Um, even if you subscribe to agw, shouldn't we be at least a little bit skeptical of estimates of ocean temperature before WWII? I mean that article has a chart of global temperatures going back to 1850. How the frog do they know what global average ocean temperatures were 35 years before the automobile was invented?

So I think that the guy mentioned in the article, Peter Thorne, is the main person responsible for assertions about ocean surface temperatures pre-WWII. I read one of his older papers (I'm having trouble finding it right now), where he describes how he estimates the ocean surface temperatures from way back then. You have to admire his confidence/optimism in making those estimates, but to me the paper was not completely distinguishable from an Onion[1] article. Apparently back then seafaring ships used to measure ocean temperature[2] by lowering a bucket tied to a rope over the side, pulling it back up and then sticking a thermometer in the bucket (full of ocean water). Iirc, the water in the bucket could have cooled or heated by up to something like 5 degrees by the time a temperature reading was taken. He then proceeds to estimate(/guess) the amount of cooling/heating by factoring in the average elevation of a ship's deck (as a function of date) and how long it would take an average person to pull the bucket back onto the deck. He also factors in things like the typical material a bucket would be made of (also as a function of date).

To me, he seems to give off a vibe of being really proud of his "rube-goldbergian" estimates, but personally, I'm not as confident as he (and apparently everybody else) is in those estimates. And since ships only traversed a small portion of the oceans in those days (the trade routes, generally), there is a lot of extrapolation involved. And assumptions about the correspondence of ocean temperatures and nearby land temperature readings.

Imo, the accuracy, or lack thereof, of pre-WWII ocean surface temperature estimates is not super relevant to the task of constructing predictive climate models. But it does seem to me to be perhaps an indication of an insufficiency of healthy self-skepticism in the climate science community.

[1] https://en.wikipedia.org/wiki/The_Onion

[2] This is actually because Benjamin Franklin did it and discovered the Gulf Stream, which is pretty cool.

And my sincerest apologies to Peter Thorne. My statements about him were completely inaccurate. If I could delete the comment, I would.

After some googling, this paper may be related to the one I recalled reading: ftp://podaac.jpl.nasa.gov/allData/gosta_plus/retired/L2/binary/docs/document/papers/3_clmchg/3_clmchg.htm

Yeah, sorry about the low quality comment. But the more sincere point was, don't claims to know the global average ocean surface temperature in the 19th century with some accuracy qualify as "extraordinary claims"? I feel uncomfortable just accepting it without a) extraordinary proof, or b) some explanation why it's actually not so extraordinary, or c) maybe error bars or uncertainty clouds on the graph?