re: #181 SixDegrees

Thanks. I’ll consider your offer, but at this point my concern is with whether or not the original raw data that HadCRUT3 was constructed from can ever be reassembled. Or, to be more succinct, if HadCRUT3 can be replicated from it’s original sources. This sort of examination of original raw data is commonplace, but my hopes that such examination will be possible are fading. It sounds as though CRU itself is unable to undertake such a reconstruction.

I haven’t seen any reason to reject CRU’s conclusions so far, but if I were citing their research I’d feel compelled to put a little star next to the reference, leading to an explanation of the circumstances, i.e., probably valid, but cannot be reproduced.

My first look, I would say no. Hadcrut3 can’t be reproduced unless CRU has something that everyone is now looking for, and that is the actual physical locations for each temperature reading with in each 5 degree by 5 degree gird, with the link to the source of that reading (IE: NCAR, GHCN, NMS, NOAA and some misc sources).

CRU says they can’t connect the each data record to the actual physical location and the source anymore.

Why is that important, because hadcrut3 has been tweaked, modified and each data record is not necessarily the raw date record that it was in the beginning.

To currently reconstruct each raw reading, you need to know what was the source agency and the physical recording station with in the grid.

That is the issue.

They say they can’t do that.

So, you have 52.2 mb of data (I have the file) with monthly temperature readings from 1850 onward and they cannot reproduce the relationship of temperature reading -> agency source -> physical recording station in the grid.