Thursday, January 9, 2014

Wiggle Matching (was:brutality)

In the blog Science of Doom there's been a lot of deliberation about the ends and the beginnings of the ice ages, or glacials as they're known in the geology department, recently. The blog host has kindly provided readers a lot of data and mathematics involved in meteorology and climatology. In part nine of the series on paleoclimates he gave a link to the NGRIP data set. And on later chapters also plots of insolation on various latitudes during the last glacial and interglacial. Previously I had found the EPICA core record.

The image is a bad example of wiggle matching combining the info mentioned. Now, I'm not claiming Toba eruption messed up the southern hemisphere circulation so the antarctica didn't get snow for 6000 years or that the Holocene climate maximum melted some 400 years of NGRIP data nor that EPICA borers dropped and shattered a five hundred year segment of the core or some other mishaps to have happened. I'm merely stating that inserting ~7000 years in this 123000 year section of paleoclimate data seems to get the records from different hemispheres to match more closely to each other. That's 5,7% of the whole set. That's still way better accuracy than many economists get.

If this was all the complications presented by the honorable host of Science of Doom in matching the ice core records to orbital theory of glacial inception (so called Milankovitch cycle theory) it would still be a hard run for anyone. But no, the orbital wiggles producing variations in the amount of insolation on different areas of the world during the ice age cycles do not seem to match well on the last glacial. I used the image for 40-50S insolation as the base for the image manipulation (no maths or curious theories here at all, at least yet) since I think the starts of the interglacials should have more to do with insolation over the oceans than over the continents. I've though organized the NGRIP and EPICA data to a same spreadsheet so I may start to manipulate those (in order to find out where the missing 6-7000 years might be). Anyway, as can be seen on the image, succeeding in this mathematical task with pretty wild theoretical assumptions would still leave matters unclear considering glacial and interglacial inception via orbital theory.

So the image is the initial artistic interpretation done before forming any hypothesis or doing any maths on a scientific problem of the exact effects of orbital variation on earth. Interested readers might want to take the quiz on

FTR (for the record): NGRIP was mutilated in 2 places and EPICA 8 times, in one case both had to be adjusted, because this is not science.


Oale said...
This comment has been removed by the author.
Oale said...

Note that applying this sort of adjustements to EPICA cores predating 123000 is not good to do (at least for me) for the lack of comparable data (maybe some oceanic core data from NH, or Lake Elgygytgyn sedimentation data goes as far with a similar accuracy as EPICA)

Oale said...

I've been noted that the dataset on EPICA I have might be one that's been since adjusted, so if you're planning to do something similar please make sure you have the correct datasets. As always, it's the responsibility of the reader to make sense of the texts and graphs here. On request, I might add some additional explanations. But only if I can make out what the question is. I'm a slow reader so long comments are likelier to get no answer.

Oale said...

there's a reason why antarctic and arctic temperature records do not need to match very closely in time, this is called the bipolar seesaw, a periodic restructuring of deeper atlantic currents due unstable ice shelves. this is explained better f.e. here:

this shoud cease to function if (or when) Greenland becomes ice free.

Oale said...

The interesting bit here is the EPICA core seems to precede the NGRIP core when the world gets warmer. (Just goes to show what a little bit of brutal unscientific torture of data can do)

I mentioned this here:,1611.msg96492.html#msg96492