Studying Temperature Data Using the Language of Science

Without greenhouse gasses, Earth would be unlivable. Water vapor is the dominant greenhouse gas.

 

Please. I did not create this thread for dullards, and I expect at the very minimum that anyone reading through the steps being followed to assemble a temperature temperature profiles for an area the size of Idaho understands the definition of the word "global".

 

Everyone knows that. The question is how much of the greenhouse effect is due to CO2.But you do agree that CO2 is a greenhouse gas, right?

 

You don't honestly believe that global warming means that every part of the globe is warming, do you? Tell me you know more about the subject than that.

 

He can't. He isn't examining data, he is limited to providing references to the claims of others. Those references say all this normalizing, aggregating, averaging, assuming and interpolating is a-okay. Just don't ask them to submit these processes and procedures to independent review…goodness no…someone might ask a question they can't answer or don't want to…such as…why aren't you providing the empirical error estimates along with the results and conclusions? Not assumed errors based on sampling density, not assumed distributions of errors because you haven't quantified the underlying error, oh no, the actual error because of all this normalization, aggregation, averaging and interpolating. Why don't you provide THAT error estimate.

It is my theory that they don't provide this information because it is difficult, and would blunt all of their conclusions if they were forced to provide the all-in error estimates.

 

Please tell me you know how to use a dictionary. When you get the dictionary to redefine global, let me know and I will rethink my position. Until then, I will stick with the definition.


glob·al adjective \ˈglō-bəl\
: involving the entire world

http://www.merriam-webster.com/dictionary/global

 

The average citizen knows no more about trace gases in the atmosphere than they do what happens when they flick that switch on the wall and suddenly "LET THERE BE LIGHT!!" is the result. Pretending that a population that believes The Big Man In The Sky rules their daily lives knows much of anything about the physical world isn't just a stretch, but is silly.

Sure CO2 is a greenhouse gas. So is methane, nitrous oxide, ozone.

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

Why do you have such a fascination with only one of them, and a relatively small one at that?

 

You might want to understand what global warming actually means before you attempt to refute it.

 

When CO2 has added 1.46 watts per square meter of climate forcing while all other GHGs have added only 1.17 watts per square meter, why shouldn't we be focused on it? CO2 may seem relatively small compared to H2O, but since H2O can only react to warming, not cause it, focusing on it would be a distraction.

 

You might learn to read what I have written before pretending I say things I do not. The bias you imply within a scientific inquiry should never exist, hence my complaint about trading honesty for scary scenarios in my OP.

An obvious tell of a zealot is assuming that just because someone is asking a question about your belief system, it is implicitly designed to refute it. Properly done science is as far from zealotry as you can get, and you would understand this if you had ever participated in a scientific inquiry. Science doesn't care what your belief system is. Neither do I. It is as likely that I can establish the validity of your belief system as refute it.

 

Feel free to focus on it all you'd like. Start an entire thread. Do your own science demo.

But this thread is about the particulars and uncertainty in manufacturing temperature profiles

 

So because you don't understand the science, that proves fraud.

That would be why nobody pays any attention to you. There are all kinds of sources you could have looked at and should have looked at before accusing everyone of fraud. But that would have been too much trouble, and would have interfered with the conclusions you had made before you began.

 

I am constantly surprised by how desperately people want to put words in the mouths of others, for whatever reason. Certainly your statement is wildly incorrect, and cannot be drawn from anything I have ever written here.

Again, commentary of a nature that only says something about your own perspective. I don't recall saying anywhere in this thread that it matters in the least who pays attention to me, in these forum ramblings and experiments. If I haven't made it clear before, this is fun and games time. No one is requiring analysis from me, there are no deadlines to meet, no cranky editors demanding a rewrite of some paragraph to make it more clear to the laymen in the audience. I can proceed at my own pace, investigate whatever I have an interest in, and lay it out as I go for those who might have an interest.

It will, however, allow me to point to this thread as a reference when yet another forum expert pretends that handing out yet another temperature profile has meaning, WITHOUT the kind of estimates I have been handing out like free candy.

Fortunate then that doing a bottoms up inquiry has nothing to do with accusing anyone of anything, and that I made no such claims. Feel free to validate your position through a thorough investigation of your own...or conversely...continue to point out references that aren't providing the level of detail I have already done.

I mentioned before the tell of a zealot...always assuming that just asking the question into their belief system can only be attempting to refute it. Do you do this because of that deep down, niggling fear that because YOU can't do your own investigation, don't do science, don't understand data, that you are subconsciously afraid you've been sold a bill of goods?

 

I'm not surprised how you were unable to recognize your own tactics being reflected back at you. It's the true believer thing.

You're following in the footsteps of famous denier Steve McIntyre. His interactions with scientists tended to go like this, which is why everyone ignores him now.
---
McIntyre: You're a fraud because my analysis shows this and this! Show me the data!

Scientist: Okay. You need to look at this and this, because you went wrong here.

McIntyre: Well ... you're a fraud because my analysis shows this and this! Show me the data!

Scientist: No, now you need to look at this and this.

McIntyre: Well ... you're a fraud because my analysis shows this and this! YOU SHOULD BE IN JAIL! Show me every last bit of your data, including all personal emails, you big fraud!

Scientist: This is a waste of time. Goodbye.

McIntyre: See! I knew it! You're hiding the data! That proves what a fraud all you scientists are!
---

 

So now make believe conversations are your proof? Can't get much more ridiculous than that.

 

Well, feel free to refute any of the data, analysis, statistical assembly or examples provided to date then. I welcome someone providing such a response but, for some reason, all you seem to be able to do is talk, and reference the work of others that you aren't even familiar with, and doesn't refute a single chart, distribution or statistical metric I've provided.

 

Determination of correlation within Idaho temperature data

Based on the previously explained statistical test, here are some results. This graph shows the relationship between groups of stations directionally reacting. For example, if 3 stations dropped in average daily temperature from Day 1 to Day 2, and 7 stations increased, I assigned a value of 70% for that day. Then I went and began making histograms of those days. A value of 50% (or less, to account for stations that recorded identical back to back daily average temperatures) signifies that as many stations got warmer as got colder, and therefore the assumed correlation on that day was 0. 100% signifies that all stations moved in the same direction on that day, they all got colder, they all got warmer. It looks like there is near equal probability of some daily event driving temperatures all in one direction as there is the ability to some to go up, some down, and demonstrating no correlation whatsoever.






This chart was constructed the same way, only using the change from Month 1 to Month 2, and comparing all the monthly average temperatures to get the relationship.





Notice the HUGE difference between the two. Temperature change from day to day is certainly correlated at the areal extent of our Idaho example, but it is highly correlated to the point of being perfectly correlated a large percentage of the time. As usual, I am providing examples of the probability involved.

I was quite surprised by the monthly data, I hadn't really thought it much, and didn’t expect it to be nearly this highly correlated. But upon reflection, it makes perfect sense. Daily variability is called weather. Clouds, storms, rain, sunshine, whatever. Weather can range in size from small to large, but it certainly is large enough to keep daily temperatures generally correlated. Cold front comes through, high pressure system, big storms. But monthly temperatures are following an entirely different cycle, which everyone here is familiar with. They are called seasons. So from February to March, and average temperature can be expected to increase as winter recedes. Same with September, as summer becomes history. At the monthly level, this profile jumps right out at you. Here is an example from the monthly Idaho data from Aberdeen Station, 1914-1939 I believe, and you can see the cycles. All the stations are operating within this seasonal cycle, so it completely expected that this high level of correlation would occur in this type of test.

 

Interesting article on temperature data tampering. I will include some of the pictures.

A Global Epidemic Of Data Tampering At The US Government

 

Based on the span of assumed correlation provided previously, would any one like to venture a guess, or their own empirical work, as to the uncertainty around the mean answer BASED ON CORRELATION alone? Express it as a % of mean perhaps, if the change in mean for any particular starting point is 2C, if I recalculate it with the proper uncertainty built in, what is range that mean might change within? +/- 25%? +/- 100%? And then we will compare that number to that provided by folks on the German data, or the empirical Idaho data we have been using.

 

Interesting that the supplied focus point are the extreme's, as compared to the very insidious issue of trending. You see, part of Hansen's game in his 1999 paper was to apply corrections that appeared to assume a level of correlation across wide swaths of data, when no so such perfectly positively correlated pattern has shown up in real data yet...although monthly data and seasons exhibit a pretty high level. That is expected however, whereas in annual data it certainly can't be taken for granted. Interestingly, every issue I've seen to date on how these correlations are handled are either stated, "these data are higher correlated across large areas" or themselves say this and then reference others, who themselves make an assumption, or build something resembling a surface...and when they do, they don't provide the corresponding correlation. Correlation measures are easy, they range from -1 to 1, are provided by scientists who aren't meteorologists right down to the exact level of the thing they are correlating (reference previously provided)...but meteorologist-climatologists apparently do not.

 

The MET office in the UK is showing some responsibility in reporting, not that the media will ever report it. Emphasis mine.

http://www.metoffice.gov.uk/hadobs/hadcrut4/faq.html

.

 

I would instantly doubt the magnitude of that uncertainty around a mean answer. Based on the magnitude of uncertainty that can be demonstrated within temperature data at small scale (let alone assumed trends within), it would appear that the article is just taking the word of the scientists who have assumed an error within the manufactured profile chosen. To date, all provided references of uncertainty do some version of this, and avoid ground up empirical observations of uncertainty like the plague.

 

Don't know if you would be interested in this post by Judith Curry.

http://judithcurry.com/2015/02/15/denizens-ii/#more-17790

 

That Curry piece isn't flattering to deniers, as it shows most of them either proudly self-proclaiming themselves as political conspiracy cultists, or falling for weird revisionist versions of logic and science to justify becoming deniers.

WUWT just did a similar call out. Pickings are getting mighty slim in denierstan concerning real science. Fortunately, Cook and Lewandowsky can probably use it all as evidence in another paper examining the mental state of deniers.

 

So they can make up more stories eh?

http://wattsupwiththat.com/2013/03/09/stephan-lewandowsky-and-john-cook-making-things-up/