Comment Re:Climate is long time periods (Score 1) 786
I asked for validated models. You have yet to offer ONE.
Manabe and Wetherald 1967.
I asked for validated models. You have yet to offer ONE.
Manabe and Wetherald 1967.
I find you quite arrogant and condescending.
So, basically, you consider it condescending that I insist that you should actually look at data. Real data. Not blog posts.
And you complain that I only gave you a link to one source of data. OK, here are data from four continents:
Berkeley Earth: http://berkeleyearth.org/
Hadley Center Climate Research Unit: http://www.cru.uea.ac.uk/cru/d...
Goddard Institute for Space Studies: http://data.giss.nasa.gov/gist...
Japanese Meteorological Agency: http://ds.data.jma.go.jp/tcc/t...
Australian Meteorological Agency: http://www.bom.gov.au/state-of...
NOAA: http://www.ncdc.noaa.gov/cag/t...
What I am asking for is a climate model in that context. A model where I can feed in ANY data and it will output a fairly accurate prediction of how that climate system will operate. This should include being able to predict roughly what the climate conditions will be like throughout the planet over a period of some years.
This is called a "global circulation model" (usually abbreviated GCM). There are dozens of GCMs around, being run by hundreds of institutions on five continents You want a GCM, pick one.
At no point, should Mann or people like him make predictions about the climate beyond the accuracy of his own models.
The models have error bars. The quoted error bars in the consensus model of global warming is plus or minus 50% on the sensitivity to carbon dioxide.
So if his models only work to a few months or a year, then his predictions should not exceed that time period.
That's not the way it works. "Months" means weather, not climate. Now, there are also weather prediction models, and these also run using global circulation models-- but it's a different regime. Climate is what happens over decades.
And, global warming also is something that happens over decades. Longer time periods averages out the noise-- the random portion of the variation--so the longer the time period, the easier it is to pick out signal from noise.
If his models are accurate to 30 or 100 years then by all means make predictions on that time scale.
Exactly! Now you've got it. Climate is long term; weather is short term. Climate predictions need to be evaluated over decades. So, your question needs to be: how accurate were the predictions of thirty to fifty years ago? I'll reference, yet again, the classic greenhouse calculation, Manabe and Wetherald 1967. Their model predictions are within the current estimated range of climate sensitivity, so it's a good test. Their model computes a warming of about 0.7C for the measured amount of carbon dioxide added to the atmosphere since 1967. The actual value is 0.6. The year to year variation about the regression line is 0.15. So, yes, so far the model looks good, well within error bars. Now, if there is the purported hiatus in warming, and it's not a statistical fluctuation (so far, it's within the historical fluctuations), and if the purported hiatus holds for roughly another decade, the prediction will move outside the measurement error. But that's equivalent to saying "although the present measurements fit the data, future measurements won't." Maybe the models will have to be reevaluated if future data shows they're wrong. But so far, the data hasn't.
From what I've been able to gather, his models are incredibly inaccurate
Citation needed.
to such an extent that they're probably not even accurate enough to handle months much less years much less decades. And yet he makes predictions that span centuries. It is absurd.
Repeat: Months are weather. Decades are climate. The longer the period, the easier it is to model.
So, you've basically said that anything that relies on observations of nature is not a science. You don't actually know what science is.
Astronomy is not a science, according to you; it's just comparing observations to models.
But, comparing observations to models is exactly what science is. Controlled experiments are nice. However, much of science is done on systems where controls are not possible-- In the absence of a second Earth, we don't have a second climate.
Which is not to say that climate science doesn't have any controlled experiments. You can measure the greenhouse effect by pointing an infrared radiometer at the sky. You can measure the absorption of carbon dioxide in a temperature controlled gas cell. The basic science behind the greenhouse effect is physics, and this is actually pretty well understood.
The way I distinguish skeptics from deniers is that skeptics are interested in the science and want to learn more, while deniers won't do one iota of work to analyze, or critically examine in any way, the conclusions that they have been carefully spoon-fed by the blogo machine.
It is amusing, in a bitter way, the lengths that deniers will go in their rationalization of why they avoid looking at data or doing any thinking of their own. But this is the signature of deniers: plenty of talk, but no critical thinking (of the opinions that support what they already believe).
I've given you links to the data. Your response is that you won't listen to others (that would be appeal to authority!) but you won't look at any data yourself either. The only remaining option is that you will just believe what you have been told.
Oh, and you say it's not about the science anyway. Well, since you have so eloquently said that you don't understand it, I will believe at least that much.
Guess what? We're already gambling the world economy. We're already placing bets: all that science has done is to begin the process of calculating the odds and the payout. And the best thing about it? For the most part, the people making the bets aren't the ones taking the risks. We're gambling with other people's money.
If you want to place a bet against global warming, my suggestion is to buy land in Kiribati. This is a nation of small islands south of Hawai'i. As long as sea levels don't rise-- its average height is two meters above sea level-- this is paradise. A wonderful retirement home.
Land there is going cheap. Real cheap.
Sounds like some of them are as inept as Al Gore has been said to be.
Oddly, in the sciences (as opposed to in politics), acknowledging uncertainty and quoting error bars is considered to be a good thing. The fact that they acknowledge uncertainty is a useful indication that scientists are not "inept."
>When there's an alternative model that fits the data, believe me, people will pay attention. Many people have looked very hard to come up with an alternative model. So far, no success.
Sadly, that isn't quite true. You've forgotten about about selection and confirmation bias.
I'm sorry, but no. The null hypothesis is that carbon dioxide emitted by humans has no effect on climate. That hypothesis has been strongly ruled out by multiple measurements. It is simply not plausible.
The greenhouse effect model-- which, basically, is a model that says human-emitted carbon dioxide has the same effect on climate as natural carbon dioxide-- is at the moment the only model that fits known observations. There is no alternate model. People have been searching for one for a long time-- there are lots of people who very much want an alternative model, and have wanted one for years. The lack of an alternate model is not for lack of trying!
In addition to the obvious (has to come up with correct values for average temperature, day-night variation, variation with season, latitude, etc.) here are three additional things an alternate model has to explain:
1. It has to explain why carbon dioxide doesn't have the warming effect predicted from simple radiative transfer models.
2. It has to come up with an alternate explanation of the observed warming over the most recent period-- a period of time in which we have had good, and increasingly better and better, measurements of the planet. (We know it's not simply a change in the overall solar intensity, for example).
3. It has to come up with an explanation for why the amplifying effect postulated in part 2 (this has to have an amplifier, since we measure the input) doesn't also amplify the greenhouse warming.
This is very difficult. So far, nobody has come up with one that isn't easily shown to be faulty. And not for lack of trying-- the person who could come up with such a model that didn't fail would instantly become the most famous climate scientist in history.
I think you made the honest mistake, that I am directly debating you with the statements I made.
The difference between you and I, is that I DO care what the alarmist are saying because they are shaping public opinion, in the media and in politics.
Putting your head in the sand and saying, but science is sound, while ignoring
And, in fact, the science is sound. You have just written three posts in this thread asserting that the science is wrong... and now you're telling me, that's not your point, you don't know or care if the science is sound, what you care about is the "alarmists" trying to "shape public and political opinion"?
OK. This is slashdot. Some of us do care about the science, and do care about people attacking science to make polical points.
The science is sound. If you don't like the alarmism, go challenge that (hopefully with actual facts) and quit making uninformed attacks on the science based on some random post you googled on some blog.
..and, by the way, what I mostly see here is you taking any possible excuse to avoid looking at the data and doing your own analysis, thinking about statistical analysis and confidence. Good job in avoiding dealing with actual science!
"the measured temperatures are a nice validation that the models are in the right ballpark"
Well, within a factor of two. That's some ballpark.
That's right: the current estimate for the climate sensitivity to carbon dioxide is 3C plus or minus 1.5C (that is: 1.5C to 4.5C). That's a big ballpark, but, indeed, that's the quoted uncertainty. Here's a graph of model simulations, showing this variability: http://earthobservatory.nasa.g...
But the interesting thing is, the deniers aren't jumping on that-- they are, instead, all about how confident the scientists are.
If someone were trying to base public policy on a set of computer models which predicted changes in, say, IQ scores of black Americans, or academic success of women in STEM fields, and the predictions were off by a factor of two, how seriously would people take those models, or the people who came up with those models? Their proponents would be laughed at by everyone who wasn't vilifying them.
Perhaps you'd think that. But, instead, you will be hard pressed to find a denier website that even acknowledges this. Pointing out that the scientists acknowledge uncertainty in their models would destroy their argument that scientists are unwilling to acknowledge any uncertainty in their models.
The New York Times has been caught lately fixing the facts to meet the narrative.
Well, first I'd like to see a citation for that, since I have no idea what you're talking about. But you post as anonymous coward, so I doubt you have any references. Anonymous cowards should be assumed to be making stuff up unless proven otherwise.
But, in any case, that was just a convenient link to an essay from Robert Muller. Any other link would be just as good. If you want data, try http://berkeleyearth.org/
The narrative is what is important to Leftists not the facts. Leftists run the New York Times and Leftists are behind the whole of AGW politics.
Basically, anonymous posters whose main argument is "It's all a leftist conspiracy" should be ignored, since you don't have any interest in actual science.
Stop moving the goalpost. I made no claim other than the last 18 years has seen no warming.
The models, which we are supposed to base policy on, have been completely wrong.
Can you substantiate that with statistics? Are they completely wrong? Analyze the statistics. What does a statistical analysis say?
I can't do this for you. Don't grab somebody else's opinion from the internet; look at the data. Analyze the statistics. What do the statistics say? Is it statistically significant?
The reason they are wrong is, they cannot predict some of the most important drivers of long term climate, i.e. ENSO.
ENSO is a driver of short term climate, not long-term climate.
It is, indeed, one of the large sources of natural variation. But let me repeat something I've said many many times, and will, I expect, will have to say many more times. Human-induced climate effects are not instead of natural variations; they are in addition to natural variations.
We will see what it looks like in 32 years.I am not arrogant enough to make a prediction.
That's the problem, isn't it? None of the deniers make predictions. Because the deniers don't have any models. Not even one.
The argument seems to be "well, the measurements fit the data to within experimental error so far (which is accurate: do the analysis!) but in the future they won't. So, based on future measurements that haven't happened yet, global warming doesn't exist."
The thing is, where do you start and stop your (or mine) cherry picked data?
The easy answer: don't cherry pick. Use all the data.
That will always determine the slope.
So, don't cherry pick. Use all the data. Statistics tells us that a longer run of data will always have a better signal to noise ratio than a shorter run.
All I am saying is, the 21st has seen no global warming.
Interesting: you just changed the question. Now you are asking, "does the data show warming over a 15 year period ending in the present?" The answer to that question is the opposite of what you just said: The data shows that the climate warmed from 2000 to present. (I linked to you some sources of data: graph it yourself.) It warmed, oddly enough, just exactly according to prediction.
And if your response is "well, fifteen years is a cherry-picked number. If it warmed over that period, so what-- if you picked 18 years instead, it didn't warm!" Excellent. Exactly: if the calculation of slope depends on whether you add a single point (in this case the high point in 1998), that means that you're not analyzing a long enough run of data. A robust statistical analysis shouldn't be sensitive to any single point.
How does that track with blaming every single weather event on global warming.
I don't blame every weather event on global warming. If you hear people who are blaming every single weather event on global warming, stop listening to them. Another thing I've said many many times, and will, I expect, will have to say many more times. Weather is not climate.
How does that track with the doomsday scenarios, the scare and fear mongering, the alarmism, etc...
I am not interested in the doomsday scenarios, scare and fear mongering, alarmism. if there weren't so many people shouting so loudly that the science is wrong, I suppose I might spend some time debunking some of the worst of these. But there are always doomsday scenarios. People saying that scientists are frauds, and science is a hoax: that annoys me.
....But the thing is, if it hasn't warmed in 18 years, you CANNOT SAY anything that is going on now is worse then 10-18 years ago because of global warming , because there has been none in that period.
I never said that the weather is any worse then 10 to 18 years ago. Global warming is a long term effect. The change in 10 years is trivial. The effect on the weather of the change in 10 years is a trivial effect on a very small amount of warming; probably not observable.
Global warming is a long term effect, not a short term effect.
For the 18 years, look no further than Ben Santer. (is that good enough for you?)
http://onlinelibrary.wiley.com...
“Our results show that temperature records of at least 17 years in length are required for identifying human effects on global-mean tropospheric temperature.”
An interesting article. Let me point out that word "at least". Instead of "at least," what does that article say about "at best"?
"While previous work has focused on a single period of record, we select analysis timescales ranging from 10 to 32 years, and then compare all possible observed TLT trends on each timescale with corresponding multi-model distributions of forced and unforced trends. We use observed estimates of the signal component of TLT changes and model estimates of climate noise to calculate timescale-dependent signal-to-noise ratios (S/N). These ratios are small (less than 1) on the 10-year timescale, increasing to more than 3.9 for 32-year trends. This large change in S/N is primarily due to a decrease in the amplitude of internally generated variability with increasing trend length"
So, 32 years was the best signal to noise ratio, in an analysis looking at trends from 10 to 32 year. They concluded: the longer the better.
So: ignore the "least". What does the data say when analyzed over a 32 year trend?
Feed in past climate data and see if your climate model can predict the past or the present accurately.
And, surprise! It does. http://www.nytimes.com/2012/07...
While I agree with most of your post, what you describe here is not science. That approach turns science on its head. The scientific method begins with a reasoned hypothesis, followed by a prediction based on the hypothesis, and an experiment to prove or disprove this prediction.
Correct. The hypothesis dates back to Arrhenius 1896 http://www.lenntech.com/greenh... The numerical calculation of greenhouse warming due to carbon dioxide was first accurately done using measured value for infrared absorption and numerical integration of the profile was done in 1967 by Manabe and Wetherald-- it's summarized in any reasonable book about atmospheric science (such as the one on my desk at the moment, An Introduction to Atmospheric Radiation, by Liou (1980), p. 188). Calculating the greenhouse effect alone (that is, assuming no change in cloudiness, and constant relative humidity), Manabe and Wetherald showed "a ten percent increase in CO2 concentration (from 300 to 330 ppm) would lead to a warming of 0.3 K." It's a logarithmic response function (Arrhenius calculated that much back in 1895, although he didn't have the data to do the complete numerical integration), so it's easy to extrapolate this to the current carbon dioxide of about 400 ppm. It comes to about 0.8 K increase by their model.
Comparing it to the data, from 1967 on... looks like the experimental result matches the prediction.
Climate "science" on the other hand does exactly what you describe here. It looks at past data and attempts to fit it to a hypothesis.
Nope. The hypothesis dates back to Arrhenius. The detailed calculation dates to Manabe and Wetherald.
In any case, while the measured temperatures are a nice validation that the models are in the right ballpark, there's plenty of other data. You seem to be unaware that there is is a lot of measurements of the atmosphere.
That's not science at all. That's little more than a statistical model. These guys believe they have their answer and are trying to fit all observations to it.
That's a description of deniers. That's not the way climate science is done.
The reason we believe that the model is more or less accurate is that there are terabytes of data confirming it. The reason we don't believe that alternative models are accurate is that there aren't any. All of the alternative models proposed so far fail when compared against the evidence.
When there's an alternative model that fits the data, believe me, people will pay attention. Many people have looked very hard to come up with an alternative model. So far, no success.
You don't seem to know much about the subject, but this is not one or two scientists doing questionable work and then everybody else saying "oh, they must be right". There are thousands of scientists working on it; supercomputer models built on five different continents; ground, balloon, and satellite measurements, terabytes of data.
I can't figure out if he's trying to be Fozzie Bear or PacMan.
"waka waka," according to B. Kliban, is one of the phrases said by cats.
Suggest you just sit there and wait till life gets easier.
