Guest post by Steve Goddard
Archimedes had his eureka moment while sitting in the bathtub. Newton made a great discovery sitting under an apple tree. Szilárd discovered nuclear fission while sitting at a red light.
There was a time when observation was considered an important part of science. Climate science has gone the opposite direction, with key players rejecting observation when reality disagrees with computer models and statistics. Well known examples include making the MWP disappear, and claiming that temperatures continue to rise according to IPCC projections – in spite of all evidence to the contrary.
Here is a simple exercise to demonstrate how absurd this has become. Suppose you are in a geography class and are asked to measure the height of one of the hills in the Appalachian Plateau Cross Section below.
Image from Dr. Robert Whisonant, Department of Geology, Radford University
How would you go about doing it? You would visually identify the lowest point in the adjacent valley, the highest point on the hill, and subtract the difference. Dividing that by the horizontal distance between those two points would give you the average slope. However, some in the climate science community would argue that is “cherry picking” the data.
They might argue that the average slope across the plateau is zero, therefore there are no hills.
Or they might argue that the average slope across the entire graph is negative, so the cross section represents only a downwards slope. Both interpretations are ridiculous. One could just as easily say that there are no mountains on earth, because the average slope of the earth’s surface is flat.
Now lets apply the same logic to the graph of Northern Hemisphere snow cover.
It is abundantly clear that there are “peaks” on the left and right side of the graph, and that there is a “valley” in the middle. It is abundantly clear that there is a “hill” from 1989-2010. Can we infer that snow cover will continue to increase? Of course not. But it is ridiculous to claim that snow extent has not risen since 1989, based on the logic that the linear trend from 1967-2010 is neutral. It is an abuse of statistics, defies the scientific method, and is a perversion of what science is supposed to be.
Tamino objects to the graph below because it has “less than 90% confidence” using his self-concocted “cherry picking” analysis.
So what is wrong with his analysis? Firstly, 85% would be a pretty good number for betting. A good gambler would bet on 55%. Secondly, the confidence number is used for predicting future trends. There is 100% confidence that the trend from 1989-2010 is upwards. He is simply attempting to obfuscate the obvious fact that the climate models were wrong.
Science is for everyone, not just the elite who collect government grant money. I’m tired of my children’s science education being controlled by people with a political agenda.
Leif,
All of the models predicted declining extent during the 21st century.
http://www.eee.columbia.edu/research-projects/water_resources/climate-change-snow-cover/images/FreiGong2005Fig3ii.jpg
So far, they are all wrong. It is astonishing that you would deny this.
Steve Goddard (06:20:28) :
All of the models predicted declining extent during the 21st century.
Beginning about 1950, so that should be your starting point. Since we have no data 1950-1965, you have to start at 1966. It is astonishing that you won’t acknowledge that.
So far, they are all wrong. It is astonishing that you would deny this.
I do not deny that. I simply point out that your analysis is seriously flawed. Proper analysis shows that over 1966-2010 the trend is flat. R^2 = 0.0001. Since the models show a decrease and observations are flat, the models in aggregate are wrong [as I have said so many times], but it is equally wrong to claim a significant increase in snow cover. Perhaps you can recognize this, so we can close this silliness. What is the #NNNN number on my graph?
Steve Goddard (05:35:23) :
Tamino calculated 99% significance on that trend line, but then applied an undocumented “cherry picking” correction
Not ‘undocumented’ since the Monte Carlo simulation is a standard procedure to examine the sensitivity to assumptions.
Leif,
Suppose that your stock broker recommended that you short a stock based on his prediction that the stock would go down. 20 years later the stock has gone up by 15% and you have lost a lot of money.
How would you respond if he told you that your lost money is not “statistically significant?” You would probably punch him in the nose.
There is no better language in the world than that of logic. It is music to my ears (which behind THAT art is pure mathematical logic). I would rather listen to logic over a good joke (which I usually don’t get anyway) any time of the day or night.
Leif,
Over the last 20,000 years snow cover has declined substantially. Over the last four billion years snow cover has increased substantially. That is fascinating but my article covered the last 20 years, not the last 44 years, not the last 20,000 years, not the last four billion years.
Pamela,
Statistics is anything but pure mathematics. It is highly subjective.
“There are three kinds of lies: lies, damned lies, and statistics.”
Steve Goddard (07:05:00) :
How would you respond if he told you that your lost money is not “statistically significant?” You would probably punch him in the nose.
Not my style. Luckily, I’m perfectly capable of figuring out on my own what the signifiance is.
Steve Goddard: “Tamino calculated 99% significance on that trend line, but then applied an undocumented “cherry picking” correction which he claimed reduced it to “less than 90%.” I was (not) very surprised to find that he found a way to disagree with my assertion that the climate models were failing.”
Tamino applied a statistical test that is known to be only valid when the starting point has not been pre-selected based on “eyeballing” the data (you can use a random starting point or use all the data, but you don’t get to choose where to begin based upon what “looks right”) and he got a value of 99%–for doing it wrong. Then he asked: Just how much does doing it wrong cause the value to be in error? And he answered the question by applying a basic procedure documented in virtually any statistic text–monte carlo analysis. The answer to the question turned out to be: doing it wrong makes the value very much in error–and if you do it right, the “statistical significance” vanishes.
Note that a general conclusion like “the climate models are failing” is a statistical assertion, because the climate models make only statistical predictions about trends, and no prediction at all about whether the snow cover on one particular year is going to be greater or less than the snow cover on another particular year. So to make any conclusion about whether the predicted trend agrees with the actual trend, you have to give up looking out your window and actually do statistics–and if you are going to do statistics, then you need to do them right.
trrll,
So what you are saying is that just because I looked at the Rutgers data and noticed that snowfall has been increasing since 1989, snowfall isn’t really increasing. Had I not looked, it would be increasing.
I didn’t realize that I had so much power or influence that I could actually change the past by looking at it. Sort of an inverse Heisenberg principle.
The stock market has gone down 2,000 points during the last two years, but not in a “statistically significant” fashion.
http://finance.yahoo.com/q/bc?s=^DJI&t=2y&l=on&z=m&q=l&c=
Therefore all the trillions of dollars which people have lost, weren’t really lost. Everyone will be so happy to hear that statisticians have solved the world’s financial problems.
Steve Goddard (15:06:43) :
So what you are saying is that just because I looked at the Rutgers data and noticed that snowfall has been increasing since 1989
Had you looked three years ago it wouldn’t have. And possibly next year or the year after that it won’t have. So, confusing weather with climate isn’t a very fruitful thing to do.
Leif,
You sound like a broken record. The top four NH snow years are 1978, 2010, 2008, 2003. I have yet to see a weather forecast which predicts trends for a period of three years, eight years or twenty years. Snow extent has been increasing for a long time and you just refuse to see it.
Weather models are good for three days, not decades.
Steve Goddard (04:52:49) :
Snow extent has been increasing for a long time
You are laboring hard to show that Global Warming causes [or is associated with] more extended snow cover. Unfortunately, that is just wishful thinking on your part, as ‘one swallow does not make a summer’. The data simply does not support your contention, but then we knew that all along, didn’t we?
Leif,
I have made it more than abundantly clear that I am not trying to associate any physical cause with the observation that NH winter snow extent has been increasing for the last 20 years. The data definitively supports my conclusion, but you knew that all along, didn’t you?
aping doesn’t become you. I think you have demonstrated that no conclusions can be drawn from the data as yet. I agree that that does not prevent people from drawing wrong or unsubstantiated ones.
Steve Goddard (04:52:49) :
broken record: The top four NH snow years are 1978, 2010, 2008, 2003.
If next two years snow cover is 40 million km2, the top four years would stay the same, and you would still say that snow cover has increased the last 24 years, right?
Leif,
If the trend changes in the future (and it will) then we will see that reflected in the graphs. That is the whole point of this article.
Having said that, there has only been one year in the last decade below 44 million, so your speculation about 40 million seems extremely unlikely.
Steve Goddard (12:26:12) :
Having said that, there has only been one year in the last decade below 44 million, so your speculation about 40 million seems extremely unlikely.
If memory serves 1981 was less than 41 million, so that is definitely within range of values expected for the observed flat ‘trend’: http://leif.org/research/Snow-Cover-1966-2010-NH-Winter.png
What is the #NNNN number in the lower RH corner?
Leif Svalgaard (11:58:57) :
Steve Goddard (04:52:49) :
“broken record: The top four NH snow years are 1978, 2010, 2008, 2003.”
And the point was that the top four years say nothing about what next year will bring. So few extremes are not indicative of any significant trend.
“Tamino objects to the graph below because it has “less than 90% confidence” using his self-concocted “cherry picking” analysis.”
Well, Tamino uses a time series of more than 40 years and you use a time series of 20 years, when everybody knows that statistical analysis of climate data should be done over periods of at least 30 years. So cherry-picking by cherry-picking, his makes more sense.
Miguel,
Does it make sense to do a linear estimation across two legs of a cyclical function? Of course not.
“Everybody knows” doesn’t carry much weight. Time to start thinking instead.