We discuss climate models a lot, and from the comments here and in other forums it’s clear that there remains a great deal of confusion about what climate models do and how their results should be interpreted. This post is designed to be a FAQ for climate model questions – of which a few are already given. If you have comments or other questions, ask them as concisely as possible in the comment section and if they are of enough interest, we’ll add them to the post so that we can have a resource for future discussions. (We would ask that you please focus on real questions that have real answers and, as always, avoid rhetorical excesses).
Part II is here.
Quick definitions:
- GCM – General Circulation Model (sometimes Global Climate Model) which includes the physics of the atmosphere and often the ocean, sea ice and land surface as well.
- Simulation – a single experiment with a GCM
- Initial Condition Ensemble – a set of simulations using a single GCM but with slight perturbations in the initial conditions. This is an attempt to average over chaotic behaviour in the weather.
- Multi-model Ensemble – a set of simulations from multiple models. Surprisingly, an average over these simulations gives a better match to climatological observations than any single model.
- Model weather – the path that any individual simulation will take has very different individual storms and wave patterns than any other simulation. The model weather is the part of the solution (usually high frequency and small scale) that is uncorrelated with another simulation in the same ensemble.
- Model climate – the part of the simulation that is robust and is the same in different ensemble members (usually these are long-term averages, statistics, and relationships between variables).
- Forcings – anything that is imposed from the outside that causes a model’s climate to change.
- Feedbacks – changes in the model that occur in response to the initial forcing that end up adding to (for positive feedbacks) or damping (negative feedbacks) the initial response. Classic examples are the amplifying ice-albedo feedback, or the damping long-wave radiative feedback.
Questions:
- What is the difference between a physics-based model and a statistical model?
Models in statistics or in many colloquial uses of the term often imply a simple relationship that is fitted to some observations. A linear regression line through a change of temperature with time, or a sinusoidal fit to the seasonal cycle for instance. More complicated fits are also possible (neural nets for instance). These statistical models are very efficient at encapsulating existing information concisely and as long as things don’t change much, they can provide reasonable predictions of future behaviour. However, they aren’t much good for predictions if you know the underlying system is changing in ways that might possibly affect how your original variables will interact.
Physics-based models on the other hand, try to capture the real physical cause of any relationship, which hopefully are understood at a deeper level. Since those fundamentals are not likely to change in the future, the anticipation of a successful prediction is higher. A classic example is Newton’s Law of motion, F=ma, which can be used in multiple contexts to give highly accurate results completely independently of the data Newton himself had on hand.
Climate models are fundamentally physics-based, but some of the small scale physics is only known empirically (for instance, the increase of evaporation as the wind increases). Thus statistical fits to the observed data are included in the climate model formulation, but these are only used for process-level parameterisations, not for trends in time.
- Are climate models just a fit to the trend in the global temperature data?
No. Much of the confusion concerning this point comes from a misunderstanding stemming from the point above. Model development actually does not use the trend data in tuning (see below). Instead, modellers work to improve the climatology of the model (the fit to the average conditions), and it’s intrinsic variability (such as the frequency and amplitude of tropical variability). The resulting model is pretty much used ‘as is’ in hindcast experiments for the 20th Century.
- Why are there ‘wiggles’ in the output?
GCMs perform calculations with timesteps of about 20 to 30 minutes so that they can capture the daily cycle and the progression of weather systems. As with weather forecasting models, the weather in a climate model is chaotic. Starting from a very similar (but not identical) state, a different simulation will ensue – with different weather, different storms, different wind patterns – i.e different wiggles. In control simulations, there are wiggles at almost all timescales – daily, monthly, yearly, decadally and longer – and modellers need to test very carefully how much of any change that happens because of a change in forcing is really associated with that forcing and how much might simply be due to the internal wiggles.
- What is robust in a climate projection and how can I tell?
Since every wiggle is not necessarily significant, modellers need to assess how robust particular model results are. They do this by seeing whether the same result is seen in other simulations, with other models, whether it makes physical sense and whether there is some evidence of similar things in the observational or paleo record. If that result is seen in multiple models and multiple simulations, it is likely to be a robust consequence of the underlying assumptions, or in other words, it probably isn’t due to any of the relatively arbitrary choices that mark the differences between different models. If the magnitude of the effect makes theoretical sense independent of these kinds of model, then that adds to it’s credibility, and if in fact this effect matches what is seen in observations, then that adds more. Robust results are therefore those that quantitatively match in all three domains. Examples are the warming of planet as a function of increasing greenhouse gases, or the change in water vapour with temperature. All models show basically the same behaviour that is in line with basic theory and observations. Examples of non-robust results are the changes in El Niño as a result of climate forcings, or the impact on hurricanes. In both of these cases, models produce very disparate results, the theory is not yet fully developed and observations are ambiguous.
- How have models changed over the years?
Initially (ca. 1975), GCMs were based purely on atmospheric processes – the winds, radiation, and with simplified clouds. By the mid-1980s, there were simple treatments of the upper ocean and sea ice, and clouds parameterisations started to get slightly more sophisticated. In the 1990s, fully coupled ocean-atmosphere models started to become available. This is when the first Coupled Model Intercomparison Project (CMIP) was started. This has subsequently seen two further iterations, the latest (CMIP3) being the database used in support of much of the model work in the IPCC AR4. Over that time, model simulations have become demonstrably more realistic (Reichler and Kim, 2008) as resolution has increased and parameterisations have become more sophisticated. Nowadays, models also include dynamic sea ice, aerosols and atmospheric chemistry modules. Issues like excessive ‘climate drift’ (the tendency for a coupled model to move away from the a state resembling the actual climate) which were problematic in the early days are now much minimised.
- What is tuning?
We are still a long way from being able to simulate the climate with a true first principles calculation. While many basic aspects of physics can be included (conservation of mass, energy etc.), many need to be approximated for reasons of efficiency or resolutions (i.e. the equations of motion need estimates of sub-gridscale turbulent effects, radiative transfer codes approximate the line-by-line calculations using band averaging), and still others are only known empirically (the formula for how fast clouds turn to rain for instance). With these approximations and empirical formulae, there is often a tunable parameter or two that can be varied in order to improve the match to whatever observations exist. Adjusting these values is described as tuning and falls into two categories. First, there is the tuning in a single formula in order for that formula to best match the observed values of that specific relationship. This happens most frequently when new parameterisations are being developed.
Secondly, there are tuning parameters that control aspects of the emergent system. Gravity wave drag parameters are not very constrained by data, and so are often tuned to improve the climatology of stratospheric zonal winds. The threshold relative humidity for making clouds is tuned often to get the most realistic cloud cover and global albedo. Surprisingly, there are very few of these (maybe a half dozen) that are used in adjusting the models to match the data. It is important to note that these exercises are done with the mean climate (including the seasonal cycle and some internal variability) – and once set they are kept fixed for any perturbation experiment.
- How are models evaluated?
The amount of data that is available for model evaluation is vast, but falls into a few clear categories. First, there is the climatological average (maybe for each month or season) of key observed fields like temperature, rainfall, winds and clouds. This is the zeroth order comparison to see whether the model is getting the basics reasonably correct. Next comes the variability in these basic fields – does the model have a realistic North Atlantic Oscillation, or ENSO, or MJO. These are harder to match (and indeed many models do not yet have realistic El Niños). More subtle are comparisons of relationships in the model and in the real world. This is useful for short data records (such as those retrieves by satellite) where there is a lot of weather noise one wouldn’t expect the model to capture. In those cases, looking at the relationship between temperatures and humidity, or cloudiness and aerosols can give insight into whether the model processes are realistic or not.
Then there are the tests of climate changes themselves: how does a model respond to the addition of aerosols in the stratosphere such as was seen in the Mt Pinatubo ‘natural experiment’? How does it respond over the whole of the 20th Century, or at the Maunder Minimum, or the mid-Holocene or the Last Glacial Maximum? In each case, there is usually sufficient data available to evaluate how well the model is doing.
- Are the models complete? That is, do they contain all the processes we know about?
No. While models contain a lot of physics, they don’t contain many small-scale processes that more specialised groups (of atmospheric chemists, or coastal oceanographers for instance) might worry about a lot. Mostly this is a question of scale (model grid boxes are too large for the details to be resolved), but sometimes it’s a matter of being uncertain how to include it (for instance, the impact of ocean eddies on tracers).
Additionally, many important bio-physical-chemical cycles (for the carbon fluxes, aerosols, ozone) are only just starting to be incorporated. Ice sheet and vegetation components are very much still under development.
- Do models have global warming built in?
No. If left to run on their own, the models will oscillate around a long-term mean that is the same regardless of what the initial conditions were. Given different drivers, volcanoes or CO2 say, they will warm or cool as a function of the basic physics of aerosols or the greenhouse effect.
- How do I write a paper that proves that models are wrong?
Much more simply than you might think since, of course, all models are indeed wrong (though some are useful – George Box). Showing a mismatch between the real world and the observational data is made much easier if you recall the signal-to-noise issue we mentioned above. As you go to smaller spatial and shorter temporal scales the amount of internal variability increases markedly and so the number of diagnostics which will be different to the expected values from the models will increase (in both directions of course). So pick a variable, restrict your analysis to a small part of the planet, and calculate some statistic over a short period of time and you’re done. If the models match through some fluke, make the space smaller, and use a shorter time period and eventually they won’t. Even if models get much better than they are now, this will always work – call it the RealClimate theory of persistence. Now, appropriate statistics can be used to see whether these mismatches are significant and not just the result of chance or cherry-picking, but a surprising number of papers don’t bother to check such things correctly. Getting people outside the, shall we say, more ‘excitable’ parts of the blogosphere to pay any attention is, unfortunately, a lot harder.
- Can GCMs predict the temperature and precipitation for my home?
No. There are often large variation in the temperature and precipitation statistics over short distances because the local climatic characteristics are affected by the local geography. The GCMs are designed to describe the most important large-scale features of the climate, such as the energy flow, the circulation, and the temperature in a grid-box volume (through physical laws of thermodynamics, the dynamics, and the ideal gas laws). A typical grid-box may have a horizontal area of ~100×100 km2, but the size has tended to reduce over the years as computers have increased in speed. The shape of the landscape (the details of mountains, coastline etc.) used in the models reflect the spatial resolution, hence the model will not have sufficient detail to describe local climate variation associated with local geographical features (e.g. mountains, valleys, lakes, etc.). However, it is possible to use a GCM to derive some information about the local climate through downscaling, as it is affected by both the local geography (a more or less given constant) as well as the large-scale atmospheric conditions. The results derived through downscaling can then be compared with local climate variables, and can be used for further (and more severe) assessments of the combination model-downscaling technique. This is however still an experimental technique.
- Can I use a climate model myself?
Yes! There is a project called EdGCM which has a nice interface and works with Windows and lets you try out a large number of tests. ClimatePrediction.Net has a climate model that runs as a screensaver in a coordinated set of simulations. GISS ModelE is available as a download for Unix-based machines and can be run on a normal desktop. NCAR CCSM is the US community model and is well-documented and freely available.
Do GCMs model the 1.4C difference in annual temperature between NH and SH ?
If they do, what prinipally causes this phenomenon ?
If they don’t…… why not ?
Well, Alexi, you should have said.
You did not say that.
So please say what in mathematical terms you wish to know so it can be answered in mathematical terms.
We can’t read your mind. Only your english.
So try again.
What mathematically are you looking for?
NOTE: be accurate. For some versions of what you’re looking for, this
http://www.qsl.net/kb7qhc/reference/bode/bode.html
is correct.
Lawrence, look it up. From a quick google:
CO2 per can of soda 6g CO2.
22 cans a week is a lot, so say 4 cans average.
4Billion people.
4×10^9x4x50x6g = 5×10^12g = 5×10^9kg = 5x106T = 5Million tons.
I’m not trying to be argumentative, but a response might be interesting. I read a comment by a skeptic that because climate model projections are “averaged over time” (ie. “linear” in nature), they don’t show possible future tipping points (chaotic tipping point reactions like the climate system actually likes to operate in).
As a result, this might lull the mainstream climate communtity into a false sense of complacency as they look at climate model’s future “linear” reponses. How big an effect do you feel that this is having on mainsteam climate science’s “global warming” projections?
Thanks
[Response: It makes no sense. Models are full of non-linearities and aren’t ‘made linear’ by averaging in time. There may be physics that isn’t included in the models that might lead to dramatic changes (c.f. the ozone hole physics that were not included in the first models of ozone depletion). But that has little to do with how non-linear the models are. – gavin]
Feedback:
rate of accretion of mass in a diffuse nebula.
4/3 pi r^3 dt
This is the volume of possible capture for a mass in space.
dt is the delta time.
But masses can move during that time only s=ut+1/2at^2.
And a is the gravitational accelleration toward that mass.
So you can do the maths yourself, but the rate of accretion depends on the mass of that object.
But dt seconds later, the mass has gone up, so “r” is bigger and so the next dt seconds later, the growth of the mass is higher than it was one timestep ago.
This is called FEEDBACK. Notice how feedback didn’t need to be mentioned to introduce the mathematics (or, indeed, the model of speed of accretion)?
Now, what don’t you understand, Alexi?
You have the start of one feedback equation. You have already been told one feedback mechanism that is relevant in climate models. Are you unhappy because it doesn’t *mention* feedback in the model description you have hold of? Well, get hold of the source code for a model with feedback.
Within that source code you will see the maths.
If you know your maths, you will see how the code sets up an iterative process. This iterative process doesn’t have “feedback” in the name because the feedback is a RESULT of running the model, not the aim of the model itself.
Similarly with the other things “you don’t know”.
Sigh. Without a real mechanism for putting in equations, you miss spotting.
r^3. Should be the size of the shell around the mass considered, so you need to differentiate it.
So sue me, I’m an astrophysicist. We don’t model how stars form, just how the grow and die.
Cumfy, Land.
Mostly.
The NH has lots of land, not a lot of water. SH has not a lot of land which leaves a lot of room for water.
Land warms up to a shallow depth (ask a gardener, no need to trust a scientist). The ocean warms to a much greater depth.
That’s one element. A real climatologist will tell you more.
Well, Mark, I am sure I already said enough even in my first post (#135). I mentioned “boundary conditions”, “differential equations”, “parameters”, “state variables of a model”, “finite-difference (or spectral) approximation of NSE”. It seems to me that even a student in EE could understand that I am not asking about army cooks or how to feed a squaddie. I was even more explicit when referring you to a pdf in post #150. The pdf contains very accurate formulas. Therefore, I suggest that the “try again” is on your side. Thanks.
Ray Ladbury (138) wrote “My sources had said that the alcohol boom had brought employment if not prosperity to the northeast.” That is my understanding as well. But the majority of the ethanol employment appears to be located further west and south.
Read Under a Green Sky, and 600mya we had anoxic ocean with a thin skim of oxygenated water, with sulfur bacteria living the high life. Got that. His point being that that’s the way it was for all time, until 600 mya. And maybe the bacteria have gotten the high hand a couple of times since, don’t know.
It appears from his chart of co2 that aside from that huge spike 250 mya that Gaia seems to be reducing the point that she will tolerate the co2. Isn’t that just like a woman; wants one thing and then something else!!
Our new dna acquisition: a gene for microencephalophy back 50,000 ya. Just coincidentally after Toba. And speaking of Toba, could the particulates have held onto water to such an extent that the enveloping atmosphere could hold onto the water and not rain it out?
Creating a mist so heavy that we had an all-enveloping cloud layer? Would make a nice explanation as to why some —here we go, off into No-No land—as to why at one time there was no Moon? Maybe because we couldn’t see it? And then—whatever it was that broke the camel’s back, and we got the Flood.
And now, shockingly, some scientists say that the water is draining back into Gaia. Is that naughty girl planning some more deviltry? Dry it out, wait around for a space spack to send it back out?
Well, Mark these posts of yours in this thread have been a delight to read, regardless of our philosophical differences.
#2 “What are the major differences between climate models and weather models? Strengths and limitations of both?”
#2 Gavin wrote: [Response: Conceptually they are very similar, but in practice they are used very differently…. Weather models develop in ways that improve the short term predictions – but aren’t necessarily optimal for long term statistics. – gavin]
So Gavin, you would say that the “chaos effect” effects (uncertainty principle), is not modeled significantly differently between a climate model versus a weather model?
Don’t weather models play to the chaos effect (the starting location of each molecule) and climate models have the chaos effect dampened out so that the climate models can follow each other for years because the chaos effect has been stongly diminished?
Ie. If you have fifteen weather models, they usually diverge significantly after a few days (hours) while climate models don’t?
Thanks
[Response: No. They will diverge at very similar rates if run at similar resolutions. The initial condition divergence is a function of the non-linearity in the dynamics (‘weather’) – it has very little to do with differences in the physics or the climatology. Climate models have exactly the same chaotic behaviour as weather models because they are essentially doing the same kinds of calculations. – gavin]
Alexi, the thread here is asking for suggestions for FAQs — I’m not trying to answer your questions here.
Wrong topic, it’s not meant for discussion; plenty of that in prior threads.
Steven Mosher, 140
Thanks. Hank also might like to read this as well,
http://www.ecmwf.int/newsevents/meetings/workshops/2008/ModellingSummit/presentations/Stevens.pdf
Re Mark Smith @105: “we need to be extremely wary of the law of unintended consequences”
Absolutely, and the rush to corn-based ethanol is a very good example, but then those who promoted and decided to proceed headlong in that direction were not really doing so to combat rising atmospheric CO2 levels.
We also need to keep in mind that we know what the world was like with lower CO2 levels, while human civilization has never had to cope with the higher levels we will see in the near future. Yet we do know those levels will create a great many unintended consequences, many of them potentially quite severe indeed.
Capthcha: waste wholly
Re Rod B @124: “Or did someone tell you that none of this stuff is any help to the economy and productivity?”
It doesn’t when more and more of your tax revenue goes to paying for “this stuff,” yet you still end up borrowing ever more to cover the rest of the bill, and then don’t make any headway paying down that debt.
I seem to recall a certain European country that used military spending to reduce unemployment essentially to zero, but it didn’t work out very well for them in the end.
The edgcm site linked in the initial post by the contributors, describes, in the Setting Up and Analyzing Climate Simulation section,
how vegetarian boundary conditions can be modified,usimg what they call a “fractional” scheme in which each grid cellis assigned a fraction of each vegetation type(such as desert,rainforest,Woodland, grassland, ………), so that every grid cell in
the GCM consists of multiple vegetarian types. This partly answers my question above, as to how to treat entities which appear within only a part of a grid cell.
http://forums.edgcm.columbia.edu/faq.php?faq=setup
It describes the flexibility and limitations of this parameter, which can be used in the GCM, as well.
“Initial Condition Ensemble – a set of simulations using a single GCM but with slight perturbations in the initial conditions. This is an attempt to average over chaotic behaviour in the weather.”
Gavin, I’m confused. So weather simulations don’t have this feature? Is this a big difference between weather and climate simulations? Can you explain this in much more depth?
“This is an attempt to average over chaotic behaviour in the weather.”
… and this is different from weather simulations in what ways?
So now you can say that climate models have this “averaging over chaotic behaviour” and weather models don’t? This is not clear to me.
Thanks
[Response: Weather forecasts use an initial condition ensemble as well. And the forecast is some kind of weighted average of the individual simulations. So there is no big difference in practice. The difference lies in what they are used for – weather models want to know the most likely trajectory of the individual chaotic path over the next few days starting from today’s conditions, while climate models are used to get the long term average of the individual paths over long enough periods so that the initial conditions don’t matter anymore. – gavin]
Suggestions for the list of FAQs:
These are not questions about the models per se but still about what the models are trying to forecast.
1) Which part of the planet does “global warming” apply to – ie which parts of the atmosphere (how high), oceans (how deep) and land surface (how deep)? (A subsidiary question would be how the components are weighted, eg does a 1 degree increase in one cubic metre of dry air increase the average by less than a similar rise in the temperature of 1 cubic metre of moist air or ocean?)
2) What are the main sources for estimates of global temperature and why do they differ?
Thanks again for the fine site.
Richard Ordway,
perhaps it’s useful to think that climate models are used to get an idea of the statistics of long-term weather conditions, but the weather itself remains chaotic and will never be predictable beyond a week or so. We can’t predict a tornado next year, but we do know when/where tornado season is.
Gary:
1. All of it. :-)
No, really. Just because the tropopause is getting colder doesn’t mean it is not being affected by global warming. If heat is being retained it isn’t heating up the tropopause.
2. Science. The main sources don’t.
Alexi 158.
Diffeerential equations.
The document you linked to was stuffed to the gills with differential equations.
So it wasn’t “missing”.
So, just on that one element, what is it you don’t understand.
The radiative forcing for CFCs is given as a linear relation for low concentrations see
http://www.esrl.noaa.gov/gmd/aggi/
Does this hold for the much (ten times or more) larger concentrations we would expect without the Montreal Protocol?
What is an approximate formula for concentraions large enough that the linear relation is no good approximation?
Which influence has the oxygen on climate? The concentration in the last 400 million years has probably changed between 15 and 30 percent, changing total atmospheric pressure I assume.
Inceasing oxygen (for example form 20.9 to 30 percent) in my opinion would:
Increasing scattering and so albedo, lowering temperature
Increasing total pressure, broadening of absorption lines, rising temperature
Increasing adiabate slope, rising SAT directly but decreasing water vapor, total effect??
Probably increasing ozone, rising temperature?
Probably decreasing methane, lowering temperature
Change on clouds, direction of effect?
Other effects?
What direction is the total effect and with size has the effects mentioned above?
Alexi, a physics based model won’t have a bit labeled “Feedback_from_water” on it because that is what a statistical model would do. The feedbacks are emergent properties of the physics included.
One reason why you won’t see terms like “feedback” in a physics based model.
A cheaper statistical model may. Then again, they may not CALL it feedback because, for example, H2O is known to be a feedback mechanism so it would be redundant to call it such.
But still explain what you’re missing about “differential equation”.
Re:153 Thanks Mark, I always underestimate google for that sort of stuff. 260 Million tonnes/yr..or the equivalent of planting 260 million trees. That’s just from soda drinks, then you’ve got the copious amounts of CO2 released from the manufacture and consumption of beer and spirits, aerosol cans etc. Why hasn’t anyone tried to raise that with the media or respective governments. Why cant you find another gas apart from CO2 to make drinks fizzy? Ok! cant be done with beer I know but surely it’s possible with soda..why not use compressed nitrogen or helium..pretty harmless inert gas?
Cumfy:>Do GCMs model the 1.4C difference in annual temperature between NH and SH ?
Mark:>Cumfy, Land.Mostly.
The NH has lots of land, not a lot of water. SH has not a lot of land which leaves a lot of room for water.
My principal concern is whether GCMs actually model that difference.
I was thinking the difference may primarily be due to the difference in polar albedo NH vs SH.
Uli writes:
Oxygen blocks a little of the ultraviolet light in the stratosphere, at shorter wavelengths than ozone. The amount blocked from reaching the surface would change slightly if the oxygen concentration changed.
> changing total atmospheric pressure I assume
Why? Is there any information about atmospheric pressure over the long term to go with information about atmospheric chemistry? I found a couple of tidbits:
Low atmospheric CO2 levels during the Permo- Carboniferous …
… at a time when total atmospheric pressure was similar or slight higher than now. …
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=130500
Pterosaurs couldn’t soar, says expert Oct 1, 2008 … A Japanese researcher has put paleo-biologists in a flap … OR the total atmospheric pressure was MUCH higher. possibly BOTH. …
http://www.physorg.com/news142086647.html
__________________
Note, I’m offering this as a suggestion for the FAQ collection.
I’m not soliciting amateur opinions or digression into discussion.
Raypierre probably knows the answer, I’ll hunt a bit more in his info.
Re: Carbonated drinks
It seems to me that the appropriate question is: where does the carbon in the CO2 in carbonated drinks come from?
If it’s from the biosphere, then the biosphere got it from the atmosphere in the first place, so it doesn’t represent an alteration to the carbon cycle which would increase atmospheric CO2 concentration longterm.
The problem with CO2 buildup is that we’re taking carbon from long-sequestered sources (fossil fuels) and injecting it into the atmosphere is a sudden burst.
I am still stuck with the aerosol influence on global temperature (posts 27, 52, 109 and 148). It apparently has had huge compensating effects by negating any rise from CO2 accumulation in the 1940’s to around the 1970’s. We should be able to trace regional effects on a macro and micro basis, showing substantial cooling in the conical wake behind polluting industrial centers – aside from the follow-on wide spread. Besides my earlier query about the recent Chinese rise in aerosol emissions (without apparent local cooling effects) there should be very good traceability of East European changes in emissions. Before the 1989 political and economical changes in East Germany, Poland and Czechoslovakia they had been the foremost industrial suppliers to the Soviet Union, with commensurate horrible pollution.
We have had detailed infrared satellite survey of the zones behind the iron curtain, i.e. there should be mapping records of before and after 1989. Did discernable warming occur?
Furthermore, if aerosols did have such a dramatic cancelling effect at the onset of WWII and during the following decades, is aerosol cooling part of the temperature models? They should, shouldn’t they? If so, do they confirm the above mentioned changes in Eastern Europe and China?
… “we’ll say for the record that the late Michael Crichton did a disservice with his denialist potboiler novel State of Fear, which abused climate scientists and environmentalists. President Bush met directly with Crichton while snubbing real scientists. President Obama can begin to set things right by showing that he is instead meeting directly with leading scientists and learning from them.”
http://www.climatesciencewatch.org/index.php/csw/details/michael_crichton_disinformation_legacy/
Tamino, #179. I answered that in the spirit of showing how MUCH CO2 humans can produce when something as minor as your soft drink can, when multiplied by the number of people on the planet, really DOES add up to a LOT.
Cumfy, #176. Yes, the models do tell the difference between sea and ice. For one thing, there’s no orography over the ocean and there’s very little water over the land. Ignoring the differences would be silly.
Deitrich, #180. Well the amount of aerosols produced were enough to kill large numbers of people annually. That’s how much they affected the atmosphere: they killed people breathing it.
If something is able to do that, why the disbelief that it is enough to affect the changes seen?
RE: #175: CO2 from soda: I think the factor of 52 (well, 50) weeks was already in Mark’s first calculation, so I think 5 million tons was a yearly value already. 260 million tons would be 1% of the yearly ~26 gigatons of CO2 we emit, which would be kind of shocking. 0.02% makes more sense.
Re: #179: http://www.asbcnet.org/resources/CO2.html seems to claim that most CO2 used in the beverage industry is a byproduct of the ammonia production process: therefore, even though this CO2 does come from a fossil source, one might argue that it would otherwise be released into the air and therefore is “free” CO2. Now, once there is a _price_ on carbon, soda companies will have to compete with selling credits based on injected CO2 underground.
Well, Mark, you ask: what am I missing in “differential equations”? You already have answered that question: “feedbacks” are missing. There is no “feedbacks” in the original, practical formulation of climate models, thank you for spelling this out. Then why would all climatologists and their skeptics break so many fences over the thing that does not exist? Why would the populist sites like RC spend so much time arguing about the non-existing thing and trying to put “quick definitions” for it? As result of being non-existent, it cannot be properly quantified, measured, and therefore causes nothing but illusion of understanding, which is quite well demonstrated here. Thanks.
Alexi, #185. Well “feedbacks” have nothing to do with differential equations.
You don’t get “feedback” in electrical amplification differential equations, either. Feedback is what you get when the output leads to the input. It’s not an add-on, it’s the result of the thing itself.
You don’t model “feedback” your model shows feedback. Otherwise you’re running a statistical model that says something along the lines of “If CO2 plus feedbacks meant an increase in insolation retention equal to 4W/m2, what would the system do”. Which is a statistical model, not a physical one. A physical one has “the water has an effect on temperature of the air”. It also has “the temperature has an effect on the water content of the air”. It doesn’t bother with “water has a feedback effect on anything that increases ore reduces air temperatures” because that isn’t needed: it’s an inherent property of the physics you’ve included in your model.
So that’s where “feedback” disappears: it isn’t needed in a physics based model because such properties are emergent, not built in.
Might as well ask “where does the dark come from when you turn the lights off?”.
Dietrich Hoecht (180) — I suspect that the East Asian aerosols are having an effect on temperatures and precipitation in North America. I breifly looked at a paper explaining how these aerosols generally move northwest up the coast and then on the westerlies across the Pacific. Locally, the weather and sky have been most unusual for about two years; these aerosols might well explain the effects I am noticing.
Who writes code for the models used by professional climate scientists? Do they write requirements for a model to accomplish x,y,z and then turn it over to a programmer to implement? I ask because of the comments about models running on the order of months. Over the years I’ve seen some impressively bad code written by smart people attempting to implement their ideas but with no formal training. Often such code \works\ but is terribly inefficient. Example, years ago I was tasked with updating an application which normally ran for 12-15 hours on a data set. When I was done it ran in 20 seconds. Same machine, same data, different person writing the code. The guy who wrote the original app was stunned, then admitted he had no formal training as a programmer. Just a thought.
[Response: Unfortunately, it is mostly by scientists (at least in the first instance). As you suggest this is not necessarily optimal, though we do get comp. sci. support. Most efficiency gains at the moment are coming from massively-parallel programming and that is unearthing a number of performance bottlenecks, however, the models do a large number of very separate things, and even if one of those was as badly coded as your example (unlikely, but possible), it would end up reducing the actual run time by only a couple of percent. But you are more than welcome to profile our code and see if you can spot something… – gavin]
Re: #188: My understanding is that while some climate code is fairly ugly (and much of it is written in Fortran which some people think is inherently ugly), the fundamental reason for the time it takes to run a model is the sheer number of calculations.
You need to run mathematical calculations in every grid cell for every timestep. Every time a model improves its resolution by a factor of 2, it increases computation time by a factor of 16: 2x2x2 physical dimensions, plus you need to perform the calculations twice as many times in order to avoid instabilities (eg, you don’t want a packet of air moving west to move further than 1 grid cell at a time: so if you halve the size of the grid cell you’ll need to halve your timestep).
And as you add in more processes: chemistry, ecosystems, 3D oceans, etc. you also increase computational demand.
I know that some models attempt to have dynamic grid cell size, so they only have high resolution where it is needed – this is a nice computational trick that saves time, but is probably limited in applicability.
Worth noting the discussion of the great natural philosopher de Selby’s views on darkness here.
As in many other of de Selby’s concepts, it is difficult to get to grips with his process of reasoning or to refute his curious conclusions. The ‘volcanic eruptions’, which we may for convenience compare to the infra-visual activity of such substances as radium, take place usually in the ‘evening’ are stimulated by the smoke and industrial combustions of the ‘day’ and are intensified in certain places which may, for the want of a better term, be called ‘dark places’. One difficulty is precisely this question of terms. A ‘dark place’ is dark merely because it is a place where darkness ‘germinates’ and ‘evening’ is a time of twilight merely because the ‘day’ deteriorates owing to the stimulating effect of smuts on the volcanic processes. De Selby makes no attempt to explain why a ‘dark place’ such as a cellar need be dark and does not define the atmospheric, physical or mineral conditions which must prevail uniformly in all such places if the theory is to stand. The ‘only straw offered’, to use Bassett’s wry phrase, is the statement that ‘black air’ is highly combustible, enormous masses of it being instantly consumed by the smallest flame, even an electrical luminance isolated in a vacuum.
(The Third Policeman)
Mark (#186): What are you trying to say is that “feedbacks are loose interpretations of statistically post-processed outputs from various fluctuating physical variables generated by a model”, correct?
I also cannot understand why do you keep the distinction between “physical model” and “statistical model”, which maintains an impression that they are somehow separate and independent. From the initial definitions of this article it follows that there no such thing as a separate “statistical model” but statistically-processed outputs from actual state variables of the running “physical” model. One model, different details.
This kind of incoherence leads to statements like “If CO2 plus feedbacks meant an increase in insolation retention equal to 4W/m2, what would the system do”. Apparently the system would provide a “feedback”, and the feedback would define another level of “forcing”. Then, the “feedback” to your “CO2+feedbacks” would cause another “feedback”, and another “forcing”, etc. So, what are you trying to state with this fuzzy construction? Would the system settle to a new statistically steady state? Would it run away? Maybe it would oscillate, chaotically? I am just trying to figure out a value of this kind of speculations about undefined “feedbacks” and “forcings” in a spatially-distributed multi-variable system.
[Response: I will add an Q on this above at the weekend, but Mark’s statements are not incoherent. The 4 W/m2 TOA forcing is the consequence of an imposed change in CO2 – all changes to LW absorption in the atmosphere as a consequence of that initial change (through water vapour, cloud or temperature profile responses) are feedbacks. The total amount of LW atmospheric absorption at the eventual (statistical) steady state will be larger than the 4W/m2 forcing (probably about 10-20 W/m2), hence implying that the eventual temperature rise will be larger than the ‘no-feedback’ response. One could envisage situations such that the feedbacks were large enough to create a runaway condition, but that has very little to do with the current climate system. – gavin]
PS Alexi, how can anyone answer you?
You started off with: I don’t understand feedback
Then it was: I don’t understand differential equations
Now it’s: Wen I say “I don’t understand differential equations”, I mean “feedback in differential equations”.
I should like to take this opportunity to offer a small critique of those other models, the simple models.
Much criticism is given to the projections and of big (AOCGCM) models, at least some of which is based on arguments using simple, some might say simplistic, models.
As is indicated in the opening presentation, big models have to meet a large and diverse set of criteria and have to do so simultaneously. It seems that simple models are sometimes produced ad hoc to highlight a specific issue meeting only a few, commonly a minimal set of criteria.
For clarity I should like to classify small models as “climate without weather” models as opposed to the big “climate with (or from) weather models”.
Ultimately they both need to meet the same set of criteria and they both need to do so simultaneously. I think the requirement for simultaneity is by biggest problem with the use of simple models to argue specific points.
By way of an example:
The use of a slab ocean (50m of isothermal water) and a climatic observation to calculate the climatic sensitivity.
Very interesting, but if one wishes to model the lag of the seasons or the diurnal lag such a model does not work. A 50m slab would (in isolation) would predict much greater lags than are observed.
Personally I think simple (climate without weather) models are important and ultimately the way forward, but they must meet the all harsh criteria of the observed climate simultaneously or they are simply to simple for purpose.
Best Wishes
Alexander Harvey
No, alexi, feedbacks are emergent properties.
If rate of accretion depends on the mass of the accreting body, then the rate of the increase in mass of that accreting body increases quickly until all available material is taken. This is feedback. It emerges from the (non feedback) collusion between rate of accretion in a uniform medium and the results OF that accretion.
With the “Statistical” and “Physics based” model, read the article and the first ~100 comments. They are different in exactly the same way as numerical solution is different from symbolic solution in mathematics.
That statement describes what a ***statistical*** model would do. The model description you linked to was NOT a statistical model. It was a ***physics-based*** model. This would be why the description of that physics-based model did not include any mention of “feedback” because, and please tell me how else to say this, because you never seem to be able to read it, *** it is an emergent property of the physics, and NOT, repeat NOT, a separate element to model ***.
The incoherence here is your inability to say what you don’t understand. Any attempt to answer must include that incoherence within itself in any attempt to explain else there is no apparent link between the (incoherent) query and the answer given.
Mark,(183), aren’t you the least bit of intellectually curious if a phenomenon appears able to cancel out global warming for 30 years? There is definitely no linkage in bracketing something as atmospheric health hazard and the discussion of the physics of climate change. Is’n this science, with all of its theories, observations, tests and ongoing learning and improving results? I don’t get your kind of argument – it’s like you are afraid of reality.
Dear Mark, you should probably try again to answer my original post (#135). Instead you gave me a lecture on how to cook for army, and what is an eating pattern of a “squaddie”, and decay of nebula. Please take a note that I do perfectly understand what a feedback is (in the normal science and engineering), and how to write and analyze/solve differential equations, both ordinary and partial. I just wanted a clarification of buzzwords you are using. So far you haven’t done a good job as an apprentice at RC. Try harder.
Gavin: your example does not make it any clearer. One interpretation of what you said is: if 4W/m2 of “initial forcing” causes 20W/m2, then the resulting 20W/m2 would in turn cause 100W/m2 forcing, then 100W/m2 would cause 400W/m2, etc. The formal incoherence arises here because you omitted to discriminate one kind of “forcing” (from CO2) from other (derivative) forcings. In a multi-variable system “forcings” do not add up in a simple algebraic way. Apparently there must be additional qualifications to restore coherency in this loosely-defined “Forcing-Feedback” paradigm.
[Response: I disagree (obviously). The 20W/m2 extra LW absorption is not a forcing, it was how the planet has adjusted to the initial change – it is not an imbalance at the TOA. (Just because the units are the same doesn’t make a feedback into a forcing). I do distinguish them – one I call a forcing, the others a result of feedbacks. It’s really not that hard. – gavin]
The Global Climate System: Patterns, Processes, and Teleconnections by Howard Bridgman, John Oliver, and Michael Glantz is an excellent book that will answer most of your questions. Proxy data, ratonal behind the models, how they have improved and some issues on paleoclimate are all covered. If you do not have the money or the time to purchase it, it is available in many libraries. The introduction speaks about the heat budget, important equations and so forth. Oh and I checked, you can get limited access through Google Scholar, so now you can read and learn, then you can ask questions that have an answer. I found the book to be indispensable, and quite frankly it would benefit many of the posters in this thread as well. Also many states have city public library ebooks that one may sign up for and view for about 3 weeks right on the desktop, many technical books can be accessed this way.
OR (if the book is too advanced to begin with)
Alexi start with reading relevant articles from Britannica, pick up a general textbook on weather and climate, or physical geography, Richard Christopherson is good, and why don’t you go to NASA.gov, PBS.org, and physics.web, and make sure you have the appropriate math background. Even non climate modelers with a solid science and math background can get the gist of the models. Some of the references and older posts on this site would really help you understand.
OR take a few intro courses in meteorology, earth science and take DQ
http://books.google.com/books?id=bV3C5VCC-0EC&dq=global+climate+systems&pg=PP1&ots=XW6n2RX7mQ&source=bn&sig=_t2m0FPZAeSWdR4afPwoPm0dNEE&hl=en&sa=X&oi=book_result&resnum=4&ct=result
Cambridge University Press.
The six box climate models seem right on and even simpler models, distinguish between solar foercing and greenhouse gases.
Great article.
I think it would be great if you guys added a short section explaining the basic way in which clouds are included in the models. A good friend of mine (who’s a scientist in an area far removed from climate) was arguing some time ago that the climate models are useless because they don’t include clouds at all or in such a crude way that their effect is totally meaningless. Apparently, he got this idea from talking to another professor in his university, who seems to specialize in studying clouds.
After saying this stuff about clouds he went on to mention something similar to the infamous “iris effect” (as the planet warms, more clouds appear, thus reducing insolation and limiting the temperature rise).
I have seen some models that include clouds, so I’m sure my friend is wrong. But it would be great to have some more info about cloud simulation in your models, to be able to counteract these false ideas.
Thanks!
[Response: Good Q. Models do include clouds, and do include changes in clouds with climate. There are certainly questions about how realistic those clouds are and whether they have the right sensitivity – but all models do have them! In general, models suggest that they are a positive feedback – i.e. there is a relative increase in high clouds (which warm more than they cool) compared to low clouds (which cool more than they warm) – but this is quite variable among models and not very well constrained from data. – gavin]
Alexi,
forcing-feedbacks on Earth are a converging series, not a diverging one. It is not necessary for a positive feedback to result in runaway warming.