A MUST SEE! Global Warming Swindle!

POSTED BY: CrevanReaver
UPDATED: Friday, July 3, 2009 16:48
VIEWED: 11643
PAGE 6 of 9

PREV PAGE 1 2 3 4 5 6 7 8 9 NEXT PAGE

CANTTAKESKY

Sunday, June 28, 2009 4:14 AM

Quote:
Originally posted by Bytemite:
And let me just say, point blank, that measurement error exceeding one degree Celsius, let alone ten degrees Celsius, would be very unlikely, seeing as how even a mercury thermometer measures to at least one degree Celsius of accuracy.

I don't mean measurement error. I mean standard error or standard deviation of averages.

For example, if you are averaging the ages of all the kids in one family, and the ages are
2, 5, 8, the average is 5. The standard deviation would be 1. So the average would be expressed as 5 ± 1.

If the ages of the kids were 4, 5, 6, the average is still 5. But the standard deviation would be 0.3. The average would be expressed as 5 ± 0.3.

This shows the range of data is tighter around the average, meaning the average is a very good representation of the range of ages, and therefore very likely to be accurate. In the former example, the opposite is true. The range is larger, the number 5 doesn't represent either 2 nor 8 very well, and is more likely to be inaccurate.

When scientists talk about averages, they are obliged to put that average in context. How tight is the range of data? They usually use one standard deviation as the ± error. There are roughly 6 SDs, 3 below the average, 3 above the average, and the average in the middle. For example, SATs have an average of 500, with an SD of 100. SAT's range from 200 to 800 (3 x 100 points below and above the average).

Now let's look at global temperatures. I think it is reasonable to assume the range of temperatures throughout the earth fall in between at least say, -15 degrees C (5 degrees F) and 45 degrees C (113 degrees F). Since this range barely covers the range in the USA, let alone throughout the planet, I'll assume it is narrow enough.

Even at this uncharacteristically narrow range of temperatures, we are looking at 30 degrees C below the mean of 15 degrees C, and 30 degrees C above. That would make the standard deviation roughly 10 degrees C.

So if you say, the annual global mean temperature is 15 degrees C, ± 10 degrees C, you would be describing the range of data you used to calculate the mean of 15.

Given that my example was unreasonably narrow, I would be very surprised if the real range of temperature data doesn't yield a much, much larger standard deviation, or standard error.

Of course, you can try to make this error smaller by averaging averages of averages of averages. But each time you average, you are supposed to carry the original range and error with you. It is called propagation of error. What climate researchers do, in a roundabout way with computer modeling, is treat averages like they are raw, measured data rather than statistical artifacts--that is, they do not propagate the error like they are supposed to. So that is why many people think the averages they show are measured data, as opposed to artificial data.

BYTEMITE

Sunday, June 28, 2009 4:23 AM

Hmm. I'd most often encountered reporting error in my works. I have done propagations of error, but I find myself unfamiliar with using standard deviation as a measurement of error. I've more often used standard deviation as a means of analysis.

In that case, I'm not sure that propagated error is given. All I know is that they're reporting numbers to 0.001 Celsius, which I took to be their statement of error.

CANTTAKESKY

Sunday, June 28, 2009 4:28 AM

Quote:
Originally posted by Bytemite:
http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts.txt

Yes. Note where they say: "sources: GHCN 1880-05/2009 (meteorological stations only), using elimination of outliers and homogeneity adjustment."

Not exactly unadjusted, is it? These numbers are all outputs of computer modeling (or as I like to call it, Sim Climate). They put in known measurements from weather stations, take out numbers they don't like, adjust the rest for "homogeneity," extrapolate into areas in which they have no measurements, and come up with an "average."

Quote:
the error is implicit at +/- 0.001 degrees Celsius.

I couldn't find this, nor how they came up with this "error." It looks like the kind of error that comes from averaging averages of averages of averages, not the kind of error that comes from raw, unadjusted measurement data. See my previous post on propagation of error.

Quote:
the temperatures can be converted back to their original, unadjusted measured temperature by adding the subtracted temperature, 14C.

See first paragraph.

CANTTAKESKY

Sunday, June 28, 2009 4:33 AM

Quote:
Originally posted by Bytemite:
In that case, I'm not sure that propagated error is given. All I know is that they're reporting numbers to 0.001 Celsius, which I took to be their statement of error.

I see. I am not looking for accuracy, the kind that comes from interpreting significant figures. I am looking for the standard deviation of the raw temperature data, as a mean of analyzing and interpreting the significance of a 0.6 deg C change in the averages of averages of averages over a hundred years.

BYTEMITE

Sunday, June 28, 2009 4:36 AM

In statistics, don't you normally remove outliers? I took it to mean numbers outside of two standard deviations.

Anyway, you asked for the numbers, I provided. Up to you, as ever, whether those numbers are convincing or not.

CANTTAKESKY

Sunday, June 28, 2009 4:40 AM

Quote:
Originally posted by chrisisall:
It's getting warmer. Is that too simple?

it depends on your purpose. If you are talking about politics or religion, simple is just fine. If you are talking about science, you need to be much, much more precise.

How much is it getting warmer, and is that extra warmth just random chance fluctuation or is it a significant change? Is it getting warmer in all areas or only most areas or only certain areas of the globe? Are these warmer areas representative of the entire planet, not just laterally on the surface, but vertically in the atmosphere? Are there any areas that are cooling? Are the methods for calculating warmth valid and accurate and reliable, or do they tend to be biased one way or another?

BYTEMITE

Sunday, June 28, 2009 4:49 AM

Also, the main page of the links explains the adjustments made.

Quote:
The GHCN/USHCN/SCAR data are modified in two steps to obtain station data from which our tables, graphs, and maps are constructed. In step 1, if there are multiple records at a given location, these are combined into one record; in step 2, the urban and peri-urban (i.e., other than rural) stations are adjusted so that their long-term trend matches that of the mean of neighboring rural stations. Urban stations without nearby rural stations are dropped.

The urban areas, they say, are averaged with nearby rural areas because monitoring stations in urban areas tend to have consistently higher temperatures.

CANTTAKESKY

Sunday, June 28, 2009 4:51 AM

Quote:
Originally posted by Bytemite:
In statistics, don't you normally remove outliers? I took it to mean numbers outside of two standard deviations.

There is no right or wrong answer here. It depends on what you are trying to accomplish. Removing outliers reduces variance, and increases the significance of small changes that otherwise would not be significant. It also unskews data if the outliers are very rare. Either way, it constitutes an "adjustment." That is the beauty and peril of statistics. You can adjust it to say whatever you want it to say.

Quote:
Anyway, you asked for the numbers, I provided. Up to you, as ever, whether those numbers are convincing or not.

Yes, you did good research. I've actually seen these numbers before, but none of them provides a standard deviation by which the mean can be interpreted in context.

When I asked you for them, I meant to make a point. Absolute SATs don't exist, the SDs don't exist. The numbers you provided were not what I asked for, and you will never find them, because they don't exist.

What would convince me is if I see absolute SATs that have NOT been adjusted or calculated with computer modeling (raw weather station averages), with an honest-to-goodness propagation of error that reflects the humoungous range of the earth's temperatures, showing a statistically significant change from year to year.

Anything short of that is religioin, entertainment, and speculation.

CANTTAKESKY

Sunday, June 28, 2009 5:12 AM

Quote:
Originally posted by Bytemite:
Also, the main page of the links explains the adjustments made.

Thanks for drawing my attention to this. I appreciate the fact that the numbers on these charts are relatively untainted. :)

You know, I don't understand why they don't use the absolute temperatures in comparison. Why don't they just say, in 1881, the Jan to Dec Annual mean was 13.89 degrees celcius, with an SD of X (propagating the error from the monthly and daily averages); in 2008, the Jan to Dec Annual Mean was 14.55 degrees celcius, with an SD of X. Why do they make you compare everything to that darned artificial climatology? It just makes them look like they are trying to manipulate data to inflate significance.

BYTEMITE

Sunday, June 28, 2009 5:28 AM

Well, I can't blame you for being skeptical.

But I can't exactly say that I think that those managing the temperature databases are intentionally hiding their standard deviations for their number sets or the raw data, nor that they are using the data to try to achieve a predetermined result.

I take them at face value when they describe how they have adjusted the data; I have no evidence to suppose that they are hiding some of their adjusting factors.

If there is a problem with the models, then it may very well be the issue of propagated error. But I also have no evidence to suppose they are perpetrating fraud. Motive, say, for government grants for research is something that might throw suspicion upon them, but is not directly condemning.

I find four different databases, three national and one in the UK, producing four very similar looking trends in temperature data very convincing that there is a statistically significant global average temperature difference from the 1880s, especially considering that they're using different methods of statistical selection, averaging, and analysis.

However, I do think there may be a substantial amount of fear-mongering and sensationalization in the alarm-blowers in regards to global warming, and that a lot of that must be taken with a grain of salt. I also think that the climate change models are far from perfect. I do not think we're really going to be able to predict how climate will change.

I only think that there has been some global warming, that human activity is the common factor for the otherwise inexplicable warming, and that I think that this will cause some climate change, which we've possibly begun already to feel the effects of. By which I mean, the documented glacial melt, which we can measure unambiguously.

I also think more intense storms may be the result of warmer temperatures, and we may have experienced some of these storms, but this will be difficult to measure empirically. Similarly, I think the droughts in Africa may be exacerbated by the effects of global warming and climate change, but again, that's hard to measure.

PREV PAGE 1 2 3 4 5 6 7 8 9 NEXT PAGE