To: Phil Jones <firstname.lastname@example.org>
Date: Tue, 12 May 2009 08:54:44 +0100
there may be some money this FY, substantial sums. Management here are
casting around for ideas. As its to be spent this FY its largely going
to be consultant work as we never have a cats chance in hell of
recruiting on that timescale. What resource do you think we could
contract from CRU (you, Harry, others?) for doing a CRUTEM4 which I
would maintain had two aims ...
1. Rescue and incorporation of recent data (I'm pinging NCDC too to see
what they could do vis-a-vis collating and sending the non-wmo US
stations and other data you may not have ... their bi-lats may have sig.
extra stations for Iran, Aus, Canada etc.)
2. A more robust error model that led to production of a set of equi-
probable potential gridded products (HadSST3 will do simnilarly so we
could combine to form HadCRUT4 equi-probable). This error model
determination would ideally be modular so that we could assess how wrong
our assumptions about the error would have to be to "matter" and what
error sources are important for our ability to characterise the long-
term trend (trivially these will be the red noise I know but then most
people seem blind to the trivial sadly ...). The HadCRUT3 paper clearly
started well down that path but a recent paper I had the displeasure of
reviewing on my way back from WMO shows its poorly understood
(deliberately so in this particular case ...).
We have a meeting Thursday. If it passes muster there we'll put it to
DECC and see what happens. No promises.
This would mean we'd have HadCRUT4 which would be HadSST3 + CRUTEM4 each
with more data and better error models well before AR5 which seems
Mr. Fraudit never goes away does he? How often has he been told that we
don't have permission? Ho hum. Oh, I heard that fraudit's Santer et al
comment got rejected. That'll brighten your day at least a teensy bit?
Peter Thorne Climate Research Scientist
Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB
tel. +44 1392 886552 fax +44 1392 885681