My belief in Global Warming is getting shaky

[Michaelangelica]

[Moontanman]

Have any of you guys checked out the Category "Science News" lately? There are a couple of news briefs that are pertinent to global warming.

[Turtle]

Originally Posted by Moontanman

Have any of you guys checked out the Category "Science News" lately? There are a couple of news briefs that are pertinent to global warming.

I read them; yes. Interesting new pieces for sure, but let's not make it a misconstruction.{bolding mine}

http://hypography.com/forums/news-in...orse-than.html

Originally Posted by Alister Doyle

"Historians of science hate to say 'this is a special time'," Crowley said.

"But when you go through the models, each step seems reasonable and you get to an astonishing conclusion that we are right at the end of a 50-million-year evolution."

Modern human societies might never have developed if such a freeze had happened slightly earlier. "Anatomically modern humans evolved only 150,000 years ago," he said.

Crowley said more tests of the projections were needed.

"It might not come for tens of thousands of years," he said. "I'm sure some headline writers will want to say 'CO2 good for the atmosphere', or 'CO2 is good for us'. That's not the case."

[engineerdude]

Hi all,

Been away on a business trip, just got back yesterday.

I started a post before I left, I realized my charts were messed up and had to pull it.

Anyways, if you are willing, here I go again.

Talking with you folks has expanded my perspective on the arguments for AGW. Based on the input you have given me, I decided to refine my arguments a bit and try to focus on cornerstones of what people believe in.

First, the assumptions:
1. Earth goes through periods where it gets warmer and cooler. It does this every few hundred thousand years. No one know why for sure.

2. CO2 lags behind temperature. This is shown in the ice core data. When the earth decides to warm up or cool down, it does it, regardless of CO2.

3. ASSUMPTION I WILL TRY TO DISPROVE CO2 acts as a feedback mechanism. Lots of CO2 make the planet warmer than it would otherwise be, low levels make things cooler.

OK, there is lots of math on both sides of the feedback argument. IPCC people show equations proving it, others show equations that show the opposite. Our climate has been quantified to such a small scale that the chance either group is right is about nil.

So, let's look at the historical record. Here is a chart I made up, it is just the Vostok data lined up nicely. Any red marks are things I put on there, everything else is just raw data from the ice core people.

[IMG][/IMG]


Here's a summary of the graphical data.

Event Temp CO2
Current Times 0 370
Event A +2 255
Event B +3.5 275
Event C +1.5 265
Event D +3.5 300
Event E +2 275

I am just looking at the peaks, since they are easily identifiable, and since we seem to be coming off a recent peak.

With CO2 levels at unprecedented highs, shouldn't it be at least as warm as other times CO2 was high? Even in recent times, global temperatures peaked +2 degrees C above current temps around the times the pyramids were built, and CO2 then was much lower than today.

For there to be some feedback correlation between CO2 and temp, high or low levels of CO2 need to have at least a minimal positive correlation to temperature. And it is just random.

Don't believe me? My wife is a statistics professor at a local university, I had her run a data correlation. For the record, she kind of believes in AGW, not for any scientific reason, just a feeling she has. Women, sheesh.

Anyways, what she ran was a correlation coefficient for this data. A value of +1 means a strong positive relationship between things, .5 means a weak positive relationship, and close to 0 means no relationship. -1 means a strong negative relationship, -.5 means a weak negative, and so on. Values range from -1 to +1.

The data came back as -.56. This indicates a weak negative relationship between temperature maximums and CO2 levels. Basically, higher overall CO2 levels kind of make temperature maximums a little lower. Exactly the opposite of what AGW people believe.

So, what do you folks say about this? How can CO2 be some sort of feedback mechanism if the record shows otherwise?

[Chris C]

engineer,

different orbital (milankovitch) conditions. In particular, more sunlight at higher latitudes, evidence is not necessarily synchronous globally. Latitude effects much less pronounced today due to different wobble and tilt effects (eccentricity itself provides a small effect but influences precession). Need to be very careful about high-frequency versus low-frequency changes (Vostok probably correlates fairly well with world temperatures for slow things, but there is a high-frequency overprint seesawed with the north that has to be filtered out first), and don't just look at the peaks. The CO2 response to temperature is time-lagged and the CO2 feedback effect is logarithmic, not linear.

The global temperature was never +2 C in the pyramid times.

[engineerdude]

Originally Posted by Chris C

engineer,

different orbital (milankovitch) conditions. In particular, more sunlight at higher latitudes, evidence is not necessarily synchronous globally. Latitude effects much less pronounced today due to different wobble and tilt effects (eccentricity itself provides a small effect but influences precession). Need to be very careful about high-frequency versus low-frequency changes (Vostok probably correlates fairly well with world temperatures for slow things, but there is a high-frequency overprint seesawed with the north that has to be filtered out first), and don't just look at the peaks. The CO2 response to temperature is time-lagged and the CO2 feedback effect is logarithmic, not linear.

The global temperature was never +2 C in the pyramid times.

Well, according to the ice core data that they have published, and I have referenced graphically in the above chart, it looks like +2 degrees above present times to me, around 8000 years ago. Look at the chart and tell me what you see.

High frequency, low frequency, logarithmic. I understand these theoretical concepts, but where in any historic record does it show some sort of correlation between CO2 and global temperatures? Please, produce some sort of actual evidence.

[Chris C]

Sounds about right for Greenland given more summer sunshine. You need to be careful about extrapolating regional information over the globe when the forcings are not global forcings (unlike CO2).

This paper has some simple explanations of things going on. The important ice core reconstructions and IPCC AR4 also discuss the feedback effect
Glacial cycles and carbon dioxide: A conceptual model

[engineerdude]

Originally Posted by Chris C

Sounds about right for Greenland given more summer sunshine. You need to be careful about extrapolating regional information over the globe when the forcings are not global forcings (unlike CO2).

This paper has some simple explanations of things going on. The important ice core reconstructions and IPCC AR4 also discuss the feedback effect
Glacial cycles and carbon dioxide: A conceptual model

The "model" this fellow has created is a pure waste of time, as far as I can tell. He is simulating a mechanism that is in no way backed up by the geologic record. One could just as well create a model which shows the earth bursting into flames or becoming a permanent ball of ice.

As a start, these people should at least back-test their models to make it match the geologic record. The model would still be mostly useless, the earth is too complex a system to begin to model with current technology.

Think I'm wrong about models being useless? Look at the weather forcast in your town for next week. See if they get the temperature correct for a day next week - I bet they miss it, at least a little. The models these climatologists use are no better than the National Weather Service, and they rely on predicting things like ocean temperatures to 1/10 of a degree out centuries. If the news could tell me next Sunday's highs and lows within a degree 99% of the time that still wouldn't be good enough - the error rate compounds so fast that computer predictions at present just are not possible.

[Chris C]

This at least started well. I strongly suggest you start with an undergraduate level textbook on this subject-- David Archer's perhaps. Instead of looking for ways to debunk AGW and "revise" arguments that can easily be invalidated, understanding the core basics is essential. You don't need to respond- that is my final comment on this matter.

[CraigD]

Originally Posted by engineerdude

So, let's look at the historical record. Here is a chart I made up, it is just the Vostok data lined up nicely. Any red marks are things I put on there, everything else is just raw data from the ice core people.
…
Here's a summary of the graphical data.

Event Temp CO2
Current Times 0 370
Event A +2 255
Event B +3.5 275
Event C +1.5 265
Event D +3.5 300
Event E +2 275

Engineerdude, what’s the source of your data (actual URLs)? Including sources is not only a sound writing style, a hypography site rule, but in this case, I’d find the information useful, as I was unable to find co2 data older than 1950 – 21676 at the usual reference, ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/epica_domec/.

Originally Posted by engineerdude

For there to be some feedback correlation between CO2 and temp, high or low levels of CO2 need to have at least a minimal positive correlation to temperature. And it is just random.
…
Anyways, what she ran was a correlation coefficient for this data. A value of +1 means a strong positive relationship between things, .5 means a weak positive relationship, and close to 0 means no relationship. -1 means a strong negative relationship, -.5 means a weak negative, and so on. Values range from -1 to +1.

The data came back as -.56. This indicates a weak negative relationship between temperature maximums and CO2 levels.

I confirm a simple population Pearson’s correlation coefficient of -0.5599 for the 6 pairs of numbers you provide.

However, when sampling data, one should not selectively choose number pairs to produce a desired correlation, because by doing this with most data, it’s possible to obtain almost any desired correlation coefficient.

If I take approximate values from where the red lines on your graphic actually touch the CO2 and temperature graph lines, I get the following number pairs:

0	260
2	255
3.5	275
1.5	265
3.5	300
2	275

The correlation of these numbers is about +0.70. Omitting the first pair, the correlation is about +0.71.

A better method is to take all available data pairs. Taking the data from ftp://ftp.ncdc.noaa.gov/pub/data/pal.../domec_co2.txt and ftp://ftp.ncdc.noaa.gov/pub/data/pal...uttemp2007.txt (Valérie Masson-Delmotte) (because the EDC3 age data from these sources don’t match, I interpolated the edc3deuttemp2007.txt data to match the integer year domec_co2.txt data with a simple line between closest matching age values), 72 number pairs, the correlation is about +0.96. Here are the 72 number pairs I used:

-1.065329194234635629 265.2
-.158947635783188996 263.0
-.5584413224027791624 264.4
-.144524664284869823 264.0
-.8416443654221001913 265.7
-.496985346676197284 267.5
.4746534680222031251 266.0
.4745670073421107581 267.6
-.817765177536387446 264.8
-.463447173246880135 264.8
-1.028548646150191341 265.3
.1915139892973107513 264.1
-1.647357693440441975 264.5
-.2102375940191850343 264.0
-.5394785393440990403 265.2
.474377719043832156 260.8
.098242453082711239 253.9
-.659147633833421398 249.7
-.70390469941769868 250.7
-.6373157675347184134 245.3
-2.051994190681778976 246.6
-2.691622679553772985 243.2
-3.135701990225695258 237.5
-4.69795153795937176 237.6
-4.33133372203730082 234.2
-4.785070264410979056 238.3
-5.287328352490421456 237.3
-4.862575279937237002 237.9
-4.792878953265623306 237.6
-3.596346840881442585 236.4
-4.482904930621442125 239.2
-4.916765333399600269 238.6
-4.838520379397218039 238.6
-4.701998174785971926 239.1
-3.813157136068430223 228.5
-1.022168530590275701 228.4
-4.986438119494193789 226.1
-4.016842794710083571 225.2
-4.536447543002152239 224.5
-4.066314442303059504 222.0
-4.241225348215755428 221.0
-4.942601049575703439 220.9
-5.648515040008219987 219.4
-5.551840764888777935 214.0
-5.273132585650637579 207.5
-5.673570850003681795 207.7
-6.867180858258674825 202.9
-7.32609938460921683 200.8
-6.847082059950032155 195.2
-8.273133418029521646 193.9
-7.763623115349911029 191.0
-7.741095955610985263 188.5
-8.971481861037097766 188.5
-9.52052954906048805 189.2
-9.80006894274300932 187.0
-9.40035853343072933 188.6
-9.072123719482342943 189.4
-8.81662161251936505 192.3
-9.72167161336401873 188.3
-10.22681115367594799 188.7
-9.4928563989063924 188.8
-9.23920143754708009 190.0
-10.09424499044343287 188.0
-8.650727869059257624 188.2
-9.29092465861427355 195.0
-9.166331761364632457 187.8
-10.07557609943346622 186.9
-9.64762431363753416 186.5
-9.42505630124173112 184.7
-9.088613015686055359 186.1
-9.2298818795162803 185.7
-9.56708826629612066 184.4

Originally Posted by engineerdude

So, what do you folks say about this? How can CO2 be some sort of feedback mechanism if the record shows otherwise?

As I note above, because you chose your small number of data points improperly, the negative correlation you found is not statistically valid. If selected in a statistically valid way, a strong positive correlation appears.

It’s important to note that correlation does not imply causation. A strong positive simple correlation coefficient does not show that increased CO2 concentration causes or is caused by increased temperature. However the claim that there is actually a strong negative correlation between these data is incorrect.