Global warming--the Data, and serious debate

[thaumaturgy]

NOAA (the National Oceanic and Atmospheric Administration) has presented this page (HERE) to describe the background to gathering temperature data.

Of note:

The observations come from the U.S. Historical Climatology Network (USHCN), a network of 1221 climate observing stations in the continental United States. These data are extensively quality controlled for errors and for small biases that may have occurred through time due to artificial changes at each observing station. These artificial changes include station relocations, different instrumentation, and changes in the landscape surrounding the station (Ibid)

(emphasis added)

So much for any proof that it is incompetent bureaucrats who allow bad data to just pour in.

But in case Glenn would still like to hold onto fears about the quality of data moving forward:

NOAA continues to work to improve the quality and representativeness of climate data provided to the public and scientific communities. In addition to advanced quality control procedures, these efforts include modernization of the USHCN by installing new, more accurate instrumentation, and ensuring proper station siting in the process. In addition, by the end of next year NOAA should have in place a U.S. Climate Reference Network, a set of 114 very high quality stations optimized for monitoring climate. The operation of the US Climate Reference Network will virtually eliminate the need for the types of corrections that have to be applied to data available today. The modernization of the US Historical Climatology Network will enable trends of regional temperature to be estimated with far fewer data corrections.(ibid)

emphasis added.

 

[thaumaturgy]

To my knowledge, most of the data around global warming trends are viewed as Temperature Anomalies, in other words annual temperatures subtracted from a multi-year mean temperature.

Indeed, NASA itself explain the use of anomalies as opposed to "absolute temperatures"

Our analysis concerns only temperature anomalies, not absolute temperature. Temperature anomalies are computed relative to the base period 1951-1980. The reason to work with anomalies, rather than absolute temperature is that absolute temperature varies markedly in short distances, while monthly or annual temperature anomalies are representative of a much larger region. Indeed, we have shown (Hansen and Lebedeff, 1987) that temperature anomalies are strongly correlated out to distances of the order of 1000 km.(SOURCE)

emphasis added.

 

[grmorton]

That's because the southern hemisphere data is basically flat. In fact, the change in the southern hemisphere is zero to within the error bars.

rising slightly. At least lets be honest about what is happening rather than spinning the data.

Edit:
If you look at magnitude instead of percent change, the area coverage in the Northern hemisphere has been decreasing at a rate of about 770 +/- 230 thousand square kilometers/decade, while the Southern hemisphere's mean value is a slight increase of about 110 +/- 150 thousand square kilometers/decade (which, by the errors, means a possibility of zero change or even slight loss).

And once again you ignre the fact that throughout history, from ice core information, when the arctic melts, antarctica ice extends. and when Antarctica melts, the arctic ice grows. Why do you ignore what I posted? Does data not matter to you?

Are you one of those guys who only talks about data that supports your position, while ignoring any data that is problematical???? If so, that is a reciipe for self-delusion.

If all you want from me is to act as if data contradictory to climate warming doesn't exist, you have the wrong person. As a guy who has made his living finding oil where others didn't because they didn't pay attention to all the facts, I can tell you that if that is what you want, you got the wrong guy.

I don't have much patience for people who only talk about data that supports their prejudices. If you are such a person, come back when you are mature enough to deal with contradictory data.

Edit 2:
Oh, and you can't cherry pick albedo data like that. Albedo will naturally vary from wavelength to wavelength. This is the really frightening thing about sea ice loss, however: sea is darker than ice, which means lower albedo. The loss in Arctic sea ice has been a major factor driving the much higher-than-average temperature increases at the North Pole.

Sigh, do you kow what the earth shine project is? It measures the amount of earth shine reflected from the moon. It covers a wide range of the visible spectrum

" Anomalies are shown
versus the 4-year average and are given in
terms of global reflected broadband shortwave
flux as well as in global albedo units. The
CERES data cover the entire Earth, for the
entire solar spectrum from 0.3- to 4-mm
wavelength. The earthshine results are primarily
for visible wavelengths and represent about half
of Earth_s surface."
Wielicki et al, "Changes in Earth’s Albedo
Measured by Satellite" Science 308 (2005), p. 825
​

​

​

And they also state the nature of the problem

"The average incident
solar radiative flux is 341 W m^-2, so that a
change in albedo of 0.01 represents a global
energy balance change of 3.4 W m^-2, similar
in magnitude to the impact of doubling carbon
dioxide in the atmosphere." Wielicki et al, "Changes in Earth’s Albedo
Measured by Satellite" Science 308 (2005), p. 825

Now, you say I am cherry picking wavelengths. I would respectfully suggest that this shows you know nothing about physics. Do you not know that given the composition of the atmosphere, one can calculate the bond albedo and geometric albedo? There are equations, rather compllicated equations, that can tell you what the variation is if you know part of the spectrum. Physical instruments don't measure all spectrums equally, so all spectral measurements are made and corrected using such equations. If you wan't to take a peak, look at http://www.aanda.org/index.php?opti...a/full/2005/46/aa3698-05/aa3698-05.right.html

So, I reject your claim that one is cherry picking frequencies. Satellite measurements are corrected using known physics to fill in the places where the satellites instruments don't go. The only way this can be wrong is if you say that the earth's atmosphere isn't ~21% oxygen and ~79% nitrogen with a mix of trace gases. You aren't saying that are you?

A question, have you ever run an atmospheric radiation heat transfer problem or programmed the interaction of radiation in the atmosphere? I have. I can assure you that your assertion is wrong.

Attached is an example picture of how albedo can be calculated from knowledge of the atmosphere.

 

[Steve Petersen]

MIT Scientists Baffled by Global Warming Theory, Contradicts Scientific Data

http://www.tgdaily.com/html_tmp/content-view-39973-113.html

Solar Minimum Heralds Colder Weather

http://www.foxnews.com/story/0,2933,366061,00.html

 

[grmorton]

LOL! That's funny. I have done fourier transforms using excel (no easy matter due to the clunkiness of Excel's abilities), and further, I work with 2-D fourier analysis of data on coatings quality defects, and I spent years chained to an FTIR (the FT stands for "fourier transform") doing chemical analyses

I do them daily on seismic data. I can't look at the jagged spectra, and predict what the seismic trace will look like. That is what I meant. I doubt you can tell me that either by looking at a FFT. If you can, you have a better multivariate processor in your head than any human I have ever met.

So, slow down here and understand what I was saying.

So, indeed, a fourier analysis of the data would mean something to me.

In the sense I am meaning that, want to try a test? I will give you an FFT of a seismic trace and you can draw it out without reversing the FFT. Want to try it? It won't mean much to you any more than I can visualize what a trace will look like from an FFT even after doing them for 39 years.

I am simply too lazy to force this data into that format. If you would be so kind as to back up your claim by doing so, I've done my part in not only fitting to a linear least squares but also a 2nd and 3rd order polynomial.

I am NOT lazy and I will do the work. The attached plot shows a peak at 64 months, and a strong peak at 1 month, which is very consistent with the data. One can see a periodicity of 1 month in the satellite temp data as well as one that is about 5 years, as I predicted.

Now, in point of fact, as far as the question of using linear regression, as you wanted to do vs fourier analysis, this still doesn't allow one to visualize the data any more than one could without the FFT. One could look at the satellite data and clearly see a periodicity of 4-5 years. The FFT verifies that. What new information has it provided? What new insight have we gained? As I said, it wouldn't mean anything.

BTW, I emailed a statistics professor friend and asked him if linear regressions are good with cyclical data. He said no. So, your linear regression is nice, but not meaningful.

I expect no less of you to fit according to your hypothesis and to back up the claim with an actual analysis. And some assessment of "goodness of fit" would also be of great value.

Well, if you truly understand FFT, then you will know that the above data (along with the phase spectrum) will be a complete description of the exact variation of temperature from the satellite. It is a PERFECT fit because I can invert this FFT and end up with the exact temperature curve. I don't think you use FFT as much as you claim or you would know this.

That's why I initially recommended fitting to a sine-wave in time-space rather than going all the way to frequency-space based FFT analysis.

Now it is my turn to LOL. An fast fourier transform is as close to a perfect fit with the data as one can get. The FFT was actually invented to deal with the data in my industry by the MIT signal analysis group.

But do, by all means, run the fit and show us your results. You seem quite convinced it is a cyclical phenomenon, so please convince us.

This last sentence clearly convinces me that you know nothing about fourier transforms. Attached is the fourier transform and the satellite data. See the jitter that takes place on a monthly basis? That is the highest amplitude frequency. See the 64 month peak? That is the underlying periodicity of the data. I would strongly suggest a perusal of the FFT part of the book John C. Davis, Statistical and data Analysis in Geology. That will let you understand that the FFT is a perfect fit to the data. But, in general, the biggest peaks are seen in the data without the fft.

 

[grmorton]

To my knowledge, most of the data around global warming trends are viewed as Temperature Anomalies, in other words annual temperatures subtracted from a multi-year mean temperature.

Indeed, NASA itself explain the use of anomalies as opposed to "absolute temperatures"

emphasis added.

This won't get you out of the problem of the raw data. Are you trying to say that the chinese stations which differ by 20 deg c wouldn't stand out like a sore thumb in an anomaly form? If you are, you are quite wrong.

You haven't commented on the Orange county register report that I posted talking about 7.5 deg of 'global warming' that has taken place in Santa Ana California. Do you think that CO2 lingers heavily over Santa Ana?

 

[grmorton]

NOAA (the National Oceanic and Atmospheric Administration) has presented this page (HERE) to describe the background to gathering temperature data.

Of note:
Of note:

The observations come from the U.S. Historical Climatology Network (USHCN), a network of 1221 climate observing stations in the continental United States. These data are extensively quality controlled for errors and for small biases that may have occurred through time due to artificial changes at each observing station. These artificial changes include station relocations, different instrumentation, and changes in the landscape surrounding the station (Ibid)
​

(emphasis added)


So much for any proof that it is incompetent bureaucrats who allow bad data to just pour in.

Boy are you gullible. If someone tells you that a used car is fantastic, only driven by a little old lady in Pasadena to and from Church, you will, of course beleive them. Do you not have one sceptical bone in your body after seeing their 'quality control?'

This is one of the funniest things I have heard in our debate.

But in case Glenn would still like to hold onto fears about the quality of data moving forward:

NOAA continues to work to improve the quality and representativeness of climate data provided to the public and scientific communities. In addition to advanced quality control procedures, these efforts include modernization of the USHCN by installing new, more accurate instrumentation, and ensuring proper station siting in the process.
emphasis added.[/quote]

Well, I guess I need to point out that that wonderful QC effort is why a 12 degree jump in temperature in Watersville Washington was allowed to continue. While they were picking their noses They didn't notice a sudden jump in temperature--yet you believe them when they say they do a good job of QC just because they say they do. Gullible.

Below also is the temperature chart (QC'd by these wonderful fellows you defend, which takes off like a rocket when they moved the weather station to the position shown in the photo. Notice the parked cars and the air conditioner. I got home about 2 hours ago and my wife got home about 3 hours ago. our car hoods are still warm to the feel. Parking cars next to the thermometer is not what I consider good QC regardless of the weather services claims. If you want to believe this stuff then you are not the person I had come to think you were.

Can you tell me why Watersville Washington's temperature went up so much so suddenly after 100 years of being much lower?

Can you tell me why Lampasas Texas (the other chart) took off when they moved the station, and why this is to be considered good QC?

If you are that gullible, you are as good a scientist as you claim.

Can you explain why, of all places on earth, Santa Ana California has undergone the most warming--7.5 deg if it is due to CO2? Does all the CO2 congregate around Santa Ana?

ARe you going to tell me in this forum where everyone can see it, that putting the thermometer above the running AC in Titusville Florida is good QC?

If you said everything tongue in cheek, it went right over my head and sounded like you were seriously defending the weather service QC. YOu didn't sound that way yesterday so I might be missing a wry joke on your part. I didn't see any smileys.

Geeminy if this is good QC, I have a bridge to sell you. I will be out of reach of the internet for a couple of days

 

[grmorton]

MIT Scientists Baffled by Global Warming Theory, Contradicts Scientific Data

http://www.tgdaily.com/html_tmp/content-view-39973-113.html

Solar Minimum Heralds Colder Weather

http://www.foxnews.com/story/0,2933,366061,00.html

Maybe when I get back I can review the coldest weather for the past 2 years as the sun went calm. but for tonight the UK is having the coldest weather in years

Coldest October weather for 34 years
[SIZE=-1]Journal Live, UK - Oct 29, 2008[/SIZE]
[SIZE=-1]WEATHER forecasters last night warned of snow for the North East in the coldest October for 34 years. Much of the region woke up to bitter temperatures and ...[/SIZE]


Early snowfalls causing chaos in coldest October for 74 years
[SIZE=-1]Herald.ie, Ireland - 14 hours ago[/SIZE]
[SIZE=-1]Tour buses ferrying sightseeing tourists across the mountain also got stuck in snow in our coldest October in over 70 years. Snow fell heavily in parts of ...[/SIZE]


'Coldest' Day In 51 Years Expected In Orlando; 2 Records Likely Broken
[SIZE=-1]Local6.com, FL - Oct 28, 2008[/SIZE]
[SIZE=-1]A cold front moving through Orlando is dropping temperatures to record lows for the end of October, likely breaking several 50-year "coldest day" marks. ...[/SIZE]


These things began about 2 years ago in the southern hemisphere and happened last year in the North, then again this year in the south, and they are starting again in the north. Believers will call it anecdotal. I call it low output from the sun--something the GW advocates don't seem to care about.

Off to the ranch

 

[Chalnoth]

rising slightly. At least lets be honest about what is happening rather than spinning the data.

I am being honest. No change is within the error bars. There is not enough evidence to suggest that the arctic sea ice is changing at all.

And once again you ignre the fact that throughout history, from ice core information, when the arctic melts, antarctica ice extends. and when Antarctica melts, the arctic ice grows. Why do you ignore what I posted? Does data not matter to you?

Well, this is just false. Antarctica, just like the arctic, is melting overall:
http://www.washingtonpost.com/wp-dyn/content/article/2006/03/02/AR2006030201712.html

Of course, as a gravity probe of the ice level, this is measuring the land ice melt.

Are you one of those guys who only talks about data that supports your position, while ignoring any data that is problematical???? If so, that is a reciipe for self-delusion.

I do hope you won't show, once again, just how ironic this statement is.

If all you want from me is to act as if data contradictory to climate warming doesn't exist, you have the wrong person.

Well, what I want from you is to stop cherry-picking data and look at the whole picture. anyway, I'm out of time...I have to meet somebody today to get some work done on a joint project. I'm out of here.

 

[thaumaturgy]

I do them daily on seismic data. I can't look at the jagged spectra, and predict what the seismic trace will look like. That is what I meant. I doubt you can tell me that either by looking at a FFT. If you can, you have a better multivariate processor in your head than any human I have ever met.

So, slow down here and understand what I was saying.

My bad, I was merely going off the point that you said:

If I showed you the Fourier spectrum, it wouldn't mean much to you.

Indeed, once fourier transformed it would mean something to me. It's either in the time or frequency domain, I don't see any others available, and I can assure you, it would make sense to me in either domain.

I am NOT lazy and I will do the work. The attached plot shows a peak at 64 months, and a strong peak at 1 month, which is very consistent with the data. One can see a periodicity of 1 month in the satellite temp data as well as one that is about 5 years, as I predicted.

Excellent! Thank you for posting that!

Now, please explain the time-series data I did on the same set and do explain the relative noise (because no matter how you slice it, there is a significant amount of noise.

But further, you are finding a significant 5-year signal in a database that has, at best, only 6 such cycles, and the "monthly" cyclcity (which I would expect) is less dominant?

Am I to conclude that the relative change in temperature on a half-decade scale has been of greater impact than the normal monthly or seasonal temperature changes in a 30 year period?

I am still skeptical.

BTW, I emailed a statistics professor friend and asked him if linear regressions are good with cyclical data. He said no. So, your linear regression is nice, but not meaningful.

I strenuously disagree. My linear regression has a p-value showing significance. YOU are under a burden to prove that the cyclicity is a better model.

No offense, but I can quantify the error in my model, you cannot (or so far have not).

So my linear regression is, at this point, much superior to your "single point" fourier transform.

Ask your stats friend to explain the Fishers' Kappa function. I would love to learn more about it. But from what I can tell, Fisher's Kappa indicates no such "cyclicity" among the noise to a 99% level of assurity.

Again, I could be wrong on this, but so far I'm the only one quantifying my potential error.

You'll also note that I have even provided you the amunition you need by providing the "Lack of Fit" data to show my model (both linear and non-linear) could be improved.

Well, if you truly understand FFT, then you will know that the above data (along with the phase spectrum) will be a complete description of the exact variation of temperature from the satellite. It is a PERFECT fit because I can invert this FFT and end up with the exact temperature curve. I don't think you use FFT as much as you claim or you would know this.

Then I think you need to explain the Fishers Kappa function to me so that statistics will line up. Unless, of course, you are going to now claim there is no "noise" or even "error" in the data.

Now it is my turn to LOL. An fast fourier transform is as close to a perfect fit with the data as one can get. The FFT was actually invented to deal with the data in my industry by the MIT signal analysis group.

So are we to assume that there is no error in the data?

No error? Really? All measurements reveal perfect agreement?

This last sentence clearly convinces me that you know nothing about fourier transforms.

Thank you. Isn't that a bit of an overstatement you think? While I am uncertain of the use of FFT in analyzing statistical or noisy data, I can assure you I am familiar with FFT. But I will defer to your superior knowledge. I will attempt over the weekend to verify this point.

I would strongly suggest a perusal of the FFT part of the book John C. Davis, Statistical and data Analysis in Geology. That will let you understand that the FFT is a perfect fit to the data. But, in general, the biggest peaks are seen in the data without the fft.

Well, it's a good thing I have a weekend then.

I am still amazed that noisy data which anyone on the planet would assume has some intrinsic error can be fit "perfectly" to reveal a "perfect" response. That fascinates me.

With your advanced knowledge would you care to comment on the Fisher's Kappa function which, if the good folks at the SAS Institute are correct would lead me to believe there is, at 99% confidence, no periodicity to the data?

I'd be very interested to learn more about this. I will be at an oceanographic institute this weekend and I'll be able to grab a copy of the Stats book you recommend.

I look forward to learning how noisy data can be utilized such that results from inherently noisy data can yield a perfect result without need of any error estimation!

This will be very interesting indeed.

 

[thaumaturgy]

This won't get you out of the problem of the raw data.

Well, apparently it does. Unless we are working under the assumption that the only data usable is that which you decree valid.

I find many times "raw data" must be processed and, gasp, "normalized" in order to correct for problems. Unless there's some reason to believe that NASA is incompetent in dealing with data as described, I should think that it is now up to you to prove that their conclusion of broader correlation of the "normalized" (anomaly) data is somehow in error.

Are you trying to say that the chinese stations which differ by 20 deg c wouldn't stand out like a sore thumb in an anomaly form?

I would really need to see how the anomaly data are subtracted from the "mean". Is the mean for the grid? If its for the station on a station-by-station basis, then there would be no problem. Trends would surivive while the relative offsets would disappear.

If it is on a grid-by-grid mean basis, then offsets become somewhat more problematic.

Do you see my point? Normalization can eliminate relative offset data.

I am still willing to assume, ceteris parabus, that NASA and NOAA are not systematically lying to me and that they are not wholly incompetent.

 

[thaumaturgy]

Glenn,
You wrote to your statistician friend, I too have one (well my wife shares a vanpool with a senior statistician). I posed my question about FT and statistical analyses. Here's his response (in part)

There are dozens of tests in the statistics literature, however. But first let's put your Fourier transform friend straight: the series is inherently a random process, and just taking the FT generates another random function, which has as much variability (and as little reliability) as the original series. Ask your friend to chop the series in half and see if he gets the same answer from the two halves, or whether either agrees in any detail with the result from the whole -- if his method is "perfect" the answer shouldn't depend on the length of the data series.

Hunting for periodic components in noisy data is a favorite activity of scientists of all kinds. Even if there are periodic processes, there are always additional chaotic or random signals as well (and the measurements themselves can never be exact) so statistics of some kind are absolutely necessary.

(Emphasis Added by Thaumaturgy)

Now I am going to still check out the reference or similar reference this weekend if I am able to, but indeed, as I feared FT is not a "magic bullet" to eliminate all variability in the data.

Unless you are of the opinion that the incompetent scientists who collect bad data non-stop somehow collect perfect data with no inherent variability.

That would, indeed, be miraculous.

 

[thaumaturgy]

Boy are you gullible.

That's the stuff! Of course it kind of smells like "conspiracy", but it's always a good way to steer the debate.

These "bureaucrats" are quite nefarious now. I am beginning to get a picture of them, let's look at them closer:

1. They are incompetent by continuing to collect bad data

2. They somehow collect "perfect" data (ie no noise so running a FFT on the data results in a "perfect" fit showing obvious periodicity)

3. They are quite crafty about their incompetence by spinning us a yarn about quality control so that the more gullible among us will buy it hook line and sinker.

They guys are amazing! I'd have to say it is hard to fathom how such nefariousness could be run out to support global warming using obviously biased data collected incompetently yet yielding detectable signals.

This is one of the funniest things I have heard in our debate.

Well, it's not so funny now that I see how strange these government scientists at NASA and NOAA and the countless other international temperature measuring bodies are.

This conspiracy of incompetent-perfection somehow is coordinated across both political and spatial regions!

If you are that gullible, you are as good a scientist as you claim.

ARe you going to tell me in this forum where everyone can see it, that putting the thermometer above the running AC in Titusville Florida is good QC?

Have I ever said this kind of thing is a "good thing"? So far in this debate I have agreed this type of thing is bad.

But, then, again, so far I'm the only one in the debate who is dealing with statistical data using statistics which by definition is how one deals with masses of data that have noise.

If you said everything tongue in cheek, it went right over my head and sounded like you were seriously defending the weather service QC.

I'm going to assume your life in geophysics has kept you insulated form "industrial quality control" like we have to deal with in manufacturing.

If you, for one microsecond, think there's perfect or even near-perfect QC in anything done by humans you are sadly sadly mistaken.

6-Sigma stuff aside, statistical process control exists because there are bad gauges and bad workers. There is, in a word, "error" out there.

I have so far seen that you have found a "group" of people who have analyzed 43% of the U.S. temperature stations and found 56% of that 43% to have a bias of 5deg (I have not been able to determine if they claim a systemic bias, but I earlier mistakenly assumed they were suggesting a consistent positive bias, I think I was wrong in that "accusation", my apologies). They cannot make any claims about the remaining majority of the stations. And they make no claims about the quality of unmanned ocean bouys and probes, they make no claims about satellite data, etc. Do they address the issues around the use of "anomaly" versus raw temp data?

Geeminy if this is good QC, I have a bridge to sell you. I will be out of reach of the internet for a couple of days

56% of 43% = 24%

Is that how you do your QC studies? Find the worst points of a limited set of gauges and assume it means your system is messed up?

I think it would be more likely that I could sell you a bridge.

At least I can quantify the potential errors in my reasoning. That's what I've been doing all along. That's what that little p-value thingie does.

What do you have except a string of jpegs?

 

[thaumaturgy]

Note the time series analysis I did earlier:

Interestingly enough, even apart from the discussion (still unresolved) around Fisher's Kappa is the preponderance of a low frequency peak in the spectral distribution.

Here's an interesting point from SAS:

Note the exceptionally high periodogram values at low frequencies. This comes from the trend in the data. Because​

periodogram analysis explains everything in terms of waves, an upward trend shows up as a very long (low frequency) wave.(SOURCE)

In my treatment of the data (which uses the annual average of the "Global" data column in the set) I am picking up a possible overall larger trend (upward?) in temperature.

Interesting.

This is getting very interesting. I'm learning more about time-series data that I thought I would! I look forward to Glenn's return from his ranch next week!

 

[grmorton]

I am being honest. No change is within the error bars. There is not enough evidence to suggest that the arctic sea ice is changing at all.


Well, this is just false. Antarctica, just like the arctic, is melting overall:
http://www.washingtonpost.com/wp-dyn/content/article/2006/03/02/AR2006030201712.html

Lets use the scientific data (which is online and scientific articles whenever we can. The journalists are not scientists and what they say is not peer reviewed.

I read that article and find something really interesting. They say that the instrument is biased to give too low a number. Thus, they did this

The averaging functions
are less than 1.0 over most of their respective
regions. Thus, they give results that
are biased low. To recover unbiased mass estimates
for each region, we scaled the estimated
mass signals to restore the original
amplitudes​

​

Of course, as a gravity probe of the ice level, this is measuring the land ice melt. This may be like the weather surface who changes the trends of bad stations to more suitable trends. Changing data is quite tough to get it right. Given the mass of ice on Antarctica, 36 cubic miles is not a large percentage.

I have published 2 papers on gravity data. I know that gravity has a high noise level and it has non unique solutions to the potential fields.

My publications
Terry Knighton, Steve Western, Glenn Morton and Robert Fleming (1999), "Development of Alternative Interpretation Models and Discriminating between Them Using a Borehole Gravity Survey and a Walkaway Checkshot Survey," Society of Exploration Geophysicists, Technical Program, Expanded Abstracts with Authors' Biographies, 69th Annual Meeting, Oct 31-Nov 5, 1999, Vol.1, p. 228-231

That paper got honorable mention

Prieto, Corine, and Morton, Glenn, (2003), "New Insights from a 3D Earth Model: Deep Water Region of Gulf of Mexico," The Leading Edge, 22(2003):4, p. 356-360
http://www.igcworld.com/PDF/Leading_Edge_April_2003.pdf
This was published in the premier exploration geophysics journal.

I do hope you won't show, once again, just how ironic this statement is.

Well, when you can cite some science that doesn't have the problems that this data does, you might have a point.

Well, what I want from you is to stop cherry-picking data and look at the whole picture. anyway, I'm out of time...I have to meet somebody today to get some work done on a joint project. I'm out of here.

Prove that I am cherry-picking. If you can't withdraw the charge.

 

[grmorton]

Well, I downloaded the paper and started to look over it. I'm afraid if the point was to "impress" it certainly did that, but sadly it now leaves me with more questions which are becoming more jarring:

Please forgive that I am wholly unfamiliar with Baysian Information Criteria, but the comparison of time-series temperatures looking for trends in temperature over time scales seems somewhat more, shall we say, mundane than the kind of analyses here.

This paper is only relevant to the issue whether I understand stat. I will freely admit that I don't understand it as well as Simmons and Yao, but I found a mistake in the pre-print which would have been embarassing. We got it fixed.

Are you now going to say that all large data sets cannot be modeled using statistics? Was that the point?

That wouldn't apply to time series of 100 years of annual average data. When you get to 3 billion, then there is indeed a problem. The data becomes it's own model. In other words, the normal tests fail for such huge data sets.

Perhaps you need to spell it out for the dummies in the audience (myself) as to how you leverage the DNA analyses and the breakdown of BIC and how that renders much more simple linear regressions somehow invalid for temperature measurements.

I didn't say that BIC had anything to do with temperatures. I was merely using this to point out that I know a bit about stat and that linear regressions are not useful for cyclical phenomenon. I work with cyclical time series, it is what exploraton geophysics is about; it is what it uses.
[/quote]

I see my fourier analysis is on page 9. Given my work schedule 12/day monday through Thursday and then 2 days at the ranch out of reach of the internet, I can't keep up. I didn't ever read page 8 until tonight and I see there are now 10 pages. Yuck.

 

[grmorton]

My bad, I was merely going off the point that you said:



Indeed, once fourier transformed it would mean something to me. It's either in the time or frequency domain, I don't see any others available, and I can assure you, it would make sense to me in either domain.




Excellent! Thank you for posting that!

Now, please explain the time-series data I did on the same set and do explain the relative noise (because no matter how you slice it, there is a significant amount of noise.

The noise to the right of the monthly signature (the highest peak) is simply algorithm noise. If you have exactly a multiple of 2 samples, you don't have a step function in the data. But if you have 336 samples then the time series suddenly stops and it takes high frequencies to represent that step function. But so long as one is aware of it it makes no impact on the decisions.

First off, the satellite temperature data you have is a yearly time series, not a monthly. I used monthly. Thus, you miss out on the monthly periodicity. You still get a periodicity of 4 years rather than 64 months but I think that is because you are not using the monthly data but are clearly using annual data. That reduces the fidelity of the signal and impoverishes the frequency content. Beyond that, since I didn't do your analysis, I can't explain it.

One more comment, by using yearly data, you have effectively filtered it. Your results should be different than mine (4 year (48 months) vs 64 months. )

But further, you are finding a significant 5-year signal in a database that has, at best, only 6 such cycles, and the "monthly" cyclcity (which I would expect) is less dominant?

Look there are other algorithms which can detect periodicities with even less data. I didn't use maximum entropy analysis which would be better. One doesn't need but one period to detect the periodicity, however, more is better. And since I was using monthly data rather than yearly data, I have more resolution than you do.

Am I to conclude that the relative change in temperature on a half-decade scale has been of greater impact than the normal monthly or seasonal temperature changes in a 30 year period?

I am still skeptical.

You clearly don't understand the satellite data. It is a global average, not a hemispheric, or Houston daily temperature. It should not show seasonal variations because it is a global average. Do you understand that? Do you understand that a global temperature won't show seasonality? Chalnoth didn't seem to understand that either.

I strenuously disagree. My linear regression has a p-value showing significance. YOU are under a burden to prove that the cyclicity is a better model.

pshah. One can use this approach to conclude that a series like 1,1,2,1,1,1,2,2,1,2,1,2,2,2,1,1,2,

is random and the fair flip of a coin. But if that is the output of a six-sided die, it is neither random nor fair regardless of what the p-value is. Those values are related to the model you re using. If you have the wrong model, you get the wrong answer and a false sense of security.

No offense, but I can quantify the error in my model, you cannot (or so far have not).

I don't have to. You have the wrong model. Noise would not go up and down rhythmically. Chalnoth a few posts ago said that with noisy data like this you couldn't get a periodic signal. I proved that wrong and proved that he doesn't understand time series analysis. I show the periods one expects with the data

Secondly you don't seem to understand that putting the data into frequency domain is an exact representation. there is no error save roundoff in the CPU

So my linear regression is, at this point, much superior to your "single point" fourier transform.[/quote



LOL LOL. This is not a single point transform. Do you know what the transform of a single point is, that is, a spike? It produces an amplitude spectrum which is perfectly flat and contains all frequencies with equal ampllitudes. Boy, you really don't know Fourier transforms very well.

As a point of information I had 357 points to transform--Just shy of 30 years of monthly data.

You are a smart guy. Don't try to argue about things of which you don't know. Be skeptical. that is ok, but don't argue about processes you don't know very well.

Ask your stats friend to explain the Fishers' Kappa function. I would love to learn more about it. But from what I can tell, Fisher's Kappa indicates no such "cyclicity" among the noise to a 99% level of assurity.

Again, I could be wrong on this, but so far I'm the only one quantifying my potential error.

My Fourier analysis of the monthly data shows strong periodicities at 64 months and one month. Quantifying the error against a bad model means nothing. And if you knew a lot about Fourier analysis, you would know, as I previously said, THERE IS NO ERROR. An inverse Fourier transform will return the same Satellite data with errors only in the round off of the computer words.

You'll also note that I have even provided you the amunition you need by providing the "Lack of Fit" data to show my model (both linear and non-linear) could be improved.

So are we to assume that there is no error in the data?

No error? Really? All measurements reveal perfect agreement?

A Fourier transform is a perfect representation of the data in frequency space. Any single-valued function can be represented as the sum of sin's and cosine's. That can be transformed back without loss of fidelity. THe FFT introduces a tiny amount of algorithmic noise, but I can assure you that it is so insignificant that If I were to transform the frequency data by the FFT, it would give back a curve so similar to the original that you wouldn't be able to tell the difference without a microscope. Believe, me, we use FFT's to process seismic data and if they introduced too much noise we wouldn't use them. Do you know what convolution is? Do you know what deconvolution is? The FFT is useful because a convolution in time becomes merely a multiplication in the frequency domain. We use FFTs to transform our seismic time series into the frequency domain so we can filter it via multiplication. Then we re-transform them back. That is quicker than a convolution operation.

Thank you. Isn't that a bit of an overstatement you think? While I am uncertain of the use of FFT in analyzing statistical or noisy data, I can assure you I am familiar with FFT. But I will defer to your superior knowledge. I will attempt over the weekend to verify this point.

Maybe, but I can tell that you haven't used them a lot even if they are of some familiarity to you. I will withdraw the claim that you know nothing about them, but, I will tell you, I have used them for 39 years on an almost daily basis. We use them also to find echos in our seismic data. If we transform a seismic section and see a particular periodicity, we know it is a multiple. Under the assumption that geologic layer thickness is random (I.e. has a white spectrum), we should not see periodicities in the seismic trace. If we do, then we use deconvolution to remove that periodicity and hopefully recover a clearer picture of the subsurface of the earth.

I am still amazed that noisy data which anyone on the planet would assume has some intrinsic error can be fit "perfectly" to reveal a "perfect" response. That fascinates me.

I think you misunderstand. The time series does have noise, but the FFT doesn't add any more, so it is a near-perfect representation of the actual trace. Maybe we have mis-communicated here.

With your advanced knowledge would you care to comment on the Fisher's Kappa function which, if the good folks at the SAS Institute are correct would lead me to believe there is, at 99% confidence, no periodicity to the data?

No. Maybe they should comment on why FFT shows periodicities on the monthly data. Even your eyeball can see the monthly periodicity. Go look at the monthly up and down, which your precious kappa function (done on yearly data not monthly) says doesn't exist. But maybe you don't want to just look at that chart. It is always a good thing simply to sit and do a gut check on the data one discusses. One can get lost in the forest of math; It is often good, as a hurricane forecaster once said, to simply go outside and look at the clouds and how they are moving.

I'd be very interested to learn more about this. I will be at an oceanographic institute this weekend and I'll be able to grab a copy of the Stats book you recommend.

I look forward to learning how noisy data can be utilized such that results from inherently noisy data can yield a perfect result without need of any error estimation!

This will be very interesting indeed.

As I said, you misunderstood. The FFT is an exact representation of the data. Strong periodicities won't show up in random noise.

 

[grmorton]

Well, it's definitely not above the noise, but the peak around 3-5 years might be the ENSO, but since that phenomenon isn't strictly periodic, it can take a very long time to pull out of the noise in a blind analysis.

No it doesn't take much to pull that signal out. I did it. Sheesh. you can't even look at the data and understand what you are looking at.

You are the one who doesn't want to explain the individual stations, which means you don't look at data do you?

 

[grmorton]

Oh, I definitely sympatize, having a long history of heated debates myself. I wasn't arguing for political correctness or against your straightforward and unrelenting approach. It was just the "herd"- and the "you all ignore the data" comments I felt was geting a bit too much in a debate where thaumaturgy had behaved in a very polite, openminded and objective manner. No doubt we can agree that we all have sheepish tendancies (though it's on a wide sliding scale), and that "your" (Morton's) demon can influence us all (The GW issue certainly being no exception).

We are all subject to Morton's demon, and I am no exception, even today. One reason I engage in debate is that it forces me to face up to any problems I have in the data. I have been in several GW debates where I was handed my head. But since I found the near city comparisons, I have been better able to defend my position. Notice how we have now gotten off of the raw data onto obscure points like FFT's and Fisher Kappa functions. The GW crowd in general doesn't like to look at the raw data. It is so bad that no conclusion about global warming can be made.

But perhaps unintentionally implying that your side is immune and all your opponents are thoughtless cowards, afraid of the data, is in my experience a poor tactic. Not only does it seem emotional and unobjective, and therefore a sure way to sabotage your own goal of getting people to look and think about your data and arguments, but it also induces the risk of lurkers entering the debate to pile on additional sweeping statements, and the thread quickly becoming a pointless flame war.

Absolutely agree that I go over the top some times. When I was 4 years old, my mother used to tell me not to be angry. I didn't feel angry and never understood what she was talking about. Such interpretations of my communication have remaind like that to this day. All I can say is that at 58 3/4, I doubt I will change.

Of course this thread is progressing well so far, so perhaps my concerns were a bit premature. But better safe than sorry I guess.

It is all about the data as far as I am concerned. But, occasionally one must force someone to actually look at it and see it from another perspective. I can take any seismic data and interpret it with an infinity of models, and they will all be consistent--that means, I can take any data set and interpret it in many way.

That sounds bad, but I learned this in graduate school in philosophy of Science (which is what I studied when in grad school), never finished the degree--got bored and had a son requiring me to get a real job). One thing I learned was that for any set of facts there is an infinity of explanations. Thus, nothing in science is settled, contrary to the claims of GW advocates.

I'm not sure I understand the problem, can you elaborate? By "good siting" these stations are expected to produce unbiased measurements and therefore should reflect the actual temperature changes. Are you skeptical of what they deem as being "good"? If you are, notice that the Peterson article are using the results of Davey and Pielke's analysis of the weather stations, one that showed that many stations were producing biased results, so it was hardly from a pro GW-biased analysis.

Peter

I am very skeptical of what they deem good. Why does a human, rather than the data get to determine what TREND in the data is good. This is (as I learned also in Grad school) what is called assuming the consequence. It is a logical fallacy. If you say that the temperature is going up at .5 deg/century and any temperature station which isn't going up at that rate is deemed to be wrong and in need of a homogeneity adjustment (as that Peterson article called it), and then the observed trend is changed to .5 deg/century, then one has assumed the answer and made the data fit the answer. That is what that article did and the GW advocates didn't notice it. It is mentioned specifically in the caption for two figures.

As I said, in Texas we call that cheatin'

 

[grmorton]

Oh my. You are familiar with who Roger Revelle is, aren't you? No offense, I'm assuming you are. He is, after all, on just about every third item with a name applied to it at Scripps Institute of Oceanography here in La Jolla. I hope you aren't questioning his "earth science" cred simply because he was an oceanographer. Some might even call him one of the fathers of modern oceanography.

If he claims that the current level of CO2 is unprecedented, I stand by my claim at least as far as CO2 is concerned.

I don't get too impressed by academics. Why? Years ago, I was area geophysicist for the Atlantic offshore for ARCO Oil and Gas. There was a guy, who will remain nameless, who was an academic expert on the east coast of the US. We paid him a heck of a sum of money to come in and look at our data and teach us what we didn't know about the East Coast. He spent the week asking us to educate him. We paid him lots of money for him to go to a seminar at our expense!

Why? Well, academics are starved for data. in the oil industry we buy data by the bucket load. When I was Geophysical manager for the Gulf of Mexico, there were some years I had a $24 million dollar budget for seismic. I also put together a $50 million purchase for 3 companies. We had more data than the academics would ever see in a career. So, if this guy says that the present CO2 rise is unprecedented, he should go look at the CO2 content of the Eocene Paleocene. It was at least 3 times the present level. At some point I will post on the last great CO2 episode and what the world was like at that time.

I hope you will flesh out your claim against Dr. Revelle. You will then have to go up against the entirety of the marine chemistry academic structure across the globe.

I am exhausted after 2 days at the ranch repairing the damage from Hurricane Ike. I had to hand wench an antenna to keep it from toppling, I roofed (not just shingles) a shed, one of 3 who lost their roofs in that storm.

(And please, re-read the quote, then reference the article in which Revelle and Seuss established the "Buffer factor" and then make whatever claims you like against the quote, but be careful to understand what the quote is in reference to.)[/quote]

I am sorry, I still have to laugh at this. CO2 in the atmosphere doesn't care whether or not it is human. The quote I was responding to, as I recall (and I am simply dead tired right now) was

"Human beings are now carrying out a large scale geophysical experiment of a kind that could not have happened in the past nor be reproduced in the future."

Given that we have gone from 280 ppm to 380 ppm, and in the Eocene period 55 million years ago the CO2 was 1000 ppm, 2.5 times higher than at present, exactly how can we claim that this experiment couldn't happen in the past--it did. I will freely admit that humans were not responsible for the past when there was 1000 ppm of CO2 in the atmosphere and if he is saying that humans couldn't cause it then, I would be forced to agree, but of what import is that fact, when the world had far MORE CO2 than we do now?

If your guy doesn't know about this, I stand by my claim that he doesn't know very much about earth history. He may know a lot about the ocean, but not about earth history.

Have you ever heard of the azolla layer in the arctic ocean? When wells are drilled in the Arctic, when they pass through the eocene, they drill through a layer of nearly 50% azolla plant fossils. Azolla is a FRESHWATER aquatic plant! It doesn't grow in salt water. It covered the arctic. Why? Because back then there was much more CO2 in the atmosphere and the amount of rain pouring into the arctic ocean made it fresh. We have been there before. This isn't a novel earth experiment. The only difference between then and now is that humans are the cause of the CO2. But the earth doesn't care about that.