"Alastair McDonald" k wrote
in message ...

Are you interpolating the values you do not have? Is the interpolation
based

Hi Alastair,

Yes, the system will provide a running average since the start of the month,
so in essence the first couple of days will fluctuate wildly (more so during
Summer due to diurnal variation) then start to settle down as the month
progresses.

The system will interpolate for figures which have not been recorded, so for
a 9 hour period, if the figures 3c,3c,4c,4c,**,4c,5c,5c,5c were recorded
then it would fill in the missing figure as 4c. The system itself updates
every 3 hours but will take into consideration the recordings sent in during
the previous 2 (hope that makes sense).

It's the same algorithm used to interpolate on Metcheck for areas between
the model grid points.

I'll run the Metcheck CET system for this month and next, see how it
performs, if it looks good then I'll get some feedback and see how it can be
improved.

Cheers,

A

The system itself updates
every 3 hours but will take into consideration the recordings sent in during
the previous 2 (hope that makes sense).

Andrew, it makes sense in terms of understanding what you are saying, but it
does not make sense in terms of the CET being a mean temperature (either mean
daily or mean monthly). Your 'instantaneous updates' may satisfy our hunger for
obtaining near-real time info. from the web but they should be daily updates of
mean temperature. Readers of your site who are not familiar with the background
of the CET may become very confused when they later hear of differing estimates
of the real CET, whether from Philip Eden or the MO or anyone else. .

You are half way to creating a new England and Wales temperature series - a 3
hourly update of temperature (and by detaching it from a CET substitute there
would be no need to calculate it as a moving average from the start of the
month.).

Julian Mayes, West Molesey, Surrey.

"JJCMayes1" wrote in message
...
The system itself updates
every 3 hours but will take into consideration the recordings sent in
during
the previous 2 (hope that makes sense).

Andrew, it makes sense in terms of understanding what you are saying, but it
does not make sense in terms of the CET being a mean temperature (either
mean
daily or mean monthly). Your 'instantaneous updates' may satisfy our hunger
for
obtaining near-real time info. from the web but they should be daily updates
of
mean temperature. Readers of your site who are not familiar with the
background
of the CET may become very confused when they later hear of differing
estimates
of the real CET, whether from Philip Eden or the MO or anyone else. .

You are half way to creating a new England and Wales temperature series - a
3
hourly update of temperature (and by detaching it from a CET substitute
there
would be no need to calculate it as a moving average from the start of the
month.).

Julian Mayes, West Molesey, Surrey.

Hi Andrew,

I agree with Julian. What I think you should do is add some more values to
your page. First, you need the daily CET for the previous day, and the monthly
CET for the previous month as calculated by you. I would like to see how they
compare with the average official CET for that day and for the month. You are
showing the current monthly one already. You could also add the cumulative CET
for this month upto and including the previous day. AISI your present value
has two errors; diurnal and monthly. I can comprehend the monthly error, but
the diurnal error makes your results strange at the start of the month and
uninteresting at the end.

It is a great idea to have a site where there is an istantaneous comparison of
the current CET with the average but I think you have made it too instaneous!

HTH,

Cheers, Alastair.