On Jan 22, 7:50*pm, Yokel wrote:
On 22/01/2012 19:11, Weatherlawyer wrote:









On Jan 22, 11:44 am, *wrote:
It's more than just an academic exercise: the number of instances of 'X'
(whatever its called) is related to the departure of the 12hr average
extremes from the 24hr average extremes. For example, this month
(January) using Hurn data, the current 24hr average minimum is 2.9
against 4.0 for the 18-06 mean& *the 24hr average maximum is 10.9 vs.
10.4 for the 06-18Z. We've now had 9 such events so far this month.
In a more 'typical' month, with no, or just one or two, events then the
difference is close to zero, and not usually more than 0.2degC.
I managed to 'dig out' an old copy of a Met Office 'Climatological
Memoranda' dealing with mean and extreme temperatures over the UK. It
has a section dealing with the differences between:
24 hour (09-09) max& *min temperatures and
12 hour (09-21) day max& *(21-09) night min.

The results, following an investigation of 38 stations between 1957
and 1970, show that, on average,
in January and December the 24 hour Max is:
0.3 to to 0.4C higher than the 12 hour day max.

The differences for minima show that:
24 hour minima were typically 0.7C lower than the 12 hour (21-09) value during the same two months.

Meaning?

Meaning that if you have a really cold night it is likely to count for 2
days instead of only 1 if you only read your min-max thermometer at 0900
each day. *This is because for much of the winter under clear skies the
minimum temperature will be only an hour or so before that time (around
or just after sunrise after all the night's cooling).

Arrrrgh!!!

Beware the trapses.

Nobody loves us. We wants it, we needs it. Must have the precious.
They stole it from us. Sneaky little weathermen.
No.
Not masters...
Yes, ...precious, ...false!
They will cheat you, hurt you.

LIE!!

If you read your min-max thermometer twice each day, at 0900 and 2100,
the rise in temperature during the day will most likely cancel out the
preceding cold night.

Below are a recent set of my recorded minima (non-standard exposure, but
that doesn't really matter for this illustration). *These are taken by
re-setting the maximum and minimum indices at times as close to 0700 and
1900 as my social routine will allow, the maximum being 0700 to 1900 and
the minimum 1900 to 0700 *except* that if the temperature is still
falling at 0700 or still rising at 1900 then the minimum or maximum
reached in the next 12 hour period will be the one recorded. *I call
this the "arithmetic" maximum and minimum as it is as far as possible
the values you would get from looking at a graph of the day's temperature..

Jan 2012
12 * *0.2
13 * *-4.7
14 * *-7.8
15 * *2.3
16 * *-6.6
17 * *-7.9
18 * *-3.5
19 * *8.6

If I had just used the 0900 values each day, this same data set could
very well have been:

Jan 2012
12 * *0.2
13 * *-4.7
14 * *-7.8
15 * **-6.5*
16 * *-6.6
17 * *-7.9
18 * **-5.7*
19 * **7.0*

The three marked with * are now much lower than before because the
temperature had hardly started to rise after the previous cold night.
In the case of 19th Jan, which was a really mild night, the "0900-0900"
recorded minimum could actually have been much lower than the 7.0C I
have put in because you might have picked up the tail end of the 17 / 18
cold night and not the 18/19 mild night. *As it happened, the mild air
arrived in time to send the temperature well above freezing by 0700 on
18th in spite of the frost earlier in the night.

The potential size of this effect is such that it could have to be taken
into account when comparing temperature records or looking for long-term
trends. *The "true" or "arithmetic" mean will actually give you the
highest minimum temperatures - if a cold day is followed by a mild night
then there will be cases when the 12-hour periods interpreted exactly
and without the "allowance" I make will also give lower minima that my
methods, but not so low as the "24 hour" protocol.

In the summer part of the year, the minimum temperature would be
expected to be many hours before the 0900 standard, allowing the rise in
temperature by then to effectively "re-set" the minimum record and so
greatly reduce this error. *As can be seen, the magnitude of this
difference between protocols will vary both by season of the year and by
weather type, so quoting an annual figure for the difference is pretty
much meaningless.

Now we have continuously recording thermometers widely available, some
thought may need to be given as how to properly define the maximum and
minimum temperatures, given that the "traditional" definition appears to
be rather wanting. *Over to those who collect and analyse such records....

Those who are going to be analysing such records will have been onto
that as soon as they went digital; I assume it was a fair bit ago.

Unfortunately they will have probably given that bone to the bods in
back rooms in East Anglia.

Pity that.

For those that do manage to view their data from the night time, I
hope you will pay it more attention.

You don't have any friends;
Nobody likes you!
I'm not listening...

I'm not listening...



Master looks after us now. We don't need you ...anymore.