How Much Money Will You Pay For Green Energy?

[Michael Hardner]

For starters: it is not clearly wrong. We have a set of tree rings which may or may not be a proxy for temperature. The hypothesis that the latter years are bad data is just a hypothesis. It is not a fact. Another hypothesis is the correlation before 1950 is a coincidence and these tree rings are not temperature proxies. You have no way to show that my hypothesis is invalid. Boreholes and other proxies are irrelevant when it comes to this question.

Yes - you can test it for statistical validity. Although I don't have the study at my fingertips, I'm pretty sure this has been done.

The rule of statistics is this: you can discard bad data if and only if you have independent evidence that the data is bad.

Independent evidence being the actual temperature readings.

You can never discard data simply because it does not fit into your model.

The issue isn't that the data doesn't fit the model - it's that the data doesn't match the known temperatures after a certain point.

Discarding data points because they don't fit the model would be junk science but that isn't being done in the way you describe.

I understand why this *could* be perceived as modifying data to make the model fit, but it isn't the case here.

In any case, we may need to agree to disagree on this point. I understand that further publications have happened using different datasets since then.

[Wild Bill]

Why not ? It's clearly wrong. There's no reason to not discard bad data. What law of statistics says that you can't discard bad data ?

Classic example:

An ad company wants to do an ad for Aspirin, having a man in a white lab coat state that "4 out of 5 doctors prefer Aspirin!" They poll doctors and it takes them 500 doctors to come up with 4 that prefer Aspirin. So they take those 4 and 1 of the other 496, discarding the 495 that don't prefer Aspirin.

Now they can legally make the claim that "4 out of 5 doctors prefer Aspirin!"

Michael, they simply discarded 495 opinions as "bad data". I'm afraid that's what your argument sounds like to me!

[TimG]

Yes - you can test it for statistical validity. Although I don't have the study at my fingertips, I'm pretty sure this has been done.

Now you throwing out scientific buzzwords. What you said makes no sense.

Independent evidence being the actual temperature readings.

This is getting ridiculous. The correlation with temps IS the variable being measured therefore it is NOT independent.

The issue isn't that the data doesn't fit the model - it's that the data doesn't match the known temperatures after a certain point.

The model in this case is a statistical model that tree rings are correlated with temperature. And yes, the divergence means the data does not fit model.

Discarding data points because they don't fit the model would be junk science but that isn't being done in the way you describe.

Actually, that is exactly how they did it. The only thing they added is a hand waving excuse for why the data was tossed out. Not good enough. There needs to be actual evidence that is independent of the temperatures.

[Michael Hardner]

Classic example:

An ad company wants to do an ad for Aspirin, having a man in a white lab coat state that "4 out of 5 doctors prefer Aspirin!" They poll doctors and it takes them 500 doctors to come up with 4 that prefer Aspirin. So they take those 4 and 1 of the other 496, discarding the 495 that don't prefer Aspirin.

Now they can legally make the claim that "4 out of 5 doctors prefer Aspirin!"

Michael, they simply discarded 495 opinions as "bad data". I'm afraid that's what your argument sounds like to me!

Your example is so simple that it's not helpful in this example.

What we have is decades of tree ring data that follows temperature quite nicely. Then around 1960 it stops... we don't know why. Was it pollution ? Was it warming ? Was it something else ? Who knows ?

Was it a COINCIDENCE that the temperatures were well reflected in tree ring growth up until that time ? Highly unlikely.

In any case, new studies have since been published that do not use tree-ring data largely replicating the 'hockey stick' result.

[Michael Hardner]

The model in this case is a statistical model that tree rings are correlated with temperature. And yes, the divergence means the data does not fit model.

Actually, that is exactly how they did it. The only thing they added is a hand waving excuse for why the data was tossed out. Not good enough. There needs to be actual evidence that is independent of the temperatures.

It's valid to say that there's a good correlation up until a certain date. You're correct that they don't know (yet) why divergence has happened but that's just a matter of time. I still think it makes sense to throw out data after a certain point - that's not the same as selecting individual points throughout the data to discard.

And in reading up on this, I can see that more studies have been published without tree rings that show the recent warming to be significant.

[TimG]

What we have is decades of tree ring data that follows temperature quite nicely. Then around 1960 it stops... we don't know why. Was it pollution ? Was it warming ? Was it something else ? Who knows ?

That is the problem. We don't know. And since we do not know we cannot know if the tree rings are valid temperature proxies and the data must be thrown out. I realize that scientists don't like it when they have to throw questionable data out but there is really no other option.

Was it a COINCIDENCE that the temperatures were well reflected in tree ring growth up until that time ? Highly unlikely.

Actually, coincidence highly likely. You can pick any natural phenomena you want and you will likely see a correlation with the temperature record over part of the period. In the case of tree rings it is known from studies of biology that they are not necessarily proxies for temperature so the requirement that they must correlate over the entire period is even more important.

In any case, new studies have since been published that do not use tree-ring data largely replicating the 'hockey stick' result.

The "hockey team" likes to play a game of pea under the thimble. i.e. they have a two or three bogus proxies that they keep combining into studies. When skeptics point out that one set is invalid they say the 'result was replicated without the bad proxy' by using the others. There are no studies which can reproduce the low MWP that excludes ALL of the bogus proxies. Edited August 9, 2011 by TimG

[Michael Hardner]

That is the problem. We don't know. And since we do not know we cannot know if the tree rings are valid temperature proxies and the data must be thrown out. I realize that scientists don't like it when they have to throw questionable data out but there is really no other option.

True, we don't know but we can be reasonably confident that it's not a coincidence.

Actually, coincidence highly likely.

No - highly unlikely. How else can you explain close correlation for hundreds of years, followed by divergence ? It's not like there was no correlation throughout - there was.

The "hockey team" likes to play a game of pea under the thimble. i.e. they have a two or three bogus proxies that they keep combining into studies. When skeptics point out that one set is invalid they say the 'result was replicated without the bad proxy' by using the others. There are no studies which can reproduce the low MWP that excludes ALL of the bogus proxies.

I just read on the wikipedia page that a study was done using Arctic ice samples alone. Leaving that aside, there have since been reviews of the Mann/Bradley/Hughes papers that validated it.

[TimG]

No - highly unlikely. How else can you explain close correlation for hundreds of years, followed by divergence ? It's not like there was no correlation throughout - there was.

It is not hundreds of years. You have to correlate tree rings with the local temperatures and the local temperatures only go back to the 1930s. So you are dealing with a situation where you are throwing away more data than you are keeping.

I just read on the wikipedia page that a study was done using Arctic ice samples alone.

Does not go back to the MWP.

Leaving that aside, there have since been reviews of the Mann/Bradley/Hughes papers that validated it.

We are talking about the Briffa Yamal Series - one of the papers that supposedly "validate" MBH. You are getting tricked by the "hockey team" and the pea under the thimble. Edited August 9, 2011 by TimG

[Michael Hardner]

It is not hundreds of years. You have to correlate tree rings with the local temperatures and the local temperatures only go back to the 1930s. So you are dealing with a situation where you are throwing away more data than you are keeping.

Sorry - you're right there. They correlate to other data before that.

Does not go back to the MWP.

Yes - it said 2000 years.

We are talking about the Briffa Yamal Series - one of the papers that supposedly "validate" MBH. You are getting tricked by the "hockey team" and the pea under the thimble.

The Arctic ice data is presumably a different series ?

[TimG]

Yes - it said 2000 years.

Cite?

The Arctic ice data is presumably a different series ?

It would not change the fact that junk science that gets the right answer is still junk science. Scientists who defend junk science are not particularily credible in my view. Edited August 9, 2011 by TimG

[Michael Hardner]

Cite?

http://www2.ucar.edu/news/846/arctic-warming-overtakes-2000-years-natural-cooling

Sorry - this wasn't done using ice measurements alone - it appears they used previously published tree ring data to plot past temperatures.

[waldo]

The other paper by MM

is just garbage

…I can’t see either of these papers being in the next IPCC report. Kevin [Trenberth] and I will keep them out somehow – even if we have to redefine what the peer-review literature is !

I think the key to this controversy is in the first sentence. Whoever stated this was speaking privately - and notably they did not say something along the lines of "we're being found out" they said "this paper is just garbage". That is a candid, though admittedly unprofessional assessment that informs us of the mindset of the reviewers.

TimG is still spinning his denial... some things never change. That McKitrickMichaels2004 paper was garbage... literally. Of course, TimG's automatic go-to, the never-ending McIntyre audit, somehow never managed to catch McKitrick's colossal screw-up (degrees vs. radians)... something McIntyre conveniently leaves out in the TimG linked reference. Accordingly, all MM2004 calculations were wrong. How did "auditor" McIntyre miss such a glaring error? Without even checking the rationale MM2004 used, others examining the paper corrected the error and re-ran the paper's regressions... with a net result that showed the warming trend reduced from MM2004's 0.27 degrees/decade, to 0.18 degrees/decade... which approximates the actual warming global trend. Strange that McIntyre never acknowledges any of this.

There also is no shortage of critical review behind the described rationale of the MM2004 paper... e.g. Benestad, R.E. (2004). Are temperature trends affected by economic activity? Comment on McKitrick & Michaels.... with the money shot:

There is therefore no evidence suggesting that the temperature trends are systematically influenced by non-climatic factors.

but... really... this is all so yesterday. McKitrick and Michael tried another kick at this same can in 2007. Again, there is no shortage of critical review of MM2007... coincidentally, none of that critical review originates from TimG's go-to "auditor", McIntyre. Go figure!

[TimG]

Sorry - this wasn't done using ice measurements alone - it appears they used previously published tree ring data to plot past temperatures.

Watch the pea under the thimble. The wiki article was written to mislead you into believing it was a no tree ring proxy study, wasn't it? Edited August 9, 2011 by TimG

[Michael Hardner]

Watch the pea under the thimble. The wiki article was written to mislead you into believing it was a no tree ring proxy study, wasn't it?

No, I just missed it.

Without the tree ring study, we're still the warmest period SINCE the MWP though correct?

[Michael Hardner]

Again, there is no shortage of critical review of MM2007... coincidentally, none of that critical review originates from TimG's go-to "auditor", McIntyre. Go figure!

Waldo returns !

I haven't mentioned the fact that MM have been quite heavily criticized as well, no, since we're not talking about their paper but about their criticism of other theories.

[waldo]

Tree ring temperatures align to those temperatures very well up to a point - and those are correlated with other temperatures from other sources.

I'm reading several "pea under the thimble" statements... Michael... you should ask TimG if he's arguing for a warmer MWP period. Ask him if he's invoking evidence for strong(er) climate sensitivity... and net positive feedback. Ask him if he accepts that his arguing for a warmer MWP means he's accepting to a higher climate sensitivity - implying a larger climate response to CO2 forcing. Ask him that, hey?

[TimG]

There also is no shortage of critical review behind the described rationale of the MM2004 paper

The issue is not whether the some people are of the opinion (and it is only an *opinion*) that the MM paper has no merit. The issue is how Jones put unreviewed editorial opinion into the IPCC report and then tried to prevent people from finding out how he subverted the IPCC process by deleting emails. Edited August 9, 2011 by TimG

[Michael Hardner]

Ask him that, hey?

I can't "proxy" your arguments, Waldo... heh heh...

[TimG]

Without the tree ring study, we're still the warmest period SINCE the MWP though correct?

We really have no data that would allow us to draw any comparisons between the MWP and today. All we know for sure is the MWP was a warm period where some proxies show it to be as warm as today but we don't have any proxy data with sufficient time and spatial resolution to make a claim about global temps.

The reality is that sometimes: "we don't know" is the only reasonable scientific answer.

Edited August 9, 2011 by TimG

[Michael Hardner]

The issue is not whether the some people are of the opinion (and it is only an *opinion*) that the MM paper has no merit. The issue is how Jones put unreviewed editorial opinion into the report and then tried to prevent people from finding out how he subverted the IPCC process by deleting emails.

Fair enough - but again *most* people wouldn't say that this is what Climategate was about.

WB seems to think that scientists faked data, faked studies and people swallowed it. In fact, there is a group of scientists who have a problem with some of the data - but most don't agree with their solution to suppress that data completely. And even if it is suppressed, we're in a warming period in the 20th century.

And now, your description of the issue is that Jones included some unsupported comments as part of his many comments into the report, and deleted emails where he (presumably) insulted the authors of the paper that he thought was "garbage".

Climategate seems to be about the MM paper getting more of a fair hearing than it did (leaving aside how fair that process was) and not about the faked data that we heard about on day 1 from Rush L., Alex J. and the rest.

[TimG]

Fair enough - but again *most* people wouldn't say that this is what Climategate was about.

There are several issues brought out by climategate. The two major ones are the subversion of the IPCC process by insiders and the deleting adverse data from graphs that went into policy reports. The latter is what leads to the accusation of faked science.

Ironically, the cover up is always worse than the crime. If the climate establishment had simply come clean and acknowledged that deleting the adverse data from the graphs was deceptive then we would not be still talking about it. By defending Mann/Jones and the set of clowns that surround them the science establish has shown that it is part of the problem.

Edited August 9, 2011 by TimG

[Michael Hardner]

We really have no data that would allow us to draw any comparisons between the MWP and today. All we know for sure is the MWP was a warm period where some proxies show it to be as warm as today but we don't have any proxy data with sufficient time and spatial resolution to make a claim about global temps.

The reality is that sometimes: "we don't know" is the only reasonable scientific answer.

I haven't heard that before. I thought there were datasets that excluded tree ring data, but that they showed the recent warming to be (something like) the warmest in 700 vs 1000 years.

I think I found a study here:

http://www.cce-review.org/evidence/MalcolmKHughes.pdf

. Our reconstructions made using tree ring

records, including those also used by Professor Briffa, are very largely similar to those

we made excluding tree ring data.

[Michael Hardner]

There are several issues brought out by climategate. The two major ones are the subversion of the IPCC process by insiders and the deleting adverse data from graphs that went into policy reports. The latter is what leads to the accusation of faked science.

Alleged subversion, I think - to be fair. The deletion of emails shows an unwillingness to comply with FOIA but doesn't show subversion.

'adverse data' means 'incorrect data' in the context above.

Ironically, the cover up is always worse than the crime. If the climate establishment had simply come clean and acknowledged that deleting the adverse data from the graphs was deceptive then we would not be still talking about it. By defending Mann/Jones and the set of clowns that surround them the science establish has shown that it is part of the problem.

They were open about the fact that there were problems with proxy data, and this was known. There was no cover up over the 'travesty that they couldn't show warming' - they had written about that problem.

[TimG]

I haven't heard that before. I thought there were datasets that excluded tree ring data, but that they showed the recent warming to be (something like) the warmest in 700 vs 1000 years.

Again, they are moving the pea under the thimble. It is true that removing the tree rings will often yield a similar shape but the resulting reconstruction is not statistically significant over periods prior to 1500. A paper with only the no-tree ring results would not get past peer review because of this which is why they insist on using the tree rings (think about it: it would be alot easier to stop using tree rings but they dont because they need them to get their papers accepted). Edited August 9, 2011 by TimG

[Michael Hardner]

Again, they are moving the pea under the thimble. It is true that removing the tree rings will often yield a similar shape but the resulting reconstruction is not statistically significant over periods prior to 1500. A paper with only the no-tree ring results would not get past peer review because of this which is why they insist on using the tree rings (think about it: it would be alot easier to stop using tree rings but they dont because they need them to get their papers accepted).

Wasn't the aforementioned study published though ?