date: Tue Nov 4 09:38:22 2008
from: Phil Jones <p.jonesatXYZxyz.ac.uk>
subject: My minutes from yesterday
to: C G Kilsby <c.g.kilsbyatXYZxyzcastle.ac.uk>
Kathryn had someone taking minutes yesterday - so you'll be getting
the official ones in a day or two. Below is what the WG has to do, plus
a few more observations - some things that won't be in the minutes but were odd.
I will try and find out more when in Exeter Thursday and Friday.
First on the mean T and range. Colin is going to liaise with Vaz
about producing the offsets from what the WG currently produces for T and R
compared to what it probably ought to do compared to NCIC. NCIC is the
MOHC gridded monthly UK climatologies - the updated data that was in
UKCIP02. The effect of doing this will get the averages for the UK sites
to be the same as NCIC for the means. Colin's not sure how long this will
take, but we need to do it. It should also improve the extreme counts.
The problem relates to us using only 115 stations to fit the WG and then
spatially interpolating these to get all 5 by 5km squares. Stations that are
frost hollows like Yeovilton will get improved, and also stations like LHR
and Ringway which are slightly warmer than neighbouring stations will also. Geoff
didn't understand why we need to do this, but he finally cam around. Colin
has done this for Manchester, Sheffield and London in Scorchio, and can send
you a series of plots if interested.
The final, final set of Change Factors will be sent by Dave to Ag by Dec 1.
Ag will then run his scripts on these and get stuff to you about 5-7 days later.
When you get these, we need to check the daily T variance to see of there are
still silly numbers. Dave apparently says they will be OK (he wasn't there,
but Geoff says he's sorted this. He may have done this by setting the changes to
1 - we will have to wait and see). Essentially they are now happy with what
they had done before - they understand why things are happening. They are
supposed to have corrected daily T var and also Pdry. So check these first
when they arrive! So things won't be that different in the revised set.
So in the meantime, you should continue your porting work to the Newcastle cluster.
Colin will produce the WG/NCIC offsets for Tmean and Range.
The PMG were impressed with the extremes when I explained what you could
and couldn't compare. We discussed what extremes we should show as maps
and what we could show as numbers - like the Tables you sent down. The maximum
number as maps should be 4, 2 for T and 2 for rainfall. I'll leave rainfall up to you
but Rmed and the dry spells above 10 days seem likely candidates. For T we should
do frost days (even though this isn't that extreme) and one high T value. Finally they
seem to understand that absolute thresholds aren't that good. We can drop a
heatwave measure. We talked about going back to doing the 90th percentile if Tx.
We can discuss this more during November. I'd also like to continue to do the
Annual highest Tx, as this is one we can directly compare with one Dave has
done - and show this if it agrees well. This should keep Lenny quiet!
BADC said they could plot the results - using the same colour bars they are
using for the. I guess you'll want to plot the maps anyway. Roger wanted us to
plot the deltas (future minus control) which I think is a good idea. Geoff wanted
us to plot future as absolute. Again we need to think about this. My thoughts
are to plot control 50% and the delta (there will be three of these for 10/50/90.
We can do many more extremes for the 8 sites you've done so far. This would
give people things that they could reproduce, as they won't be able to reproduce
the maps. Again we can discuss this more. We could expand on what you
currently have or limit it to these. In any plots of the 8 sites - don't join up
series. Just plot the symbols against the 8 sites. Some people tried to
interpolate between the locations !!!!!!
I said we could also add the 10/90th percentiles from the Obs. We do this
by making the 10th &ile the 3rd value when ranked and the 90th %ile the 27th
(approximately). Doing this will ensure the WG control is within the obs range.
Some other things
1. We are expected to go to the review meeting on Jan 13/14 in Reading. We might
be asked to make a presentation and/or answer questions. Geoff sees this as
being very useful and getting the responses done fairly quickly. Kathryn has been
asked to try reduce the time between that meeting and the launch. The review
period is Nov 28 - Dec 16. After Dec 16, we will get the comments back, and be
expected to respond to as many of the easy ones before Jan 13/14.
2. The HC have told James Murphy and Dave Sexton not to talk to Roger till
after Dec 1. He is a bit fed up with this - he can't understand some of their comments.
Anna is sending around the worked examples and we've all got too much else
on to have time to look at these. Only Geoff and Kathryn are responding, which
isn't ideal. I've not got time and I guess you haven't either.
3. UKCIP is staying in DEFRA. Almost all other climate work is moving to DECC.
DECC is apparently trying to set up something which seems quite similar to
UKCIP, so UKCIP are quite worried. Not sure what is going on - so this is where
I hope to find out more in Exeter talking to Bob Watson and/or Chris Sear.
4. We need to liaise with Ag on the TD and the User Guidance. I'll email Ag later
to get a copy of this. Kathryn told me where it is on a password protected site,
but I can't find that password anywhere. For the TD, this will be to ensure that
it reproduces what will be in the extreme tables. On the User Guidance it is
ensure that the text about how to use the model variants is still there. I still
think that those at the PMG don't understand that this guidance (from our
Newcastle meeting in 2007!) is important. Geoff says why isn't it in our WG
5. I reiterated that PET will be MORECS. Explained again why we can't do MOSES.
6. We need to tell Roger who will be manning our helpdesks for the launch. We need
to do this fairly soon. UKCIP want to have someone visit all the first points of contact.
UKCIP will farm these out, but try and respond to as many as possible with their
120 FAQs. Yes they have this many. The HC haven't been through and signed off on
many of theirs - another bone of contention for UKCIP.
7. Finally Geoff and Roger want some examples of the output from the WG. Maybe
Colin can send some daily and hourly stuff for LHR. Doesn't need to be much - a few
months or days. Geoff says why don't we plot an example of the output! I said
this would be silly as it is just weather data.
8. Kathryn has been surprised that none of us (HC, BADC, UKCIP, Newc, UEA) have asked
for more money given the delay. Apparently we can. We may not get it, but if we
can think of a good reason we might. Need to get something in soon and a revised
contract by March. DEFRA is expecting to take a big hit in the next spending review,
so we are very unlikely to get anything after March 2009.
We can extend the contract end date, but need to inform them of this.
In Exeter the rest of the week - here today if you want to call.
Apologies if this is a but disjointed, but based on my notes.
Prof. Phil Jones
Climatic Research Unit Telephone +44 (0) 1603 592090
School of Environmental Sciences Fax +44 (0) 1603 507784
University of East Anglia
Norwich Email p.jonesatXYZxyz.ac.uk