On Tue, 6 Jan 2004 14:53:22 -0000, "Philip Eden"
philipATweatherHYPHENukDOTcom wrote:


"Dave.C" wrote in message
...
We usually get a long thread every year on this subject. Philip or Martin
step in and all becomes clear for about 10 minutes. Then when it happens
the
next year I'm still confused.

Try:

http://groups.google.com/... huge link snippified

Philip Eden :-)

Thanks Philip, I remember that thread now. I did look, but failed to
find it. I've read it now.

What concerns me is not so much what time of day the readings are
taken, it doesn't seem to be a great problem if readings are
attributed to the "wrong" day as long as none are missed altogether.
But with the "night min/day max" method (18-06-18), whatever times are
chosen, is it not true that some daily extremes can be completely lost
to the long term record? Even if 21-09-21 or other times are used
instead?

Long winded example: It's 30th January in Fareham and until this
morning, it's been mildish all month. Yesterday's min was +6C and the
max +8C. But the temperature has been dropping this morning since 3 am
as much colder air is advected South and is down to +3C by 06GMT. It
continues to get colder and by 10 am, is down to -2C and frosty. But
the sky is now clearing and the wind is dropping out. Ooh, I think to
myself excitedly, we've just had our first air frost of the month and
lowest minimum of the winter! At last!

By 11am, the sun is shining brightly and the afternoon maximum
eventually creeps up to +6 degrees in my old wooden Stevenson screen.
But the cold snap is short lived here, of course. Tonight (30th/31st)
has a minimum of +4C, occurring at 19GMT, as cloud increases ahead of
a warm front from the West and the temperature starts to rise.
Followed by days and days of mild Westerlies.

Is it not true that using the 18-06-18 method (as in SYNOPS), the
minimum today (30th) would be recorded as +3 (at 06 GMT) and
tomorrow's minimum (31st) as +4 (at 19 GMT)? My month's minimum, and
the one and only air frost, has been lost to the official records
forever (not just pushed into the wrong day or month) hasn't it?

If true, that's what concerns me most about the 18-06-18 method. It
seems to strengthen greatly the case for the seemingly more logical
method of recording one max and one min for each successive 24 hour
period. Far better to have the maximum or minimum temperature
attributed to the wrong day or month than to have them lost
completely, I'd say.

Unless, of course, I've got it completely wrong in me advancing old
age...

--
Dave