On December 15 2011 10:04 Capook wrote: The current models contain hundreds of fudge factors that must be fit to past data. They all do it in different ways, and their predictions diverge rapidly. Also, the predictions of a single model often changes wildly (well outside the statistical error bars of the previous model) when new physics is added.
To the OP: Are elements of climate models still adjusted "by hand", and is this in the initial conditions phase or the evolution phase? Also, the multi-model dataset used in the fourth IPCC assessment report seems to have a fairly well-constrained set of model results, at least in terms of mean temperature - were these results somehow sterilised to ensure that "rapidly divergent" results did not make it into the report?
Capook: do you know what 'diverge rapidly' means in the context you are using it in?
hi ipunatic, good question.
the question of overfitting / tuning / free parameters is a scientific grand challenge for climate models. it is simply not possible to test all reasonable paramater combinations due to the fact, that coupled ocean - atmosphere models have to be run thousands (!) of model years to achieve equilibrium, which is incredibly expensive on modern high performance computer.
so the answer is: yes, there are elements that are adjusted by hand, or by expert judgement. there is however a whole science evolving around the idea of a perturbed physics parameter ensemble (where I am also currently active), that tries to put a mathematical framework around the concept of tuning.
one thing you should keep in mind is the fact, that those perturbed physics ensembles usually have quite a similar behavior as ensembles of opportunity as the IPCC ensemble, which is essentially a mix of randomly produced climate models around the world.
if you do bayesian analysis with these type of ensembles, both ensembles will in the end lead to the same - as of now structural - uncertainty range in climate sensitivity.
these adjustments are btw done in something called "control run", a thousand year equilibrium run with forcing of 1880 or so, pre-industrial. all experiments later on (historical, projections,...) are then done with this set of parameters, so at least it is possible to evaluate this specific model version against observations in a weak sense.
On December 15 2011 22:16 optical630 wrote: whats the worse case scenario if the north pole melted forever
The absolute worst case scenario I heard of was, the gulf stream could stop sending warm water from the Caribbean to Europe, and Europe experiences some kind of Ice Age. Best case scenario is, nothing happens and new shipping routes from Europe to East Asia open up.
Would also depend on the quantity of frozen water in the north pole - water levels rising a few feet would could be pretty disastrous to coastline cities.
On December 15 2011 10:04 Capook wrote: How about this one, from a genuine "skeptic" (yours truly). I'm only "skeptical" about some aspects of the science--my main gripe is with the policies advocated by most climate change people.
It's clear that the earth is warming, and from basic physics it seems likely that co2 is the cause. But that's not too interesting for policy. What matters is how much the warming will continue (say, global mean temperature in 2050), and what the environmental consequences will be. To answer these questions requires simulating the entire earth, which is basically impossible. The current models contain hundreds of fudge factors that must be fit to past data. They all do it in different ways, and their predictions diverge rapidly. Also, the predictions of a single model often changes wildly (well outside the statistical error bars of the previous model) when new physics is added. This is because--as everybody knows--the overwhelming source of error is systematic, in that the included physics is not correctly modeled and much physics known to be relevant is not included. In no other field of science (certainly not mine) would the quantitative results of such simulations be taken seriously, and yet in climate science policy documents you see statistical error bars slapped on results of whatever the latest simulations are (with some averaging over different models), and pedaled as scientific predictions to world leaders.
Here's what I would want to see before quantitative simulation results can be considered reliable. For each model, train on only 1/2 the available data, and then verify that the remaining half is reproduced before making predictions. Verify that all secondary aspects are reproduced as well (such as measured ice melt rates, etc.). For each code, verify that each time new physics is introduced, the results remain consistent with previous results. And across all codes, make sure that agreement remains within statistical error bars for each code. I have seen two climate simulation presentations made for physicists, and none of them showed any of this. I have also directly asked some experts and none could produce such tests. I therefore assume that these criteria are (roundly) not satisfied. In that case the simulations can be regarded as providing further evidence for the hypothesis that co2 caused and will continue to cause warning, but can not be used to infer anything quantitative about the future.
The bottom line is we know next to nothing about what the scope and impact of global warming will be. This should be straightforwardly communicated to policymakers so they can make the best decisions. In my own personal opinion, it makes no sense to spend tons of money on combating co2 emissions based on the idea that (e.g.) droughts and floods *might* cause food shortages for people in 50 years, when these same people *are* facing food shortages and many other problems right now. If we spent a tenth of what is proposed to be spent to combat global warming on current problems in the third world, it would be truly a testament to the charity and generosity of western liberalism and environmentalism. On a final personal note, I count myself both a liberal and an environmentalist, but not a gambler. Since the impact of co2 emissions is so incredibly uncertain and there are so many certain problems now, let's spend our limited money, time, and energy fixing the certain ones.
hi capook,
awesome post. the methodology you suggest is not complex in itself and evident to all students of machine learning or related topics. I was baffled myself that it was not used consistently in climate research so far.
the reasons are clear: global circulation models did not start as prediction models but as tools to understand processes and to complement data. now, they are used to make projections of some type, and the focus is clearly on the more hard evaluation tests you suggest. I know that the section on model evaluation of the forthcoming IPCC report AR5 will be much more clear on this part, that so far the evaluation procedure lacks scientific rigour in some parts.
Still, your argument "we know next to nothing about what the scope and impact of global warming will be" is plain wrong. Even with very simple, one line energy budget models you can estimate global warming of certain magnitude. The radiative transfer models that can estimate warming for a single air column due to enhanced CO@ concentration are "parameter free" in the sense that they are just numerical approximation of well know physical equations. They do predict a warming of around 1 degree centrigrade per doubling of CO2. There is close to non euncertainty in this number.
As I mentioned earlier in this thread, the feedbacks that enhance this direct effect are the cause of uncertainty, i.e. clouds, water vapor, vegetation change, ice cover change, x.. This adds up to the current understanding of around 3-5 degrees warming per doubling of CO2, which is also consistent with DATA (20th century) and RECONSTRUCTIONS (paleo time periods).
Bottom line: while there is room to improve evaluation of climate models and it is currently done, see tendency to falsifiable decadal predictions, the result does not mean, we know nothing of climate change. It just indicates again, that the problem is complex.
On December 15 2011 22:16 optical630 wrote: whats the worse case scenario if the north pole melted forever
its not really dangerous. floating sea ice will not influence the sea level, only the salinity distribution of the ocean. therefore it MIGHT lead to a shut down of the meridional overturning circulation (which one big part of is the gulf stream) and to colder climate in Europe, but I am sure that we would also be able to cope with that.
The problem about ice is land ice, glaciers and especially greenland and antarctica.
A climate that would meld the Arctic permanently, would also melt greenland (few meters! sea level rise) and in the end Antarctica (dozens of meters of sea level rise). Both effects would be very slow, though, so its nothing that will abruptly happen this century,
On December 14 2011 20:29 Zerksys wrote: For the OP:
Regarding the comment about how higher temperatures mean higher crop yields, this is simply not true. For the vast majority of things that we grow other than the super hardy plans like sugar cane and corn, higher temperatures are bad. This is due to the fact that certain enzymes that run photosynthesis work only at very specific temperatures and rising temperatures would do nothing to help this. The higher carbon dioxide part is true, yes, but the toll taken by the enzyme function in plants is catastrophic at higher temperatures. You often find things like this in nature where living things have adapted to exactly the conditions of the planet at the time and changing the system even a little bit would do lots of damage.
hi zergsys, thanks
I agree 100%. Higher temperature most certainly will not increase net primary productivity of the Earth's ecosystem.
On December 15 2011 10:24 adacan wrote: 3 questions. Are there any climate scientists who dispute AGW?
How can people who have no background in the study of climate change dispute the scientific consensus?
Why doesn't this happen with other scientific topics? (i mean it does happen with evolution but to a much smaller extent)
Sorry if these questions seem loaded but I am genuinely curious.
hi adacan
from my experience:
a) no
b) its easy, write a blog, let the blog be cited by another blog and in the end its automatism. the internet is free. additionally, many scientific results are not sharp in their conclusions because of inherent uncertainty. this uncertainty is used by agents (in a game theory sense, not matrix ~) that have motivations to confuse public opinion on the state of climate change science.
in other words: the public debate that should be ongoing on the moral and ethical and practical way of adressing the climate change problem, is transported in the scientific domain by creating a fake view, that science is still divided. science is divided in specific questions, the main theory of AGW is proven beyond reasonable doubt at this point in time.
c) it does, in medicine concerning the efficacy of new drugs. and this already explains it: everywhere, where agents in the civil society, usually big companies or associations of a certain worldview, have a vested interest in a certain scientific result, they will try to interfere in the science domain, because it is much easier there to raise doubt - scientists are just prone to admitting uncertainties, its because our daily life exists of these uncertainties, we have to remain sceptic to survive in the science environment.
On December 13 2011 07:23 QuXn wrote: man made global warming probably exists, but it is not nessessary to take any direct action stopping/delaying it, as the finite resources of the earth make it impossible to cause any real harm to humankind. anyway, nobody cares about it because of financial trouble in the world.
another point: im AGAINST all kinds of subsudies to create ethanol or any other form of biofuel, because it literally KILLS people by the millions (rising food prices,...) and generates just 1.5 times the amount of energy used to produce it!
His presentation was really good, but when he said 70% of my portfolio is in gold and/or silver, it became very suspicious.
It then became even more suspicious as he started speculating how much gold/silver prices are going to increase, because of the exponential difficulty of getting those resources.
I wont deny that the climate changes, it has been doing so since the dawn of time as ice ages amongst other events show. What i do deny is that there is scientific proof that human activity plays a major role in this. In my personal opinnion the impact of humans on climate change is maybe 1% at best (we are not advanced enough, nor is our economic activity big enough to make a substantial difference) and the impact of nature (the solar activity, comets, solar wind etc) and natural cycles 99%. This does not mean we should not try eliminate that 1%, but i dont think it realy matters in "the grand sceme of things"
Maybe when earth has 100 billion people, all living on oil and with no advances in "clean" technology we would have a major impact.
I have two items I'd like discussed. A) This was already indirectly addressed in an earlier post but: I had a professor of physics specializing in electromagnetic phenomena and he was opposed to the idea of carbon dioxide having significant impact on the basis of the strength of water absorption. I did my semester paper on the window of wavelengths blocked by carbon dioxide and how this prevents heat from escaping, but he still disagreed claiming that the amount of water vapor in the atmosphere dwarfed co2 by so much that even in the ranges where water doesn't absorb it still was so prominent that the affects of co2 would be many orders of magnitude lower. I didn't do any follow up research because it was the end of the semester but this still bugs me today. That said B) I really, honestly don't think ANY models for predicting how climate change will effect mankind are reliable. I argue that we have enough evidence to say global warming influenced by man's actions is occurring, but the extraordinarily complex and chaotic nature of convection and climate cycles on earth result in an inability to reliably predict the outcome of this. This is compounded by trying to project THOSE results onto ecological factors. When we add yet another layer of unpredictability in the form of future human behavior, the uncertainty reaches a point where I no longer am willing to expend significant resources to counteract this trend. But I am a very lazy Nihilistic moral relativist so it's extraordinarily difficult to get me to act on anything.
On December 16 2011 09:14 Rassy wrote: I wont deny that the climate changes, it has been doing so since the dawn of time as ice ages amongst other events show. What i do deny is that there is scientific proof that human activity plays a major role in this. In my personal opinnion the impact of humans on climate change is maybe 1% at best (we are not advanced enough, nor is our economic activity big enough to make a substantial difference) and the impact of nature (the solar activity, comets, solar wind etc) and natural cycles 99%. This does not mean we should not try eliminate that 1%, but i dont think it realy matters in "the grand sceme of things"
Maybe when earth has 100 billion people, all living on oil and with no advances in "clean" technology we would have a major impact.
hi rassy,
your post is unfortunately very typical as a combination of 2nd and 3rd type of climate change denial
1st type: its not happening (took 3 decades, but nobody serious still does that type of arguing) 2nd type: its not us / we cannot do anything about it (majority today) 3rd type: its happening and its us, but it wont be too bad. (starts to grow in numbers and its not really denial anymore, it moves into the right direction of discussing policy issues in the policy realm and not the science realm)
to your arguments: yes, climate changes according to external changes, in the past orbital changes, volcanoes, vegetation changes, solar activity changes ++ today the biggest driver is man made co2 change, your argument does only strengthen this hypothesis.
in the same direction: your personal opinion estimate of 1%at best is simply untrue. there are studies that show that at least 3/4 of the warming trend in the 20th century is due to anthropogenic induced change in CO2 concentrations, probably closer to 90%.
and @ your last point: it is not sustainable to run Earth as of today with 7 billion people and half of them essential with a 100 year civilisation lag, it certainly wont be possible to sustain 100 at any point ~
and @ your last point: it is not sustainable to run Earth as of today with 7 billion people and half of them essential with a 100 year civilisation lag, it certainly wont be possible to sustain 100 at any point ~
Seriously? Do you understand technological innovation or economies of scale at all? People said in 1900 that the Earth could not sustain more human beings. Boy were they wrong.
and @ your last point: it is not sustainable to run Earth as of today with 7 billion people and half of them essential with a 100 year civilisation lag, it certainly wont be possible to sustain 100 at any point ~
Seriously? Do you understand technological innovation or economies of scale at all? People said in 1900 that the Earth could not sustain more human beings. Boy were they wrong.
Not sustainable by our current technology and living standards. Perhaps future innovations will improve things.
The 1900's thing was wrong because the guy thought that agricultural output could not increase exponentially, but it did thanks to the various advances in farming.
On December 15 2011 22:09 nam nam wrote: What about carbon dioxide sequestration? I know there have been a few prototype plants using this but I'm not sure how the results have turned out. I think it was quite cost ineffecient the last time I read about it.
Carbon dioxide sequestration is a technology to avoid dealing with other problems. It can be used at coal/oil/gas-fired powerplants, oil rigs and probably a few other very carbon-intensive endevours. In powerplants in Denmark you can reduce CO2 significantly through scrubbers, but scrubbers use a lot of sodium hydroxide to keep it down and after use it is just an unclean gypsum-mass ready to be deposited in closed down mines in Norway for indefinite storage. Using sequestration could therefore result in a benifit of less waste to manage. On oilrigs the last oils after a reservoir is almost emptied, it is normal to use salt water to pressurize the reservoir. Salt water is bad for that in many ways but the cheapest thing near an oilrig. Using CO2 to pressurize the reservoir is a far better method and will improve the amount recovered by 8-16%. It is most likely not enough to make it worth economically, but if it has an environmental effect it could be worth it for commercial reasons and if the CO2 actually to some extend could be reduced to CH4 in the reservoir, it is another advantage.
However I do not see sequestration being a good idea as a stand-alone solution. It is very expensive to dig as deep as it is necessary and the pump is also very expensive. I think a number mentioned was, like 3 times the cost of a coalfired powerplant in Denmark for the sequestration plant? or at the moment about 240€ for each ton CO2 in the first real test as opposed to the 45€ the companies' first guess was at. Compared to the about 60€ for each ton CO2 when scrubbing...
Sequestration in itself is theoretically sound at least for a period of time. The process relies on trapping the CO2 in a deep reservoir. All very well and sound. However, the longer you run this process the more spread out the plume of CO2 will be and at some point you will have either a very high pressure or CO2 exiting the trap: Sequestration is only a temporary solution that could result in serious uncontrollable release for later generations.
Before we look into how things change other systems, please make sure that the studies and research you're reading up with has error bars and justification. <3
You seem to be spending a lot of time responding to people, I hope you have the time to watch these 4 videos. I was wondering on how you would respond to the time scale of most people who believe in global warming. We have millions of years of data on the earth's climate but everyone who talks about global warming only seems to focus on less than 50 years of weather trends. When looked at from an overall picture it seems that the earth is behaving perfectly normally.
Also, is there any empirical data that shows CO2 can actually cause a drastic increase in temp and that humans are creating a large (>30% in my mind) of CO2 emissions? Most of what i've seen shows a logarithmic correlation instead of something to actually be worried about and that humans create and extremely small portion of CO2 emissions.
Climate change is one of the least problems for humanity atm, ofcourse it can change with new findings. Climate change is pretty solid fact, but is it caused by humans? Noone has solid proof of that. And thats about global, ofcourse cuting down 250k square kilometers of forest will turn the place into desert, and changes local climate. Or mount Kilimanjaro, some people say climate change has destroyed the glaciers on it, but the fact is that it was caused by cuting down the forests and stoping the water circulation thro evapotranspiration (you could fill 1/3 of sahara with rainforest, and it will be able to sustain itself). I study hydrobiology and paleolimnology makes current climate warming panic look like a foolish joke. Nature can handle climate change, but it cannot handle massive ammouts of chemicals, heavy metals etc in their circulations (water, carbon, nitrogen etc). And ofcourse the destruction of ecosystems, wich wont restore for thousends of years. These are the real dangers, climate chance is a joke compared to those. Nature WILL survive climate change with realtive eas, compared to the other dangers caused by humans.
Can add to my knowledge all biofuels so far have failed miserably (in some cases you need to spend 1 t of oil to produce 1 t of bioethanol), but there is one chance. Biggest potential is in the algea, but to squeezing the oil out of them is extremeley energy costly. But some laboratories in USA are currently working on cyanobacteria, who will extract oils from themselves, so it floats, and you can easily harvest it, but its still young research field.
yes yes, bad english
And ofcourse theres also the homo sapienses civlization suicide in near decades, but thats more of a moral, philosophycal question. But extremely serious non the less, specially when USA is turning into totalitarian police state.
EDIT: i allraedy regret posting, since i am still a student and my knowledge is limited, but i have my fair share of free seimars and lectures about the subjects. But still i get some of the things wrong. Thats why shoudlnt really post, just read, but oh well i am human lol.
On December 19 2011 09:21 LeibSaiLeib wrote: Climate change is one of the least problems for humanity atm, ofcourse it can change with new findings. Climate change is pretty solid fact, but is it caused by humans? Noone has solid proof of that.
What do you think happens to all the CO2 we release then, which has almost doubled atmospheric CO2 since the industrial revolution? CO2 is a pretty well-documented greenhouse gas, and the laws of physics are anything but ambiguous on this one.
Tell me what you think solid proof would look like.
Can add to my knowledge all biofuels so far have failed miserably (in some cases you need to spend 1 t of oil to produce 1 t of bioethanol), but there is one chance. Biggest potential is in the algea, but to squeezing the oil out of them is extremeley energy costly. But some laboratories in USA are currently working on cyanobacteria, who will extract oils from themselves, so it floats, and you can easily harvest it, but its still young research field.
Young is quite right. Algae is still a moonshot for now, but its always good to get ideas rolling!
On December 19 2011 09:19 WirelessWaffle wrote: You seem to be spending a lot of time responding to people, I hope you have the time to watch these 4 videos. I was wondering on how you would respond to the time scale of most people who believe in global warming. We have millions of years of data on the earth's climate but everyone who talks about global warming only seems to focus on less than 50 years of weather trends. When looked at from an overall picture it seems that the earth is behaving perfectly normally.
Also, is there any empirical data that shows CO2 can actually cause a drastic increase in temp and that humans are creating a large (>30% in my mind) of CO2 emissions? Most of what i've seen shows a logarithmic correlation instead of something to actually be worried about and that humans create and extremely small portion of CO2 emissions.
Human CO2 emissions, as a percentage, are small compared to natural CO2 emission: we emit 29 gigatons of CO2, which is a lot. But that's 4% of emissions. Why are we in trouble then?
Natural CO2 emissions are balanced out by natural CO2 reduction, because that's how the biosphere has evolved on Earth for hundreds of thousands of years. When we add more CO2, the sinks do not suddenly take in more CO2. They take in just as much CO2 as they have before. The CO2 levels in the atmosphere are well documented.
But suppose you were to think that the level of CO2 we added was too negligible to make a difference, then you could apply the same reasoning to particulate matter. Humans emit barely any particulate matter compared to natural sources and sinks, yet take one look at the skies in Beijing (or 1940 Pittsburgh, which were entirely black) and you knew that you were looking at something else entirely.
And yes, in the millions of years Earth has been around, the climate has shifted. This caused extinctions on massive scales. The kinds of shifts in CO2 we're seeing now have happened only on much, much exponentially larger time scales in the past, and the chance they'd happen right when the industrial revolution starts? You don't need to be a mathematician to see that's a pretty big coincidence.
On December 19 2011 09:19 WirelessWaffle wrote: You seem to be spending a lot of time responding to people, I hope you have the time to watch these 4 videos. I was wondering on how you would respond to the time scale of most people who believe in global warming. We have millions of years of data on the earth's climate but everyone who talks about global warming only seems to focus on less than 50 years of weather trends. When looked at from an overall picture it seems that the earth is behaving perfectly normally.
Also, is there any empirical data that shows CO2 can actually cause a drastic increase in temp and that humans are creating a large (>30% in my mind) of CO2 emissions? Most of what i've seen shows a logarithmic correlation instead of something to actually be worried about and that humans create and extremely small portion of CO2 emissions.
Human CO2 emissions, as a percentage, are small compared to natural CO2 emission: we emit 29 gigatons of CO2, which is a lot. But that's 4% of emissions. Why are we in trouble then?
Natural CO2 emissions are balanced out by natural CO2 reduction, because that's how the biosphere has evolved on Earth for hundreds of thousands of years. When we add more CO2, the sinks do not suddenly take in more CO2. They take in just as much CO2 as they have before. The CO2 levels in the atmosphere are well documented.
But suppose you were to think that the level of CO2 we added was too negligible to make a difference, then you could apply the same reasoning to particulate matter. Humans emit barely any particulate matter compared to natural sources and sinks, yet take one look at the skies in Beijing (or 1940 Pittsburgh, which were entirely black) and you knew that you were looking at something else entirely.
And yes, in the millions of years Earth has been around, the climate has shifted. This caused extinctions on massive scales. The kinds of shifts in CO2 we're seeing now have happened only on much, much exponentially larger time scales in the past, and the chance they'd happen right when the industrial revolution starts? You don't need to be a mathematician to see that's a pretty big coincidence.
You still haven't answered my second paragraph; yes humans are creating more CO2 but you still haven't showed me a correlation to climate change; also the climate change we're currently experiencing falls in line with normal trends of inter-glacial periods.
I do agree that pollution (your beijing/pittsburg examples) is a very serious problem but the bigger problem is that it's being overshadowed by CO2 emissions when I can't find any data saying CO2 emissions should be our number one priority compared to other forms of pollution.
Also the climate changes that caused mass extinctions (that I know of) were shifts into colder temperatures except for one (and again, that I know of) at the end of the Triassic period. That shift to warmer weather is currently theorized to be caused by ~10 gigatonnes of methane being released into the atmosphere.
TLDR; Humans are creating CO2, but what data is there to support that CO2 is a culprit in global warming? Why are obviously more toxic forms of air pollution being put on the back burner compared to CO2 emissions? If we're at the end of an interglacial period why are we only focusing on the earth heating up? It seems that everyone is tunnel visioning on CO2 emissions when there are other problems which seem more pressing.
EPT
