cases, the energy scales, spatial scales and number of important processes are much more restricted, and this allows them to do much more "exact" science. We, on the other hand, are looking at the entire planet, across every scale from the molecular to the global, with a single realisation of history and an inability to do lab experiments (we lack the time and the second (control) Earth).

The real difficulty I have with such puritan conceptions of science is that they are usually belied by other beliefs. Usually, climate models are dismissed as inadequate (in the way described above) by people who purport to be disturbed by the idea of (say) HadCM3 claiming representational skills, but who are quite happy to accept macroeconomic forecasts of, inter alia, the future cost of greenhouse gas mitigation policies. Now if you don't accept climate models because they (a) fail to capture interactions between the small and large scales; (b) are too coarsely resolved; (c) deal with non-linearity too poorly for one's purist tastes, then you have absolutely no business wahtever accepting ANYTHING that comes out of a macroeconomic model, and very little business accepting anything from microeconomics, either. [Presumably evolution, plate tectonics and a variety of other commonplaces are taboo as well, to say nothing about many of the common heuristics we use to navigate our way around this messy (non-linear, etc.) world.]

So I think it's a position akin to those found in first year philosophy classes when people claim to be sceptics about the real world (suspending belief in the table in front of them; the chair thay are sitting on). It's easy to make the claim, but impossible to actually live up to. The moment the student leaves the tutorial room they act, of course, as though the real world really does exist, in conflict with their stated scepticism. The moment the climate sceptic takes seriously socio-economic forecasts, or any number of other quite ordinary scientific claims, they undermine the epistemological piety they wish to proclaim. As with many other sorts of fundamentalism, it's difficult to see why those of us who have to live in the real world - with its messiness, non-linearities and unattributable phenomena - should either emulate them, or take their policy recommendations seriously.

[And that's not even going near the whole Bayesian/degrees of belief turn in the philosophy of science.]

Dave

---

why is it always the ones that insist on 99% confidence for global warming but they take absolute hearsay as "truth" to go to war, or just blind faith to believe in "Creationism" over science etc? ;-)

The researchers will have taken into account the size of model that most crunchers are willing and able to commit to. (There are still periodic complaints on the boards about the size of the models - length of time to complete and HD space required to store them after completion). The present grid-size resolution is allowing lots of models to be completed; the number of completed models is also important for a spread of results. Future models will be including more factors, and I think some of them will take more computing time/power.

The experiments are designed to show likely global climate change, but not at the local micro-climate level.

The Japanese Earth Simulator will have smaller grid sizes. But it takes an area the size of 4 tennis courts full of servers.

Cpdn results so far seem to correspond well with climate changes that we are already noticing in the real world in Europe.

---

__________________________________________________

&gt; The researchers will have taken into account the size of model that most
&gt; crunchers are willing and able to commit to. (There are still periodic
&gt; complaints on the boards about the size of the models - length of time to
&gt; complete and HD space required to store them after completion). The present
&gt; grid-size resolution is allowing lots of models to be completed; the number of
&gt; completed models is also important for a spread of results. Future models will
&gt; be including more factors, and I think some of them will take more computing
&gt; time/power.

That's right, but Moore's law is helping us all the time. It took us six weeks (including the Christmas dip) to go from 50K runs to 60K runs. That was pretty impressive. Ellie's going to slow it up for a while with the sulphur cycle stuff, but I'm sure we'll get back to the same number of completed runs/month pretty soon.

&gt; The experiments are designed to show likely global climate change, but not at
&gt; the local micro-climate level.

That's right. Mostly we're looking at continental scales, though we do have a few ideas about having a look at some stats on smaller scales, too, though that's a bit further down the line.

&gt; The Japanese Earth Simulator will have smaller grid sizes. But it takes an
&gt; area the size of 4 tennis courts full of servers.

It does have awesome graphics, though. I saw a demonstration in Sapporo and was really impressed. You can see some cool equatorial wave structure in the ocean, and little eddies off the gulf stream. [Tolu would love it, I reckon.]

Dave

---

&gt; That's right. Mostly we're looking at continental scales, though we do have a
&gt; few ideas about having a look at some stats on smaller scales, too, though
&gt; that's a bit further down the line.

By embedding a higher resolution grid region in a lower grid resolution model?

---

____________________________<br>
<a href="http://www.boincforum.info/boinc/">boinc forum</a> and <a href="http://www.uk4cp.co.uk/">United Kindom</a> team, my climate change <a href="http://www.livejournal.com/users/mike_atkinson/">blog</a>.

This is my first post ever to a CPDN-related forum. I have been silent lurker though a heavy donator to CPDN. I have four P4-class threads running full time on the model, and would have a fifth but I fried the processor while assembling the machine.

I have been silent because I will find myself at odds with nearly everybody on these boards. Perhaps "odds" isn't enough to describe the difference. Loggerheads may be a better term, at least with regard to politics, religion, and climate. I choose not to say anything on these forums because for the most part, I think that there is no tolerance for dissent. Indeed, my observations of board contributors lead me to conclude that there are a good number of people for whom ecology is religion, global warming dogma, and man's causality thereof immorality. These are the same people who would dismiss my religion and politics because they require aspects of humanity that transcend science. Science triumphant? No way. What an inhuman view of the world. If sciece is the pinnacle of human existence, I for one perceive such existence to be hopeless.

That said, this thread has definitely piqued me. I am a mechanical engineer by training, and I specialized in combustion. It's the only graduate course I took! Mr. Frame's contribution here really resonates with me. Why? Because he points out the great intellectual dishonesty which can exist within any single mind. But more importantly, he bellies up to the real issue, which is an unwillingness of people, particularly those with the "curse of the intellectual" to say "I believe."

Do *I* believe there is global warming? I consider the question irrelevant. MUCH more relevant is, "If there is global warming, to what degree is it destructive?" And, "If there is global warming, AND it is overwhelmingly destructive, to what degree did man cause it?"

THESE are the questions that motivate me to contribute to CPDN. Is the grid size in CPDN good enough to answer my questions with any serious certainty? No. Furthermore, grid size notwithstanding, the model parameters probably do an inadequate job parameterizing phenomena which should really be part of the computations: fluid flow, heat transfer, precipitation, etc. But it all must start somewhere, no? This is a perturbation analysis. It doesn't exist to predict results; it exists to validate models. And those models will serve as the shoulders for somebody who can shrink the grid, and refine, refine, refine.

What's so wrong with saying "I believe?" It's human. It's natural. It's NOT unscientific. Faith and science can coexist. If they couldn't, why would we bother with this?

---

Hi Dave R

Welcome to the forum; there may be plenty more contributors to cpdn with similar views to your own who may need the impetus your post provides to speak up.

Your two questions, 'If there is global warming, to what degree is it destructive?' and 'If there is global warming AND it is overwhelmingly destructive, to what degree did man cause it? are both of paramount importance.

But I think that the question you consider irrelevant, 'Do *I* believe there is global warming?' is also very important. Here where I live, S England, most people are even now aware of signs showing that GW is already happening. There are already studies being carried out into the possible future environmental effects of GW (see Mike Atkinson's separate post listing publications from a variety of sources). I don't think anyone yet really knows what the global impact of GW will be, say on agriculture, because the global climate predictions are still so uncertain and local predictions even more so. But a rise in sea level seems inevitable, and bearing in mind that so many people live on very low-lying plains near the sea, and that this effect of GW will be catastrophic for them, I feel it would be wrong to delay action.

We don't know whether human activity is the whole problem, but I think the patterns of the CO2 (and other gases) emissions point to it being part of the problem ie the part we can actually do something about.

I think the cpdn experiment DOES exist to predict results, but in order to do this, the models have to be validated first. And, as you say, refined again and again.

I don't personally think that, even if GW is shown to be largely caused by human activity, this means that people are immoral. Industrialisation and the widespread use of fossil fuels started long before anyone guessed that this could have important effects on climate and air/water quality. However, if you do know, and are capable of taking action but choose not to, then the moral situation is different. More than a question of morality, this just seems common sense.

I hope you post again, as I am interested to see what you think.

---

__________________________________________________

As Mo says, personal experience and the consequences for that individual are important motivators for action. For the person of faith, though, the possible impact of human and natural activity on others, and upon the world, is always of concern. The question 'why should I care if it won't affect me?' hardly arises for them. That is shared, of course, by many people with strong commitment to a personal philosophy even though they may not regard themselves as religious.

It is hardly surprising, therefore, that a project like CPDN should attract a significant proportion of people with strong religious or political views, and it would be extraordinary, and even worrying, if they all agreed with each other about these things. I do not think, though, that those beliefs should govern their judgement on whether these changes are happening, or the reasons for them. We would regard as very strange a statement such as <i>'as a Christian [substitute Buddhist, Humanist etc] I believe that the average world temperature will increase by 2-5.4% over the coming century'</i> but we would accept a statement such as <i>'the scientific evidence is that there is a significant possibility of climate change, and as a Christian I believe that we must act because of the consequences'</i>.

---

&gt; This is my first post ever to a CPDN-related forum. I have been silent lurker
&gt; though a heavy donator to CPDN.

Welcome along, and thanks for your contribution.

&gt; THESE are the questions that motivate me to contribute to CPDN. Is the grid
&gt; size in CPDN good enough to answer my questions with any serious certainty?
&gt; No. Furthermore, grid size notwithstanding, the model parameters probably do
&gt; an inadequate job parameterizing phenomena which should really be part of the
&gt; computations: fluid flow, heat transfer, precipitation, etc. But it all must
&gt; start somewhere, no? This is a perturbation analysis. It doesn't exist to
&gt; predict results; it exists to validate models. And those models will serve as
&gt; the shoulders for somebody who can shrink the grid, and refine, refine,
&gt; refine.

Quite.

&gt; What's so wrong with saying "I believe?" It's human. It's natural. It's NOT
&gt; unscientific. Faith and science can coexist. If they couldn't, why would we
&gt; bother with this?

I'd actually go further than that. Faith is inherent in belief about the world. Kant pointed out that you can make statements that are absolutely true, with no messy issues about the world, but that these "analytic" truths are tautologies, the truth of which is built into their definitions and the logical operations between them (classical logic and maths (arguably) are in this bin). On the other hand, as soon as you actually want to claim something about the world, you are left with various assumptions, the validity of which one can never ascertain. This, much of Hume, Kuhn and the history of science fatally undermines (at least in my mind) the notion of scientific certainty. We are defining, guessing, re-defining and re-guessing all the time, and at every point along the way we are believing some propositions and disbelieving others. Moreover, we are believing some propositions more than others, and adjusting things up and down as we go.

We always explain the world to the extent that we can, and it is never the "full world" (a point made nicely by Nancy Cartwright in The Dappled World). In the case of climate research this is kind of obvious (see title of this thread) but it's true of other science, too: we get our ideas about stellar evolution from a snapshot of different stars, rather than the sort of longitudinal study one probably requires to make literal claims about how stars change (akin to looking at a photo of a farm to work out the life-cycle of sheep, rather than actually watching a sheep be born, grow, grow old and die). We ofen get our ideas about dinosaurs, evolution, plate tectonics, etc in ways that don't correspond to the longitudinal claims we are making, too. To me, this isn't too big a deal as long as we remember that these claims (like all claims about the world and its operating system(s)) are contingent on the various assumptions we have made along the way.

Dave

---

&gt; This is my first post ever to a CPDN-related forum. I have been silent lurker
&gt; though a heavy donator to CPDN.......

&gt; I have been silent because I will find myself at odds with nearly everybody on
&gt; these boards. Perhaps "odds" isn't enough to describe the difference.
&gt; Loggerheads may be a better term, at least with regard to politics, religion,
&gt; and climate. I choose not to say anything on these forums because for the
&gt; most part, I think that there is no tolerance for dissent. Indeed, my
&gt; observations of board contributors lead me to conclude that there are a good
&gt; number of people for whom ecology is religion, global warming dogma, and man's
&gt; causality thereof immorality. These are the same people who would dismiss my
&gt; religion and politics because they require aspects of humanity that transcend
&gt; science. Science triumphant? No way. What an inhuman view of the world. If
&gt; sciece is the pinnacle of human existence, I for one perceive such existence
&gt; to be hopeless.


Welcome and thank you for sharing your views. The CPDN forums are no different to any 'community' of disparate individuals. Some are more vociferous than others, some choose to be open and candid about their political and ethical beliefs whilst some prefer to keep their opinions to themselves. It can be difficult to join in an ongoing debate where it seems so many regular contributors know each other well. We must each make our own choice whether to speak or not but if the debate on occasions seems one-sided then it is surely incumbent on those with opposing views to put their case. As a 'silent lurker' you will be aware that no matter how strongly any of us feel on the issues to hand, no-one uses the CPDN boards as a platform for <i>personal</i> attacks on anyone else's beliefs. Given the importance of the issues and the strength of feeling involved (from whatever viewpoint) it is remarkable that so few 'flame wars' occur. New contributors are always welcome, and for those who may have reservations, I would only ask them to consider the measured and thoughtful responses which your post has received.

Marj

---

_________________________________