The Holocene Deniers

[grmorton]

Then why the whole song and dance number of how your data showed you that Siberia wasn't melting? Are you able to hold a discussion without resorting to straw men and red herrings all the time? You argue just like a YEC, and I'm not surprised to hear you used to be one.

There is a difference between what you all should expect--massive measurable temperature rise in the Arctic and having some warming melting some permafrost. If the world's temperature has risen according to what the IPCC says, ,we should most clearly see it in the Arctic regions. We don't. That doesn't mean either that the world hasn't warmed, nor does it mean that the permafrost isn't moving north right now. What it does mean is that what should be the case if your catastrophist scenario is true, isn't true.

I've already told you why I'm "missing the point". Your point is COMPLETELY IRRELEVANT. I don't know how many other ways I can tell you that. The earth warming in the past through natural causes has absolutely NOTHING to do with the earth warming today through unnatural causes. As so many people have already told you, your entire OP is a huge straw man, as well as irrelevant.

I just love the loves and kisses from you.

For science's sake, check your facts mr Morton.

Does Permafrost Affect Trees? | eHow.com

Sigh, what part of ABOVE do you not understand. I said that trees can't live IN permafrost. You point me to a site which talks about trees living above the permafrost--in guess what--dirt that ISN'T permanently frozen.

"Survival is more difficult for trees that grow above permafrost, though several species have adapted to the conditions. In ground where the permafrost layer begins closer to the surface, only black spruce trees can survive"

I thought you might have missed that above.

http://en.wikipedia.org/wiki/Drunken_trees




Wow, more peer reviewed sources from the guy who always tells me to use peer reviewed literature. Trees dont' live in permafrost
If you are basing those assertions on your previous one about trees not growing on permafrost, they are completely unfounded. Evidence needed.

Think again, Billion Barrel Morton.

As a moniker it lacks originality--has something of a copycat persona, and lacks the alliteration of the original.

 

[grmorton]

I like how AGW deniers think that they can use 1 years information (or less) to document a century-spanning trend.

Given that you snipped whatever caused you to make that comment it is difficult to reply to. Seems that you dont like data very much as you ditch it before making comments.

 

[Contracelsus]

That is an interesting graph. I don't see the same info in it as you do. According to the way I read the graph, Scenario B says we should have had about .8 deg C temperature rise since 1960 up through about 2007.

I was relying on the actual points on the graph presented itself. I believe those only go up to 2005? I don't know for sure.

But it looks to me like the observed data is reasonably close to the predicted data over more than a decade.

And Hansen clearly outlines in his paper the forcing functions he used, so his model must be onto something.

Perhaps you could show me on the graph specifically where the model has broken down.

There is also something else that needs to be considered. Look at your chart at 2005. Scenario B predicted that there would be a .75 deg C rise in temperature since 1960. But that conflicts with the IPCC claim that over the past 100 years we have had an increase in temperaure of .74 deg C. ( see page 30 of http://www.ipcc.ch/pdf/assessment-report/ar4/syr/ar4_syr.pdf)

It seems that according to this criterion, Hansen's Scenario B is off by a factor of 2. I don't know, maybe a factor of 2 is also to be considered 'on target'. I wouldn't think so but if you do, please explain why.

I don't recall mentioning anything here about the IPCC report. I was addressing the issues raised earlier about the climate forcings, specifically CO2 (but clearly Hansen's 1988 paper addressed other forcings as well).

If a scientist in 1988 predicts a trend and over the next 12 years or so the trend of actual observed data agrees within some error with the predicted trend I don't see how the model is flawed.

Maybe I'm missing something. Maybe we should just, for the time being, focus on one graph, one model, one set of predictions in order to address the data at hand.

 

[grmorton]

If I might jump in here for a second. In relation to the two Iowa stations, I have made a histogram of the differences between the two stations:

Now, this is what is called a "heavy tail distribution". It has a large number of values that fall far out into the tails. Sure this is not perfect data, but then what is? Look at the histogram bars close in to 0. These two bars account for almost 15,000 data points.

In fact, assuming I counted right, the % of data measurements greater than +20deg difference is about 2% of the data. That's all. Remember this data set represents about 20,960 data points. This is daily data that goes back, what, about 60 years?

In fact the median is -1deg. That doesn't look too bad. And if I understand the debate correctly the absolute temperature isn't what is important, it's the trend in temperature changes.

So if the difference between these two is only 1degree F over 60 years that doesn't seem like a bad data set. The fact that the histogram is reasonably symmetrical would seem to indicate a lack of systematic error.

I've looked a few other of these couples that Mr. Morton has listed on his Blog. Here's the histograms:

Two stations; in in Iowa and one in Illinois.

Two stations close to each other in Georgia. The median difference here is about 0 deg F. Now this one has 3 days out of about 57 years worth of data where there was a -60degree difference.

3 data points out of 20,515 data points? That's 0.01% of the data.

Here's two stations close to each other, one in Mississippi and one in Louisiana that Mr. Morton mentioned in his blog.

While there's clearly some errors in the data and outliers (which can be dealt with rather easily) this would seem to indicate that various couplets of stations are not necessarily systematically flawed in some way. Certainly not enough to indicate that calculations of TRENDS based off the temperature data taken across the country would necessarily be erroneous.

Finally someone who actually likes to analyze data rather than call names. I have, like you done the same thing with many datasets, I can't remember if I did it on this one or not.

To me, median difference doesn't capture the problem. Lets take two towns which one day Town A is 20 degrees higher than Town B. But then, the next day Town B is 20 degrees higher than Town A. Averaging the differences means that the average problem is zero.


I just looked up my Belle Plaine and Toledo spread sheet. I had not previously run a histogram on this data set. So, I decided to look a bit at the distribution. Contracelsus will be thrilled, I know, just thrilled, that I agree that the number of days above 20 deg discrepancy is very small. But unlike him, I didn't stop there and I looked at the number of days which were different by 10 degrees and by 5 degrees. And I split it out by whether or not it was hotter in Belle Plain (BP) or Toledo (T).

BP>T by 20 deg..... 11
T >BP by 20 deg.... 29

BP>T by 10 deg......436
T>BP by 10 deg......167

BP>T by 5 deg.......2035
T> BP by 5 deg.......878

The first thing to notice is that there is a bias towards Belle Plain, which was seen on my graph of yesterday.

Now lets look a bit deeper at the 5 deg or more difference. Just shy of 10% of the days are warmer in Belle Plaine by 5 degrees or more. Just over 4% of the days are warmer in Toledo by 5 degrees or more.

That means that the total number of days greater than 5 degrees difference amounts to an amazingly large 15% of the days.

Now, given that if you go look up their 5 day forecasts they are always within 1 degree of each other, often being the same. So what should we allow for the acceptable error between these two towns. Certainly having 15 percent of the days being greater than 5 deg different should be a cause of concern, I would think.

I then decided that we could accept a 3 degree interval (even though such a temperature gradient is that of a very large and strong cold front's. So, at 3 degrees or greater we have a temperature gradient that should yield thunderstorms as strong cold fronts.

Here are the results of that

BP>T by 3 deg ....3588 days--17% of the days
T>BP by 3 deg.....1806 days--8.5% of the days

Again a clear bias in which 25% of the days have temperature gradients above that of the strong cold front--remember I used 3 as the cut off. 4 deg F/over 18 miles gives a result higher than all but the worst of cold fronts.

A　suggested　set　of　criteria　based　on　the　horizontal temperature　gradient　has　been　devised.　A　weak　front　is one　　where　　the　　temperature　　gradient　　is　　less　　than　　10[deg]F per　　　　100　　　　miles;　　　　a　　　　moderate　　　　front　　　　is　　　　where　　　　the temperature　gradient　is　10　[deg]F　to　20　[deg]F　per　100　miles; and　a　strong　front　is　where　the　gradient　is　over　20　deg]F　per 100 miles.
THE COLD FRONT

In order to believe this data set and beleive that there is no problem, one must believe that 25% of the days have a strong weather front BETWEEN the two towns.

Do you believe this Contracelsus? I don't

 

[grmorton]

I was relying on the actual points on the graph presented itself. I believe those only go up to 2005? I don't know for sure.

But it looks to me like the observed data is reasonably close to the predicted data over more than a decade.

And Hansen clearly outlines in his paper the forcing functions he used, so his model must be onto something.

Perhaps you could show me on the graph specifically where the model has broken down.

I think I know what the problem is. Hansen's observed data since 1960 seems to be a bit on the high side. He has about a .68 deg C rise in temperature since 1960 as the measured data. Global climate at a glance doesn't go higher than about half a degree.

I never take Hansen's word for much so I compare what he says with other data sets put out by NOAA, which is what I did.

I don't recall mentioning anything here about the IPCC report. I was addressing the issues raised earlier about the climate forcings, specifically CO2 (but clearly Hansen's 1988 paper addressed other forcings as well).

I know you didn't mention IPCC but I don't think it is rude of me to mention it, especially since Hansen's 'observed' temperature rise seems to be a huge chunk of the claimed warming. Why shouldn't I compare Hansen's prediction with what the IPCC says? Things ought to fit together.

If a scientist in 1988 predicts a trend and over the next 12 years or so the trend of actual observed data agrees within some error with the predicted trend I don't see how the model is flawed.

What is the error? What is the acceptable error? You didnt specify what level would make you feel that he missed the mark. That is a problem with post-facto claims of success.

Maybe I'm missing something. Maybe we should just, for the time being, focus on one graph, one model, one set of predictions in order to address the data at hand.

I never focus on just one thing. I have learned that if you are going to find oil, one must not focus on just one thing. All facts must fit together facts from all sorts of sources. All of it must be woven together into an understanding of what is going on. To me, focusing on one thing means you miss the problems.

 

[Contracelsus]

I just looked up my Belle Plaine and Toledo spread sheet. I had not previously run a histogram on this data set. So, I decided to look a bit at the distribution. Contracelsus will be thrilled, I know, just thrilled, that I agree that the number of days above 20 deg discrepancy is very small. But unlike him, I didn't stop there and I looked at the number of days which were different by 10 degrees and by 5 degrees. And I split it out by whether or not it was hotter in Belle Plain (BP) or Toledo (T).

BP>T by 20 deg..... 11
T >BP by 20 deg.... 29

If I do a chi square test based on these two data points alone I get a x[sup]2[/sup] = 8.1, a df = 1 for this table, and f[sub]e[/sub] = 20 (calculated) which, when I look it up at 95% confidence there's no statistical difference between those two.

BP>T by 10 deg......436
T>BP by 10 deg......167

Another chi square test shows the chi sqare value = 120 with 1 df which means that it has a p-value = 2.2x10[sup]-16[/sup]. Have to say those are not statistically different.

BP>T by 5 deg.......2035
T> BP by 5 deg.......878

Chi square = 459.5, df = 1, p = 2.2x10[sup]-16[/sup]

The first thing to notice is that there is a bias towards Belle Plain, which was seen on my graph of yesterday.

I am unsure if that bias is proven out by these three examples. The data look pretty evenly spread around the median. It looks like there may not be a systematic bias down to 5 degrees difference.

In order to believe this data set and beleive that there is no problem, one must believe that 25% of the days have a strong weather front BETWEEN the two towns.

Do you believe this Contracelsus? I don't

That's why we're running the stats, right?

Not sure I see a systematic bias here. Certainly not by the stats, unless I'm running the numbers incorrectly (which is possible. I don't do any of this stuff for a living).

 

[grmorton]

Contracelsus mentioned the two towns across the Missisippi River from each other, St. Joseph, LA (SJ) and Port Gibson, MS (PG) I pulled up my spreadsheet on these two towns. I performed the same analysis I did above for Belle Plaine and Toledo IA.

SJ>PG by 20 deg or more....686 days...2.4%
PG>SJ by 20 deg or more......2 days...0.007%

SJ>PG by 10 deg or more....895 days...3.2%
PG>SJ by 10 deg or more.....38 days...0.13%

But now, we will look at how many days are greater than 5 deg F. this was again chosen because that creates a 0.2 deg per mile gradient limit on what should be acceptable because a .2 deg per mile gradient is a very very strong weather front.

SJ>PG by 5 deg or more....3252...11.7%
PG>SG by 5 deg or more.....494....1.8%

This is clearly a better dataset than Belle Plaine/Toledo but even so, to believe that there is no problem with this data set one must believe that there are strong cold fronts between the two towns 13.5% of the time. That seems a little excessive to me.

In the chart below I run a 365-day running average over the data and you can see that the bias changes from one town to the next for years at a time.And you can see that there are problems at the later years of the time series.

I think this data is pretty bad and not at all usable for teasing out tiny increases in global warming. Shoot, the corrections that must be made to these data sets are greater in magnitude than a century's worth of global warming.

 

[grmorton]

If I do a chi square test based on these two data points alone I get a x[sup]2[/sup] = 8.1, a df = 1 for this table, and f[sub]e[/sub] = 20 (calculated) which, when I look it up at 95% confidence there's no statistical difference between those two.




Another chi square test shows the chi sqare value = 120 with 1 df which means that it has a p-value = 2.2x10[sup]-16[/sup]. Have to say those are not statistically different.



Chi square = 459.5, df = 1, p = 2.2x10[sup]-16[/sup]



I am unsure if that bias is proven out by these three examples. The data look pretty evenly spread around the median. It looks like there may not be a systematic bias down to 5 degrees difference.



That's why we're running the stats, right?

Not sure I see a systematic bias here. Certainly not by the stats, unless I'm running the numbers incorrectly (which is possible. I don't do any of this stuff for a living).

Did you simply count them? Did you look at the 365 day running average? Remember one thing I am doing is subtracting the values before doing anything else. You also have to be sure that you have temporally aligned the data--just checking here.

Since one can actually SEE the bias, I am surprised by your results. But I would caution you that bias isn't the only problem here and to focus only on that would be to miss the point here.

I am substracting two very closely spaced data sets. That is as close as we can get to repeating a temperature measurement. We can't go back to 1952 and repeat the temperature measurements to see what the real error in the data is. But we can compare two towns very close together to see how big is their difference.

I averaged the difference in temperatures. I got 0.83 deg F for the average of the daily temperature differences of Belle Plaine and Toledo. There is on average a degrees worth of bias, but it varies over time--you can see it in the running average. In any electrical or seismic signal, the average is the DC shift--that is, the bias.

I did a standard deviation on the temperature differences and got 4.07.

Thus, if one looks at these two temperature streams as repeated attempts to measure the same thing--the temperature of the area, then the temperature is accurate to only within 4 degrees. That seems to be a problem as far as I am concerned.

And you didn't answer my question. Do you believe that for 25% of the time there was a real temperature difference of greater than 3 degrees, which would create the same temperature gradient (deg/mile) as a very strong cold front? Do you believe that?

I really do want an answer to that

edited to add. I guess we better be sure we are getting our data from the same place. This is where I am getting mine http://cdiac.ornl.gov/epubs/ndp/ushcn/usa.html

 

[grmorton]

If I do a chi square test based on these two data points alone I get a x[sup]2[/sup] = 8.1, a df = 1 for this table, and f[sub]e[/sub] = 20 (calculated) which, when I look it up at 95% confidence there's no statistical difference between those two.




Another chi square test shows the chi sqare value = 120 with 1 df which means that it has a p-value = 2.2x10[sup]-16[/sup]. Have to say those are not statistically different.



Chi square = 459.5, df = 1, p = 2.2x10[sup]-16[/sup]



I am unsure if that bias is proven out by these three examples. The data look pretty evenly spread around the median. It looks like there may not be a systematic bias down to 5 degrees difference.



That's why we're running the stats, right?

Not sure I see a systematic bias here. Certainly not by the stats, unless I'm running the numbers incorrectly (which is possible. I don't do any of this stuff for a living).

I didn't hardly sleep last night I should have thought to ask this before. In your chi squared test what is your expected theoretical frequency distribution? Where did you get it? Are you merely testing for Gaussianicity? If so, I fail to see how one can have your result and my count at the same time. We need to get to the bottom of it.

Are you running this on each station alone? Are you running it on the difference between the two towns, which is what I am doing?

 

[Baggins]

Once again Mr Morton doesn't want to address the political reasons for his rejection of the work of those far better qualified than him in climatology.

He may have me on ignore now, perhaps someone else would like to bring this up?

 

[Contracelsus]

Did you simply count them? Did you look at the 365 day running average? Remember one thing I am doing is subtracting the values before doing anything else. You also have to be sure that you have temporally aligned the data--just checking here.

I just used the counts you provided. In each case I set up a table:

f[sub]o[/sub], f[sub]e[/sub] and ran the chi square test.

The f[sub]e[/sub] were calculated (as from the stats textbook I have here) by merely summing the two f[sub]o[/sub] values and dividing by two.

So for instance in the 5 degrees sample I set up the table like this:

f[sub]o[/sub] f[sub]e[/sub]
11 20
29 20

The x[sup]2[/sup] is calculated as usual as the sum of the two

(f[sub]o[/sub] - f[sub]e[/sub])[sup]2[/sup]/f[sub]e[/sub]

The test is an attempt to see if in this number of "tests" is there really a statistical difference in the number of days T>BP and BP>T. Would we expect to see this difference by random chance alone or is there a directionality to the bias.

So what do I believe? If I'm doing the stats right here then it looks to me like the counts you provided don't give me enough reason to say there is a directional bias at 20deg, 10deg or even down at 5deg.

But like I said, I could be doing this wrong. I am not a scientist. I don't do this stuff for a living.

Since one can actually SEE the bias, I am surprised by your results.

In the two Iowa stations (Belle Plain and Toledo) I don't see the bias in the histogram of differences between readings. (And I have attempted to make sure they line up and I've subtracted only days that are the same)

My fascination is that, even when we think we see a bias sometimes that isn't a real bias. That's why statistics are fascinating things.

But I would caution you that bias isn't the only problem here and to focus only on that would be to miss the point here.

I was addressing your description of the bias between stations in Post 164.

 

[Thistlethorn]

There is a difference between what you all should expect--massive measurable temperature rise in the Arctic and having some warming melting some permafrost. If the world's temperature has risen according to what the IPCC says, ,we should most clearly see it in the Arctic regions. We don't. That doesn't mean either that the world hasn't warmed, nor does it mean that the permafrost isn't moving north right now. What it does mean is that what should be the case if your catastrophist scenario is true, isn't true.

We do see it it the arctic regions.

I just love the loves and kisses from you.

When you stop it with your straw men, red herrings and outright lies, I'll stop treating you harshly.

Sigh, what part of ABOVE do you not understand. I said that trees can't live IN permafrost. You point me to a site which talks about trees living above the permafrost--in guess what--dirt that ISN'T permanently frozen.

You do understand that an area with permafrost can still have a layer of top-soil that isn't frozen, right? So, your assertion that trees don't live in permafrost is just wrong. Can't you admit a mistake?

"Survival is more difficult for trees that grow above permafrost, though several species have adapted to the conditions. In ground where the permafrost layer begins closer to the surface, only black spruce trees can survive"

I thought you might have missed that above.

I didn't. I thought you would understand what that meant. Apparently you didn't.

Wow, more peer reviewed sources from the guy who always tells me to use peer reviewed literature. Trees dont' live in permafrost

If you want me to find peer-reviewed literature that says the same things that those sites say, I can certainly oblige. Wikipedia has this nice sourcing system after all. And, you're still wrong. Trees live in permafrost. They have even adapted to permafrost conditions by having a wide but shallow root system instead of a deep one. It's all there in those non-peer-reviewed links.

If you are basing those assertions on your previous one about trees not growing on permafrost, they are completely unfounded. Evidence needed.

Clever but wrong. Trees do live in permafrost. I live in a northern country. Areas of my country are permafrost areas. Those areas have trees.

As a moniker it lacks originality--has something of a copycat persona, and lacks the alliteration of the original.

Yes, but I can't help that your last name is Morton and not Borton. On the plus side, it's based on truth, not a lie like yours was.

 

[Thistlethorn]

Finally someone who actually likes to analyze data rather than call names.

I think our collective irony-meters just exploded.

 

[Thistlethorn]

Once again Mr Morton doesn't want to address the political reasons for his rejection of the work of those far better qualified than him in climatology.

He may have me on ignore now, perhaps someone else would like to bring this up?

I'll quote you for him, Baggins. He's still responding to me.

 

[grmorton]

I just used the counts you provided. In each case I set up a table:

f[sub]o[/sub], f[sub]e[/sub] and ran the chi square test.

The f[sub]e[/sub] were calculated (as from the stats textbook I have here) by merely summing the two f[sub]o[/sub] values and dividing by two.

So for instance in the 5 degrees sample I set up the table like this:

f[sub]o[/sub] f[sub]e[/sub]
11 20
29 20

The x[sup]2[/sup] is calculated as usual as the sum of the two

(f[sub]o[/sub] - f[sub]e[/sub])[sup]2[/sup]/f[sub]e[/sub]

The test is an attempt to see if in this number of "tests" is there really a statistical difference in the number of days T>BP and BP>T. Would we expect to see this difference by random chance alone or is there a directionality to the bias.

So what do I believe? If I'm doing the stats right here then it looks to me like the counts you provided don't give me enough reason to say there is a directional bias at 20deg, 10deg or even down at 5deg.

But like I said, I could be doing this wrong. I am not a scientist. I don't do this stuff for a living.

In the two Iowa stations (Belle Plain and Toledo) I don't see the bias in the histogram of differences between readings. (And I have attempted to make sure they line up and I've subtracted only days that are the same)

My fascination is that, even when we think we see a bias sometimes that isn't a real bias. That's why statistics are fascinating things.

I was addressing your description of the bias between stations in Post 164.

As to bias, one can get that merely by averaging the difference in temperatures. If the two closely spaced towns measured the exact same temperature every day, then the bias would be zero, that is, the average of the series would be zero because there would be no temperature difference on any day in the series. This would also be the case if the temperature difference between the towns had a symmetrical histogram. But, given that the average is NOT zero, it says that one town is hotter than the other. I don't see how you can avoid that conclusion.

Physics comes into this at some point. Which is why once again, I am going to ask you the quesiton you have not answered. Do you believe that it is reasonable for 25% of the days to have temperature gradients between these two towns greater than a strong weather front and yet not have strong winds and thunderstorms which are associated with the temperature gradient.

Temperature gradients are the physical basis for the winds. If those temperature differences are real we should see a strong correlation between the temperature differences and the wind direction. I would bet a steak dinner that that won't be the case.

Finished for today. Gonna watch a movie.

 

[grmorton]

I had written: "
Finally someone who actually likes to analyze data rather than call names.
"
Thistle replied:

I think our collective irony-meters just exploded.

Do you have anything to say about the data? If not, my comment stands.

 

[grmorton]

Before starting the movie, I thought I would show everyone the inferred average temperatures of various places on the globe during the Paleocene-Eocene Thermal Maximum--a time of extreme warmth which didn't do anything to the earth in the way of destroying it.

Of course everyone today is scared to death of a very minor rise in temperature compared with this time.

Night, Movie starting.

 

[Contracelsus]

As to bias, one can get that merely by averaging the difference in temperatures. If the two closely spaced towns measured the exact same temperature every day, then the bias would be zero, that is, the average of the series would be zero

The median is 1degree F. That isn't that bad.

The histogram appears quite symmetrical indicating no directionality for bias.

The chi squared tests on the 20degrees, 10 degrees and 5 degrees differences also indicate no bias (there does not appear to be reason to believe that there is a significant difference in the number of days Belle Plain is hotter than Toledo versus when Toledo is hotter than Belle Plaine.

If I have miscalculated the chi square statistic please tell me. Because as it stands now I do not see that there is any reason to claim a statistically significant bias above 5 degrees F difference.

because there would be no temperature difference on any day in the series. This would also be the case if the temperature difference between the towns had a symmetrical histogram. But, given that the average is NOT zero, it says that one town is hotter than the other. I don't see how you can avoid that conclusion.

As I said the histogram of differences looks quite symmetrical to me.

Physics comes into this at some point. Which is why once again, I am going to ask you the quesiton you have not answered. Do you believe that it is reasonable for 25% of the days to have temperature gradients between these two towns

But I don't see there being a systematic bias. There are going to be errors. I've got two thermometers in my back yard that are about 3 feet apart and they often read differently. (But there is a systematic error in them and one of them, when it is above 90deg F will shoot up to 100deg F while the other stays around 95deg F.)

A systematic bias, even if it existed, would still have to affect the TRENDS calculated based on these. If one consistently read 3 degrees higher than the other yet it still tracked with the other (ie showed an increase at the same rate as the other) it would still be effective for the present discussion.

It is like an "offset" in a measurement instrument. If I were to move the little wheel on my bathroom scale so it consistently read 5 lbs heavier than the actual weight and then I proceeded to gain weight I could still tell you how much weight I gained, if not the absolute weight I was.

 

[Contracelsus]

Physics comes into this at some point. Which is why once again, I am going to ask you the quesiton you have not answered. Do you believe that it is reasonable for 25% of the days to have temperature gradients between these two towns greater than a strong weather front

I am not as concerned by 25% at 3 degree difference between stations. When I look at the histogram I see a rather tight histogram centered around 1 deg F difference and that is from over 20,000 observations taken over the course of 60 years. I don't see evidence for a massive systematic bias of much more than 1 deg F and the fact that the tails are so low and tail off so fast impresses me even more.

I am really impressed that people going back more than a human generation were able to collect data with changing technology, hiring who knows who to do the job, and still wind up with that kind of distribution.

Again, I'm also not convinced there is an effect that would hide a trend which is what is ultimately used in the data.

If there is some reason to believe a trend is somehow hidden I'd be very interested in seeing that.

It is even more impressive with the stations that show a median of 0 degrees difference as I pointed out in one of the earlier posts.

If this is as bad as the data gets I'm afraid I can't get worked up to throw it all out.

 

[Baggins]

Before starting the movie, I thought I would show everyone the inferred average temperatures of various places on the globe during the Paleocene-Eocene Thermal Maximum--a time of extreme warmth which didn't do anything to the earth in the way of destroying it.

Straw man. No one is claiming that AGW or climate change will destroy the Earth.

Of course everyone today is scared to death of a very minor rise in temperature compared with this time.

Human civilisation didn't exist to be adversely affected in the Eocene.

People aren't worried about this increase in temperature in relation to past ones they are worried about how it will impact human society.

I find it hard to understand why a seemingly intelligent man would post such rot.

He almost certainly knows it is rot as well.