Monday, October 08, 2012

Sea Level Rise and Climate Contrarians' Judgment

Generated in Excel from data here :

On the Twitter #AGW hashtag we find the following:

 cleric and idiot  Strikes Out With Breathtaking Stupidity  

The link is to a page on a contrarian website "Real Science" where Steve Goddard says "Sea ice has no impact on sea level, Antarctic sea ice is at the highest level ever recorded at either pole, and sea level is not rising in Tuvalu. Three whiffs for the hockey boy – send him back to strike out in little league."

Goddard links to the original  Huffington Post  article where Michael Mann is interviewed about Arctic sea ice melt.  Mann explains "we [will] really start to see sea level rises accelerate," as the Greenland and the west Antarctic ice sheets disappear. Unlike with the melting of sea ice, these ice sheets would introduce vast quantities of water into the world's oceans"

So Goddard is wrong to imply that the interview suggests that sea ice melt will cause sea levels to rise. 

He is also wrong about "Sea Level is not rising in Tuvalu". The graph Goddard publishes shows a clear upward trend. I point this out to @gdthomp01.  

He replies:
only an  Kool-Aid drinker like you would "see" a trend in that sea level graph but no trend here --> [points to a cherry picked selection of recent global temperatures.]
Your eye, is biased by your green agenda you try and push. Keep trying.

So I go to the data, and with a little help from @bobfishell and @bobirving99 (thanks guys) I manage to paste the data into Excel, to generate the chart you see at the head of this post - which clearly shows an upward trend of sea level in Tuvalu, of 3.57mm per year.

So Steve Goddard and Cosmocon are wrong on two counts: Mann did not claim that sea ice melt would affect sea levels, and there is a rising trend in Tuvalu sea levles. It is pointless to expect them to retract or apologise, because the tactic of the Right is to put confident claims out there and not to worry about whether or not they can be substantiated. The aim of the game is to create impressions in the minds of the onlookers, by means of headlines, not to try to represent the truth. 

In the process, we see also how they use pejoratives and derogatory language to belittle those who disagree. I am pretty used to this stuff now. There is no point in responding to it at the same level, but there is a point in continually re-iterating the fact that the contrarian hypothesis is incompatible with the facts. 

PS since we are talking about ice, people will want to now why Antarctica's sea ice sheet is growing.


Bob Fishell said...

We already is no point in arguing directly with these people, as they have a clear agenda and aren't interested in the facts. We might mention that the Arctic is losing sea ice at a rate 8 times that of the Antarctic, but only in the interest of reaching those who are reading the nonsense contrarians are spewing and need to know what's really happening. But thanks for the mention. Glad to help.

Cosmoscon said...

Wow, is this the first trendline you've plotted with Excel? You left off the R-squared value that would show the goodness of curve fit. You'll notice R-squared is 0.107 which is horrible. You can also play a game by just taking the last 11 years and see the trend line is negative!

You get a B+ for at least trying to use the real data.

DocRichard said...

You are incredible. Literally.

You know as well as I that contrarians stock-in-trade is taking a short series that matches your POV and parading it triumphantly as the nth final nail in the coffin of climate change.

There are 17 points here - the points offered by Watts as indication that sea level is not rising in Tuvalu. I pointed out that to my eye the trend was upwards. You blamed my eye. I go to the trouble of learning how to put up a trend line. You query the power of the data that Watts provided, and invite a shorter series that will match Watts' aims.

Answer me one thing: do you think this Republican Fuqua is out of order or not?

cosmoscon said...

Hey Doc,

What I was trying to say was that the data set you are trying to prove has a trend is nothing better than random numbers from a statistical sense. An R-squared value of 0.10 is meaningless. To prove it, I used a random number generator to obtain the following data set in order (with the ranges from the Tuvalu data 6750 - 7020). graph it and you'll see its R-squared value is higher than your sea level data. Both are characteristic of random numbers. It's basic statistics.


cosmoscon said...

And I don't know how politics works in your country but in America we don't blindly follow every political opinion of people who have the same party affiliation. Not sure what that link from an Arkansas candidate has to do with climate data but not too many sane people in America support killing children for rebelling against their parents.

Anonymous said...

Very interesting. Can someone explain to me why Cosmocon's point about the data presented is incorrect? I am not a sceptic and support the work of this blog, and I also know Doc is usually very thorough and very rigourous, but it does seem that Cosmocon has a good point re that specific data

DocRichard said...

The point is, Anonymous, that Cosmocon drew attention to a post on a "sceptic" site which asserted that sea level in Tuvalu was not rising. The site published the data before you as a graph. I said it looked as if it was rising. Cosmocon said I was misperceiving. So I took the trouble to make a regression figure to overcome the misperception accusation, and now he is saying that the data which formed part of his original anti-AGW posting is not adequate.

Bob Fishell has a point. Some contrarians are not worth engaging with. Cosmocon is one of those who deny that CO2 is a real greenhouse gas, so I really should not be wasting time with him. It would be different if he was civil and polite, but it is a pretty unpleasant experience discussing anything with him.

Cosmoscon said...

So Doc not only refuses to address his lack of knowledge in statistics and the folly of saying there is a trend in data with an R-squared of 0.10, but now he is telling lies. Doc, I have always said that I consider CO2 a Greenhouse gas and to say otherwise is a big lie.

Prove where I said that CO2 isn't a GHG and I'll retract calling you a liar.

DocRichard said...

In your piece here:
You write: "So the results of three different peer-reviewed papers show that over a period of 36 years, there is no reduction of OLR emissions in wavelengths that CO2 absorb.Therefore, the AGW hypothesis is disproven".

OLR = Outgoing Longwave Radiation. Greenhouse theory predicts that OLR will decrease over time as an effect of increasing CO2 levels. The papers you quote bear that out, but you prefer to use raw data and regard the disaggregation techniques used to sort out the CO2 part as cheating.

If the CO2 has increased and there is no decrease in OLR, then CO2 is not a greenhouse gas (GHG).

Of course, you can also say that you believe CO2 IS a GHG, because you are able to hold two mutually contradictory thoughts in your head.

On your page here
David Appell and others try to clear up your misunderstandings, but basically your position is that climate models may not be used in climatology. Which is a little like saying that CAT scans are not to be used in medicine. (CAT scans use computers too).

I will not waste time asking you to withdraw the "liar" slur, because I suspect that the grace to do that is not in you.

DocRichard said...

StHelensoregon in the discussion under your piece, says "You seem to have no idea of all the science that come before, but believe you can overturn it all with some simplistic notions published in a biased, nonscientific magazine. Science has far higher standards than this."

She does have a point you know. To believe that you can just dive in and correct a specialist paper in a peer-reviewed journal by someone who is a specialist and a scholar is pure hubris.

Is it that you believe that climate science is a huge conspiracy?

Anonymous said...

Ok, this is tiresome and I really knew better than to get involved.

But, that data (whoever posted it and for whatever reason and regardless of the validity of anything else this Cosmocon says) is inadequate and the graph / poor regression coefficient of 0.1 do not actually show what they purport to show with any degree of statistical significance...right?

DocRichard said...

Yes, which discredits one of Watts' points. The other is discredited in the original post.