Moderated Global Warming Discussion

[r-j]

There is a simple and real fact that the people who fear and hate coal don't know. It's a fact and it won't change.

A chunk of quality coal, the size of a man fist, in a modern engine, will produce more energy to do useful work, than a man can do in a day. Oil is the same sort of concentrated energy.

 

[r-j]

From the beginning he’s tried to use regional and/or short term data to make a generalized claim.

More nonsense. Do you just make things up like that in every subject?

Smart Tamino, stating the obvious once again in a vain attempt to educate climate change deniers.
Too Little Time.

He seems like some sort of idiot.

None of the five main global temperature data sets shows statistically significant trend since 2003. But if they’re adjusted to remove the influence of known extraneous factors — el Nino, aerosols, and solar variation — there is one which does show a statistically significant trend despite being limited to a mere ten years and two months. That’s the lower-troposphere data from UAH:

Its estimated trend since 2003 is 0.018 deg.C/yr, even higher than its estimated trend since these data begin in December 1978.
http://tamino.wordpress.com/2013/04/02/too-little-time/

Using the trend generator from http://www.skepticalscience.com/trend.php his problem seems obvious. The 2003=2013 UAH trend actually looks like this

Trend: 0.0033 ±0.0403 °C/year (2σ)
β=0.0033320 σw=0.0049037 ν=16.855 σc=σw√ν=0.020132


It's one reason I don't waste time reading blogs

 

[Reality Check]

He seems like some sort of idiot.

It seems idiotic to state without evidence that anyone seems idiotic.
Too Little Time.

None of the five main global temperature data sets shows statistically significant trend since 2003. But if they’re adjusted to remove the influence of known extraneous factors — el Nino, aerosols, and solar variation — there is one which does show a statistically significant trend despite being limited to a mere ten years and two months. That’s the lower-troposphere data from UAH:
(graph)

Using the trend generator from http://www.skepticalscience.com/trend.php your problem is obvious.
This is not the adjusted to remove the influence of known extraneous factors el Nino, aerosols, and solar variation - lower-troposphere data from UAH.
Not realizing this could make people think that you are "some sort of idiot" .

You need the other trend generator to explore the Foster & Rahmstorf adjusted data. This can produces a similar graph with "Trend: 0.0309 ±0.0209 °C/year (2σ)".
N.B. this is probably not the source of Tamino's graph since it does not have the Skeptical Science watermark.

 

[Reality Check]

Not if we believe that the trends are useless, as we have been told over and over. The NCDC doesn't give enough data to know that the graphs I posted actually mean anything!

No one says that the trends are useless.
The NCDC doesn't give enough data to know that the graphs that anyone has posted have significant trends.

But graphs of any relevant data over periods of more that 20 years will have significant trends!
The more data you have, the less uncertainly in the trends there are. Using Tamino's example at the SC calculator that you brought up:
UAH Adjusted from

2003: 0.0309 ±0.0209 °C/year (2σ)
1993: 0.0188 ±0.0060 °C/year (2σ)
1983: 0.0151 ±0.0031 °C/year (2σ)
​

Note how the uncertainties decrease with more data.

Thus r-j: you have still debunked your own claim that winters are getting colder with these examples of winters getting warmer over periods of 20 years or more .

 

[Pixel42]

Depending on the person, ten years is fine for ice, twenty years is enough to freak out over globally, but a hundred years of hard data is useless. Your science, it's irrefutable.

For goodness' sake! It doesn't depend on the person, or the accuracy of the
data, or anything except the statistical significance.

If something is changing very rapidly then just 10 years of data may be enough
to show a statistically significant trend. If something is changing very slowly it may take 100 years of data to show a statistically significant trend. If something isn't changing at all then a thousand years of data won't show a statistically significant trend.

Some of the graphs of trends on the NCDC site may well be statistically significant, but since its tool (unlike the SkS one) doesn't calculate it we don't know. We can make a guess based on how steep the slope is and the amplitude of the variations, but we don't know.

This is not rocket science, it's statistics 101.

 

[macdoc]

what doesn't much matter is a small state with a tiny +/- in the middle of a continent.

 

[DC]

UhHuh.

http://www.realclearpolitics.com/articles/2013/04/02/meet_the_new_climate_deniers_117759.html

yeah sure the Deniers think this shows AGW is not happening...
yes it is still if you not only look at surface temps. but take into account the whole System. scientists are doing that and Show that there still is a huge deal of warming.

Just because the surface temps do not rise that much atm doesn't mean the science was wrong. it doesn't Change anything about CO2's absorbtion of IR Radiation.

but sure the Deniers want to make it look as if that debunks AGW. that is disconected from reality. very dangerous.

 

[Lennart Hyland]

So its been like 10 pages now of discussing local winter weather? And I still dont know what r-j's agenda/opinion about AGW is...

 

[r-j]

For goodness' sake! It doesn't depend on the person, or the accuracy of the data, or anything except the statistical significance.

Finally, we are getting at least close to an answer from you. No wonder you wouldn't give a straight answer when I asked you about what you were willing to accept.

Of course the accuracy of the data matters when it comes to calculating a trend. You are completely wrong there. Now your inability to understand a trend for weather data is starting to make sense.

If something is changing very rapidly then just 10 years of data may be enough to show a statistically significant trend.

There is where statistics and a long time period actually matter a lot. If you only have data from 1993, or 1979, and you want to say "something is changing very rapidly", in regards to global data, you don't have enough data to even know, much less say there is a trend. You are trying to compare too short a period of time, in regards to change.

If something is changing very slowly it may take 100 years of data to show a statistically significant trend. If something isn't changing at all then a thousand years of data won't show a statistically significant trend.

There it is. How can you know if something is changing very slowly, if you don't have 100 years of data? More importantly, as we can see in many of the temperature graphs from the NCDC, what seems like rapid change is actually no change at all, over a hundred year period. If you just take the last 17 years, you get these drastic changes. Some show warming, others show cooling.

Just like the ice in summer levels, looking at a short period of time makes it look like a drastic reduction. If you had 100 years of data, you could say the importance of the ice reduction is meaningful. Same for the satellite sea level data since 1993. Trends are regularly stated from a very short time period. From very questionable data sources.

In fact, CO2, sea level, and ice extent all are regularly shouted as clear trends. And, of course an increase of .0018 degrees Celsius a year is considered important, if it shows warming.

But 1.6 degrees a decade, from data that has no accuracy problems, is dismissed with the wave of a hand, and some insults for good measure.

And you think you are winning the war. To convince skeptics that you are right, and everybody who questions you is an evil denier.

Priceless.

 

[DC]

Finally, we are getting at least close to an answer from you. No wonder you wouldn't give a straight answer when I asked you about what you were willing to accept.

Of course the accuracy of the data matters when it comes to calculating a trend. You are completely wrong there. Now your inability to understand a trend for weather data is starting to make sense.

There is where statistics and a long time period actually matter a lot. If you only have data from 1993, or 1979, and you want to say "something is changing very rapidly", in regards to global data, you don't have enough data to even know, much less say there is a trend. You are trying to compare too short a period of time, in regards to change.

There it is. How can you know if something is changing very slowly, if you don't have 100 years of data? More importantly, as we can see in many of the temperature graphs from the NCDC, what seems like rapid change is actually no change at all, over a hundred year period. If you just take the last 17 years, you get these drastic changes. Some show warming, others show cooling.

Just like the ice in summer levels, looking at a short period of time makes it look like a drastic reduction. If you had 100 years of data, you could say the importance of the ice reduction is meaningful. Same for the satellite sea level data since 1993. Trends are regularly stated from a very short time period. From very questionable data sources.

In fact, CO2, sea level, and ice extent all are regularly shouted as clear trends. And, of course an increase of .0018 degrees Celsius a year is considered important, if it shows warming.

But 1.6 degrees a decade, from data that has no accuracy problems, is dismissed with the wave of a hand, and some insults for good measure.

And you think you are winning the war. To convince skeptics that you are right, and everybody who questions you is an evil denier.

Priceless.

Do you actually think you are fooling anyone here with your ignorant nosnense and your cheap attempts to expose others lack of knwoledge in which you fail misserably?

your trend was dismissed because it is a small Region and you want it to mean more that it actually does.

all you exposed here is your ignorance and arrogance to think you are in a Position to lecture others. when you which to be taken serious, i recommend changing attitude and start debating climate instead of weather and global rather than regional.

 

[DC]

Same for the satellite sea level data since 1993. Trends are regularly stated from a very short time period. From very questionable data sources.

do you have evidence for this Claim? define a very short time period and what questionable sources are you talking about?

 

[r-j]

it doesn't Change anything about CO2's absorbtion of IR Radiation.

absorbtion
Although it’s “absorbed” and “absorbing,” the correct spelling of the noun is “absorption.”
http://public.wsu.edu/~brians/errors/absorbtion.html

The temperature data doesn't change anything, ever. It's just our best idea of the heat balance on a global scale, and weather and climate on a local one.

Just because the surface temps do not rise that much atm doesn't mean the science was wrong.

When I read something like that, it's hard to imagine you aren't being sarcastic.

If somebody predicted a warm month of March for 2013, and then it actually was record cold, if they said "just because the temperatures were cold doesn't mean it wasn't a warm March", I can't imagine they are being serious.

Who would say something like that?

 

[lomiller]

Not if we believe that the trends are useless, as we have been told over and over. The NCDC doesn't give enough data to know that the graphs I posted actually mean anything!

No one has told you trends are useless. What you have been told is that trends need to be statistically significant to use them.

The NCDC doesn't give enough data to know that the graphs I posted actually mean anything!

They give you sufficient data to calculate trend and significance. It’s not their fault you are not interested in doing so.

 

[lomiller]

More nonsense. Do you just make things up like that in every subject?

Don’t project. I take it from this you are ready to admit your previous suggestion that winters are getting colder are false?

He seems like some sort of idiot.

He’s a statistician who specializes in time series has published papers on the topic and has co-written papers on climate trends.

Using the trend generator from http://www.skepticalscience.com/trend.php his problem seems obvious. The 2003=2013 UAH trend actually looks like this

Garbage in garbage out.

 

[r-j]

No one has told you trends are useless. What you have been told is that trends need to be statistically significant to use them.

Pure nonsense.

The NCDC data is claimed, by you, not to contain enough information to be able to determine if they are "statistically significant". You and others claim that if you can't determine if they are "statistically significant" because they do not give enough information. You then say, and I quote, "trends need to be statistically significant to use them".

I then describe that you said the NCDC data is useless for creating a trend. And then you say, and I quote, "No one has told you trends are useless.".

That you can't see this, once again, priceless.

 

[r-j]

One important thing I learned from this thread, concerning the trends and the global climate data. They "adjust" the data to get rid of things like the PDO, pollutants, clouds and other random elements that influence the temperatures of the planet.

That's good to know. It explains why the actual data, the measurements of temperature, often do not match what we see in the global temperature data.

Do they also remove the solar variations?

 

[Belz...]

Pure nonsense.

Actually, it was syntaxically and grammatically correct, and I understood what was meant. It also makes sense that statistically-insignificant data isn't very useful. That's not nonsense.

 

[lomiller]

Pure nonsense.

Statistical significance is not “pure nonsense” It’s an essential element of any valid statistical analysis.

The NCDC data is claimed, by you, not to contain enough information to be able to determine if they are "statistically significant". You and others claim that if you can't determine if they are "statistically significant" because they do not give enough information.

Neither I nor anyone else has said any such thing. There is sufficient data available to determine if a trend is statistically significant or not. What people have been telling you is that many of the “trends” you refer to are NOT statistically significant.

 

[DC]

One important thing I learned from this thread, concerning the trends and the global climate data. They "adjust" the data to get rid of things like the PDO, pollutants, clouds and other random elements that influence the temperatures of the planet.

That's good to know. It explains why the actual data, the measurements of temperature, often do not match what we see in the global temperature data.

Do they also remove the solar variations?

oh i might learn something new here.

how does one adjust the data to account for pollutants PDO clouds etc?

and why do you say that measurements of temperature do not match what we see in global temp data? got an example?

 

[r-j]

There is sufficient data available to determine if a trend is statistically significant or not. What people have been telling you is that many of the “trends” you refer to are NOT statistically significant.

That is at least partly right, but very wrong in essence.

There are numerous examples of claims made about trends, and time periods. Tamino (poor tamino) even had his ten year trends drug into the fray. (wasn't he being deliberately wrong to make a point?)

It started with Florida, included to Cali, Illinois then Tenn and Kentucky, jumped to the corn and soy belt, them Montana, as proxy for Canada. Because somebody mentioned Illinois that became a battleground state.

I mentioned it show a trend towards colder winters.

You asserted that Illinois had cooled over the last 30 years, I put the parameters in correctly and proved that the data actually showed a 0.6F per decade increase in temperature. That cannot correctly be called a climate trend (don't be fooled by the tool's use of the term) because (a) the tool does not calculate the statistical significance and (b) a single state is not a large enough area for 30 years' worth of data to show a meaningful trend.

And there we have it. When I posted the hundred year trend, that also was dismissed.

30 years of data is just about the minimum that is likely to show a statistically significant trend, provided you're looking at data for the whole world. The smaller the area you're looking at the longer time period you need before you can say that any apparent trend is genuine, i.e. is a change in climate, not just variations in the weather.

That is actually true, and not just for the whole world. You need a long time period to determine climate. No exceptions. If the climate changes a lot in a hundred years, temps up and down, rainfall up and down, it's actually called a variable climate. Parts of Africa are well known for drastic climate change in short periods of time. Mostly due to rain of course.

IIRC the contiguous US covers less than 10% of the earth's surface, so even its 30 year winter trend (warming of 0.51F per decade) is of debatable significance.

Part of the problem is the lack of understanding about climate, weather, and regional temperature data. Some of it seems just pure cussedness. Nobody ever claimed Illinois is a proxy for the globe, nobody ever said the world is showing a trend of colder global temperatures. The strawmen are all over this topic at this point.

Reading weeks of the thread in a shorter time, it appears that at core the heat was about vague terminology. His main assertion boils down to "for some locations, and some time periods, the regression trendline for winter temperature is negative". Actually, nobody really doubts that - with noisy data we would expect it.

This belief that data is noise, it might be part of the problem. temperature data from the NCDC stations is not noise, even if daily weather is of course. The monthly mean, the yearly mean, it's how you separate out the noise, so you can see what is happening over time. rainfall data especially is looked at to see trends, as a decrease in rain is a vital bit of information. Due to the noisy nature of weather, a trend is almost useless for predicting the next year. But over a long time period you can get a good idea of probabilities.

Note that some here have not dismissed the possibility that there could be some real world negative trend in winter temperatures on some timescale and for some locations - only that rj's level of analysis has failed to distinguish signal from noise well enough to be significant evidence for it.

There has been a lot more than that siad, but it's just opinions and anecdotes, so it matters little.

All the opinions in the world won't change the rainfall in the corn belt.