• This is a political forum that is non-biased/non-partisan and treats every person's position on topics equally. This debate forum is not aligned to any political party. In today's politics, many ideas are split between and even within all the political parties. Often we find ourselves agreeing on one platform but some topics break our mold. We are here to discuss them in a civil political debate. If this is your first visit to our political forums, be sure to check out the RULES. Registering for debate politics is necessary before posting. Register today to participate - it's free!

No Significant Trend Can Be Discerned from Global Tempature Data

Tim the plumber said:
Well, if the system is highly complex with many variables in play then there is no reason to consider CO2 to be all that significant a driver of climate then is there.

The entire basis of the AGW cult is that CO2 is this primary driver of the temperature changes we have seen. If it is not then there is nothing to worry about.
Click to expand...

That's a bizarre leap. It's complicated, therefore it can't be bad?
 
Deuce said:
That's a bizarre leap. It's complicated, therefore it can't be bad?
Click to expand...

No.

I will try to get this across for those who really don't want to think so please bear with me.

If the rise in temperatures is all down to CO2 then the IPCC's predictions hold true. At least for the lower end of them where there is no problem really.

If the rise in temperatures since 1970 (or whenever) is due to lots of factors only one of which is CO2 then the IPCC's predictions don't hold true and we will be well below their lowest numbers. Thus there is nothing at all to worry about.

Only if there is some sort of vast increase in temperatures and the rate of rise of temperatures do we need to think about this at all.
 
Tim the plumber said:

No.

I will try to get this across for those who really don't want to think so please bear with me.

If the rise in temperatures is all down to CO2 then the IPCC's predictions hold true. At least for the lower end of them where there is no problem really.

If the rise in temperatures since 1970 (or whenever) is due to lots of factors only one of which is CO2 then the IPCC's predictions don't hold true and we will be well below their lowest numbers. Thus there is nothing at all to worry about.

Only if there is some sort of vast increase in temperatures and the rate of rise of temperatures do we need to think about this at all.
Click to expand...

And what about other factors having a cooling influence? Wouldn't they suggest the effects of CO2 are stronger?

The sun has been flat for decades, and most recently took a dip. And yet, temperatures have not declined.
 
Last edited:
Deuce said:
And what about other factors having a cooling influence? Wouldn't they suggest the effects of CO2 are stronger?

The sun has been flat for decades, and most recently took a dip. And yet, temperatures have not declined.
Click to expand...

Factors that have changed.

The cooling factors, it is agrued by the IPCC, have remained the same. They hold that all the warming is a result of CO2 changes. And that the rate of warming will somehow hugely increase..... because of.... well it will!!!
 
Tim the plumber said:

Factors that have changed.

The cooling factors, it is agrued by the IPCC, have remained the same. They hold that all the warming is a result of CO2 changes. And that the rate of warming will somehow hugely increase..... because of.... well it will!!!
Click to expand...

You're suggesting they have no basis for predicting an acceleration of temperature rise.
You're suggesting they just made that up out of nothing.

:roll:
 
Deuce said:
You're suggesting they have no basis for predicting an acceleration of temperature rise.
You're suggesting they just made that up out of nothing.

:roll:
Click to expand...

Correct.

Given we are in the place on the graph with the maximum rate of CO2 increase as a percentage of the total CO2 and that more CO2 has a progressively smaller effect thus doubly the impact of increasing CO2 should be highest in terms of energy inbalance now why will the rate of temperature increase ever be as high as now?

You are the one who is convinced it will be so perhaps you can tell me the answer.
 
Deuce said:
They choose the start of the instrumental temperature record.

And no, they don't ignore data prior to that.
Click to expand...

Yes we know they don't ignore it- they just make it up . ;)
 
LowDown said:
Data from weather stations on the earth's surface are available from 1880 to the present day. According to professional statisticians there is no trend in this data that can properly be distinguished from natural random variation...
Click to expand...

Natural variation is not random. But LowDown did manage to reach his second sentence before losing all credibility, and even the slightest hint of progress should be encouraged, so three cheers for that!

This really is one of the most bizarre threads I've seen on this topic, and that's saying a lot. Okay, so if you have 20 or 50 or 100 attempts at it, you can make a random time series that looks kind of like the global temperature series.

Does this tell us that the global temperature record is incorrect? No...
Does it tell us there is no greenhouse effect?
Does it tell us that there's been no increase in greenhouse gases?
Does it tell us that there've been big increases in solar output (contrary to records suggesting it peaked in 1958)?
Does it tell us what impacts volcanism or oceanic oscillations or cosmic radiation or alien death rays or anything else have had on the climate?

Does it tell us anything whatsoever which is in any way relevant to our understanding of the climate? No, of course not. But that doesn't stop an army of charlatans and useful idiots from pretending that it does :roll:
 
Last edited:
Deuce said:
"No statistics are valid"

Climate Skeptics, 2016
Click to expand...

I am pretty sure they picture a bunch of hippies sneaking onto glacier and permafrost every night with electric blankets.
 
Mithrae said:
Natural variation is not random. But LowDown did manage to reach his second sentence before losing all credibility, and even the slightest hint of progress should be encouraged, so three cheers for that!

This really is one of the most bizarre threads I've seen on this topic, and that's saying a lot. Okay, so if you have 20 or 50 or 100 attempts at it, you can make a random time series that looks kind of like the global temperature series.

Does this tell us that the global temperature record is incorrect? No...
Does it tell us there is no greenhouse effect?
Does it tell us that there's been no increase in greenhouse gases?
Does it tell us that there've been big increases in solar output (contrary to records suggesting it peaked in 1958)?
Does it tell us what impacts volcanism or oceanic oscillations or cosmic radiation or alien death rays or anything else have had on the climate?

Does it tell us anything whatsoever which is in any way relevant to our understanding of the climate? No, of course not. But that doesn't stop an army of charlatans and useful idiots from pretending that it does :roll:
Click to expand...

You clearly don't understand the problem. How do we know for certain that the temperature record from 1880 to now is the result of a trend showing a gradual increase of about 0.8 degrees versus the result of random variation?

Statistical analysis of the data in the proper fashion, which is to analyze it as a time series, which is what it is, indicates that you can't tell whether it's random or due to a trend. The reason being that a time series can have that much random variation, which is what we can easily see from the graphs. It's a different sort of variation than what you see with a series of random numbers, and it looks a lot like the temperature record.

All that is required to have a time series is to assume that a given state of the climate in time determines the next state. And we know that climate scientists agree with this because that's the way their climate models work. So if you think that climate models can accurately model the climate then you must also agree that random variation of a time series is a valid model to compare to the temperature record. And we do that and find that, yes, the whole thing could be random variation.

And as far as our understanding of the climate is concerned, what this means is that we don't have adequate data to conclude anything about the climate. And as Wunsch, the MIT professor, said, this leaves us with a choice about what to say about the data. Are we going to be honest about it and tell people that we can't come to any conclusion, which will mean we lose our funding and possibly our job? Or do we cook up some **** and bull story about it?
 
LowDown said:
You clearly don't understand the problem. How do we know for certain that the temperature record from 1880 to now is the result of a trend showing a gradual increase of about 0.8 degrees versus the result of random variation?
Click to expand...

There has been a ~0.8 degree increase in global surface temperatures. That's not the result of a trend; trend is a descriptive term, not prescriptive. It is the result of a certain causes, which can be understood with some degree of confidence through empirical investigation. That you think you're asking a serious question whether changes in the climate might be random shows that you have approximately zero understanding of the subject.

LowDown said:
Statistical analysis of the data in the proper fashion, which is to analyze it as a time series, which is what it is, indicates that you can't tell whether it's random or due to a trend. The reason being that a time series can have that much random variation, which is what we can easily see from the graphs. It's a different sort of variation than what you see with a series of random numbers, and it looks a lot like the temperature record.
Click to expand...

This time series is even more easily reproducible through random variation:
spot-dec10-fig2.jpg


Therefore you would conclude that we can't tell whether the annual number of nonimmigrant visas issued in the United States is random or due to a trend. That 'logic' is utterly and irredeemably stupid of course: There are a number of more and less obvious factors influencing that time series, most if not all of which can be identified and quantified to some extent.

The fact that random variation could produce the same results means precisely two-thirds of diddly squat.

That anyone imagines that it does mean something says quite a lot about the intelligence of our 'sceptic' community, sadly. Maybe some of your alleged statisticians need funding too?
 
Mithrae said:
There has been a ~0.8 degree increase in global surface temperatures. That's not the result of a trend; trend is a descriptive term, not prescriptive. It is the result of a certain causes, which can be understood with some degree of confidence through empirical investigation. That you think you're asking a serious question whether changes in the climate might be random shows that you have approximately zero understanding of the subject.



This time series is even more easily reproducible through random variation:
spot-dec10-fig2.jpg


Therefore you would conclude that we can't tell whether the annual number of nonimmigrant visas issued in the United States is random or due to a trend. That 'logic' is utterly and irredeemably stupid of course: There are a number of more and less obvious factors influencing that time series, most if not all of which can be identified and quantified to some extent.

The fact that random variation could produce the same results means precisely two-thirds of diddly squat.

That anyone imagines that it does mean something says quite a lot about the intelligence of our 'sceptic' community, sadly. Maybe some of your alleged statisticians need funding too?
Click to expand...

You don't understand the statistics. From the graph you posted we can say that the overall trend is not significantly different from zero. It's the same with the global temperature data. That there may or may not be things influencing the data that could be accounted for means nothing, especially if your theory about what influences it is wrong.

Now, saying that the trend from 1880 to now is not significantly different from zero isn't the same as saying that the trend from 1976 to 1999 is zero. I haven't checked, but I suspect that the trend is significantly greater than zero for that time period.

The question a statistician would ask is how likely it is that a given data set could be duplicated by random variation with a trend of zero. If the chances are less than 5% then you'd have significance, but in a time series you'd have to have a pretty pronounced looking trend above the randomness in order to be able to see it. You look at how much the data changes in going from one time point to the next. If it goes up more than it goes down it seems pretty obvious that there would be a positive trend, but then you have to consider variance.

It's also not the same as saying that there is no positive trend. There might be such a a trend, we just can't be sure from the available data.
 
LowDown said:
You don't understand the statistics. From the graph you posted we can say that the overall trend is not significantly different from zero. It's the same with the global temperature data. That there may or may not be things influencing the data that could be accounted for means nothing, especially if your theory about what influences it is wrong.

Now, saying that the trend from 1880 to now is not significantly different from zero isn't the same as saying that the trend from 1976 to 1999 is zero. I haven't checked, but I suspect that the trend is significantly greater than zero for that time period.

The question a statistician would ask is how likely it is that a given data set could be duplicated by random variation with a trend of zero. If the chances are less than 5% then you'd have significance, but in a time series you'd have to have a pretty pronounced looking trend above the randomness in order to be able to see it. You look at how much the data changes in going from one time point to the next. If it goes up more than it goes down it seems pretty obvious that there would be a positive trend, but then you have to consider variance.

It's also not the same as saying that there is no positive trend. There might be such a a trend, we just can't be sure from the available data.
Click to expand...

The positive temperature trend since 1880 is statistically significant.
 
LowDown said:
You don't understand the statistics. From the graph you posted we can say that the overall trend is not significantly different from zero.
Click to expand...

No, you can't. Even in that graph the trend is different from zero, obviously; whether it is significantly different depends on the hypothesis you're considering and hence the parameters of the null hypothesis. Otherwise, you'd have a universal rule that no trend can ever be significant unless it rises at least N degrees above the X axis, which is utterly absurd. Highly stable systems can have tiny yet statistically significant trend changes. The fact that you don't realise any of this shows that you've really got no idea what you're talking about :roll:

In the case of global temperatures vis a vis anthropogenic climate change (or anthropogenic global warming), the null hypothesis is that all observed changes can be accounted for by changes in insolation, cosmic radiation, oceanic circulation, volcanism, changes in albedo (and anthropogenic aerosols). None of those are random phenomena over the 1880-present timeframe: We know when major volcanoes erupted, we know that there was a peak of solar activity in the late 1950s, we know that over the last 130 years the net effect of the major oceanic oscillations has been close to zero (though contributing some to both the mid century cooling, and the early and late century warming), we know that anthropogenic aerosols have had a net cooling effect...

If the sum of all other factors and their associated uncertainty ranges cannot account for all of the observed warming at a given probability threshold - and to the best of our knowledge, they cannot - then the null hypothesis (that GHG increases are unnecessary to explain the observations) is rejected and we have a statistically significant result for the hypothesis that GHG increases have contributed to the warming.

Which we do. Playing with random graphs is an utterly clueless exercise which tells us nothing about the climate, only about the general intelligence of self-proclaimed 'sceptics.'
 
Mithrae said:
No, you can't. Even in that graph the trend is different from zero, obviously; whether it is significantly different depends on the hypothesis you're considering and hence the parameters of the null hypothesis. Otherwise, you'd have a universal rule that no trend can ever be significant unless it rises at least N degrees above the X axis, which is utterly absurd. Highly stable systems can have tiny yet statistically significant trend changes. The fact that you don't realise any of this shows that you've really got no idea what you're talking about :roll:

In the case of global temperatures vis a vis anthropogenic climate change (or anthropogenic global warming), the null hypothesis is that all observed changes can be accounted for by changes in insolation, cosmic radiation, oceanic circulation, volcanism, changes in albedo (and anthropogenic aerosols). None of those are random phenomena over the 1880-present timeframe: We know when major volcanoes erupted, we know that there was a peak of solar activity in the late 1950s, we know that over the last 130 years the net effect of the major oceanic oscillations has been close to zero (though contributing some to both the mid century cooling, and the early and late century warming), we know that anthropogenic aerosols have had a net cooling effect...

If the sum of all other factors and their associated uncertainty ranges cannot account for all of the observed warming at a given probability threshold - and to the best of our knowledge, they cannot - then the null hypothesis (that GHG increases are unnecessary to explain the observations) is rejected and we have a statistically significant result for the hypothesis that GHG increases have contributed to the warming.

Which we do. Playing with random graphs is an utterly clueless exercise which tells us nothing about the climate, only about the general intelligence of self-proclaimed 'sceptics.'
Click to expand...

If the trend of global temperatures is not significantly different from zero then changes in other factors, like CO2, don't matter.
 
Deuce said:
The positive temperature trend since 1880 is statistically significant.
Click to expand...

What I'm saying is that analyzed properly it is not.
 
LowDown said:
If the trend of global temperatures is not significantly different from zero then changes in other factors, like CO2, don't matter.
Click to expand...

It is significantly different from zero :roll: Despite considerable evidence to the contrary from your posts, I don't really believe that you are a stupid person: You do understand how absurd it is to suppose that every single time series has an identical value N, below which no trend is ever significant. You do understand that statistical significance is determined by context depending on the hypothesis in question; you do understand that climatic influences are not random; and you do understand that using random numbers as a point of comparison is hopelessly idiotic.

You are simply hoping that there are enough dupes around here that your propaganda gets through in spite of these facts.
 
Mithrae said:
It is significantly different from zero :roll: Despite considerable evidence to the contrary from your posts, I don't really believe that you are a stupid person: You do understand how absurd it is to suppose that every single time series has an identical value N, below which no trend is ever significant. You do understand that statistical significance is determined by context depending on the hypothesis in question; you do understand that climatic influences are not random; and you do understand that using random numbers as a point of comparison is hopelessly idiotic.

You are simply hoping that there are enough dupes around here that your propaganda gets through in spite of these facts.
Click to expand...
His string title before this one was "an article from WUWT".
This one was more a Stealth article from.... WUWT/WTFUWT/LOLWUWT... with the Blog source Hidden under some words.
But the two studies WUWT used/abused, OTOH, were in full view/proudly posted.
There's never been a board section so dominated by a Single website (BLOG no less), and in this case it's infamous, and an exception, and differs with All/virtually All mainstream science orgs.
 
Last edited:
LowDown said:
If the trend of global temperatures is not significantly different from zero then changes in other factors, like CO2, don't matter.
Click to expand...

So, the sun doesn't matter?
 
[h=1]More on ‘Vilifying Rose’[/h]Posted on 14 Dec 16 by JAIME JESSOP • 4 Comments
David Rose has been talking to David Whitehouse of the GWPF about the reaction to his Mail on Sunday article in which he criticised the global warming hype which we saw at the peak of El Nino warming earlier this year. He seems genuinely astounded by the reaction and one can hardly blame him. … Conti
 
Mithrae said:
It is significantly different from zero :roll: Despite considerable evidence to the contrary from your posts, I don't really believe that you are a stupid person: You do understand how absurd it is to suppose that every single time series has an identical value N, below which no trend is ever significant. You do understand that statistical significance is determined by context depending on the hypothesis in question; you do understand that climatic influences are not random; and you do understand that using random numbers as a point of comparison is hopelessly idiotic.

You are simply hoping that there are enough dupes around here that your propaganda gets through in spite of these facts.
Click to expand...

All you've got is a lot of vague hand waving about this and that. Let's see something specific.

If you can model the temperature with random numbers then the temperature is random. That is what has been amply shown.

Let's see if you can follow this argument and offer substantive comments:

The state of the weather at any given time in great part determines the next state. Therefore, randomness in weather is not relative to some baseline state but to the previous state. You don't have summer days alternating with cold winter days, for example. One winter day leads to another gradually going to spring and summer. One warm year is likely to follow another. This is the behavior of a time series.

Therefore statistical models that assume entirely random states relative to a hypothetical mean are not appropriate to model the climate.

So, if temperatures are in a time series then drift in the data can be due to randomness in the way temperatures change from one time point to the next. So the thing to look at is the change from one time point to the next. We therefore difference the data and create a vector consisting of the difference between one year (to eliminate seasonality) and the next. This eliminates the drift.

If the magnitude in increases from one year to the next is significantly larger than the decreases, then we have a positive trend, and we have proven that, in all likelihood, there has been a significant increase in global temperatures not due to random effects.

So in R, if gt is global temperatures from 1850 to the present (HADCRUT4 Global) then we get

gt.d = diff(gt) as the differenced vector.

and if t is the years, and t2 is that squared, then we do a general linear regression:

Call:
lm(formula = gt.d ~ t + t2)

Residuals:
Min 1Q Median 3Q Max
-0.85105 -0.07320 0.00014 0.07340 1.05075

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.730e-03 9.364e-03 0.185 0.853
t -3.366e-06 2.163e-05 -0.156 0.876
t2 1.732e-09 1.047e-08 0.165 0.869

Residual standard error: 0.1394 on 1996 degrees of freedom
Multiple R-squared: 1.389e-05, Adjusted R-squared: -0.0009881
F-statistic: 0.01386 on 2 and 1996 DF, p-value: 0.9862
Click to expand...

So the temperature difference from 1850 to now is not significantly different from zero since the range of the SEM includes zero. In other words, the change in temperature from 1850 to now is no different from what we could expect of random variation. This should be no surprise since we a looking at changes in the order of tenths of a degree.
 
LowDown said:
All you've got is a lot of vague hand waving about this and that. Let's see something specific.

If you can model the temperature with random numbers then the temperature is random.
Click to expand...

So you're asserting that the temperature is random :shock:

Your grasp of logic is obviously even worse than your understanding of statistics. But this goes far beyond even that, because you really shouldn't even need to recognise your specific logical flaws to see the absurdity of your conclusion. We're talking about absolutely basic facts of reality here: Global temperature trends are caused by solar variation, planetary albedo, oceanic oscillations, atmospheric composition etc. etc. This is utterly basic stuff that most high school drop-outs probably understand if you use small words and ask the right questions. But you do not.

How does it feel, to look like you're dumber than a high school drop-out? You're probably not, really, but do you honestly think you're going to persuade anyone by giving them that impression?
 
Last edited:
Mithrae said:
So you're asserting that the temperature is random :shock:

Your grasp of logic is obviously even worse than your understanding of statistics. But this goes far beyond even that, because you really shouldn't even need to recognise your specific logical flaws to see the absurdity of your conclusion. We're talking about absolutely basic facts of reality here: Global temperature trends are caused by solar variation, planetary albedo, oceanic oscillations, atmospheric composition etc. etc. This is utterly basic stuff that most high school drop-outs probably understand if you use small words and ask the right questions. But you do not.

How does it feel, to look like you're dumber than a high school drop-out? You're probably not, really, but do you honestly think you're going to persuade anyone by giving them that impression?
Click to expand...

Well, so the answer is no, you can't.

No, I'm not saying that the temperature is actually likely to be random. I'm only saying that you can't prove that it's not random with this data (HADCRUT4 Global annual or monthly means).

That being the case, then attempting to assign causality for a trend in the temperature to various factors is a fool's errand.

I don't agree with the mainstream on this. That's the point.
 
LowDown said:
Well, so the answer is no, you can't.

No, I'm not saying that the temperature is actually likely to be random.
Click to expand...

You and others have said and shown that something similar to the temperature record can be modeled with somewhat random numbers. You further said that if you can model the temperature with random numbers then the temperature is random. So you are saying that the temperature is random. You said it right from the beginning in the opening post ("natural random variation") and the theme has continued up to the present. There's no way to excuse this as simply an atrocious inability to think and communicate clearly - this is what you are claiming, or at least what you want people to believe has been proven.

The trouble is that the claim is utterly and demonstrably ridiculous.

So you are equivocating and obfuscating and trying to bull**** your way past the simple and obvious fact that global temperature variations are not random, so playing around with random time series means absolutely nothing.

It tells us nothing about the climate, it tells us nothing about greenhouse gases, and it tells us nothing about the temperature record itself.
 
You must log in or register to reply here.
Back
Top Bottom