On Friday, August 14, 2015 at 7:29:21 PM UTC+1, Dave wrote:
Official Teignmouth Met Office Station rainfall 24hrs to 10am 14-Aug-15 was 50.1 mm or 1.972 inches in old money.

I can only remember one or two other occasions when 50mm has been exceeded.

50mm has a return period of 28 months in Penzance (based on 23 years of 09:00 to 09:00 records, so not a particularly rare event here.

Some years have seen 50mm recorded on 2 occasions, a mixture of winter & summer events.
1997 saw 53.1mm 7/11 & 52mm 3/8
2002 saw 84.5mm 13/11 & 62.5mm 31/12
2009 saw 65.2mm 24/4 & 51.8mm 16/7

My most notable rainfall event was when I recorded 90.2mm in 2 hours on 2/6/92 in Wootton Bassett (pre royal!) 92mm was recorded at Purton, 2 miles way, virtually none at RAF Lyneham. It made it into the MetO hydrometeorological bulletin. I even got a visit from the MetO (1 person. not the entire organisation!) and was handed satellite images (a novelty in those days) and a copy of the bulletin, a very pleasant surprise.

Graham
Penzance

Thanks Graham, I think the occurance of +50mm events in Penzance will be more frequent in Penzance than Teignmouth.

With Teignmouth being in the rain shadow of Dartmoor, we neec special synoptic setups for large rainfall totals in Teignmouth. Today was one of those, with a starionary front/low triple point in Lyme Bay.

"xmetman" wrote in message
...
On Friday, 14 August 2015 09:25:32 UTC+1, wrote:
"xmetman" wrote in message
...
My memories as an observer on a wet day or night shift, were of hourly
observations where I reported total hourly rainfall typically in the
range
0.2 to 2 mm per hour. Admittedly my memory isn't fantastic, but nowadays
if
it rains (and I know I live in Devon now) it pours!

I do have an automatic weather station to watch rainfall intensities
which
might skew things, but as an observer we always had the tipping bucket
gauge that gave off an audible click in the office as the counter
updated,
so it's not that I'm any more attentive than I was back then. Don't get
me
wrong, there were many occasions in showers or thunderstorms when you
could
report closer to 10 mm in a single hour, and I don't think the frequency
of
those events has changed. It's the other slow-moving frontal events I am
thinking of, when instead of reporting several hours of slight (0.6
mm/hr)
occasionally moderate rain (0.6 to 3.9 mm/hr), and perhaps an hour of
heavier rain (=4.0 mm/hr) as the front passed through, but in recent
years
all you just seem to get is a succession of hours of heavy rain.

Perhaps it's that I'm finally coming to terms (after 12 years) with how
rain falls in this part of the country and it's just taken me a long time
to notice it!
================================

Yes you must be.

===================
Anyway here is my best guess of rainfall accumulations across the UK for
yesterday 0700 and today at 0645 UTC (13/14 August 2015). I must say the
total is spot on for here in mid-Devon. There is a sharp line running
down
through the British Isles marking the extent of the rain, and it also
captures a number of drier pockets in East Anglia, West Wales and Central
southern England. see http://wp.me/p3yVic-1hS
===================

Haytor is 51mm. How are you estimating your totals Bruce. Surely you are
not
assuming dwell times for the whole of the radar return? E.g if you have
quarterly reports then if you use a dwell time of 0.25 hours then you will
normally over-estimate. We found that out in the Nowcasting department way
back in the 90s when you were in the Office. Better to use 5 minute or
even
1 minute radar data but there are more sophisticated algorithms that look
at
the standard deviation of returns over the whole period and essentially
fit
an accumulation model to the data. Yours is an excellent poor man's
representation of the data and does give a quick overview but I'm worried
that your readers wil take it as gospel when it could be so much better?
Do
you reference your methodology anywhere?


Will

I have a very simple approach - I keep three arrays of intensities to match
the rainfall intensities low/medium/high:

TIntensitiesLow : array[0..8] of double = (0, 0.01, 0.5 , 1 , 2, 4, 8,
16, 32);
TIntensitiesMid : array[0..8] of double = (0, 0.25, 0.75, 1.5, 3, 6, 12,
24, 48);
TIntensitiesHigh : array[0..8] of double = (0, 0.5, 1, 2 , 4, 8, 16,
32, 64);

As you can see the high intensity array matches the lower of the seven mm/hr
values displayed on the weather radar page. I suppose I could have added
another couple of arrays to cover the mean and the max of the range as well.
By default I use the medium setting which as you can see I've reduced from
the lower range value by around 25%. This is what I normally use when
running the accumulation, I look up the pixel colour and take a quarter of
the 60 minute intensity and add it to my running accumulation for each
pixel.

As you point out if we had more frequent weather radar images estimates
would be more exact. Increasing the size of the image would enable to show
more detail, but to be realistic this is not a big issue, because an 18 hour
accumulation takes about 25 seconds to run and create a tabulated list of
accumulations for 3100 places across the British Isles, which makes it a
usable tool. The big pain is that the Met Office don't produce a single
weather radar image (they do in the DataPoint service but the last time I
looked it was not hi-res and has a weird projection), so I have to stitch a
grid of tiles together before I can even display an image in a viewer.

As for fooling the user, I always make it clear in my blog or in the images
that I generate the accumulation is an estimate. I see it as being no more
misleading than an original image, which at times can be very misleading,
with precipitation from medium level cloud that never reaches the surface,
bright banding in winter, and 'spoking' from numerous sites especially
Chenies caused by nearby trees (which I have blogged about).

I noticed recently that Weather Online are doing a similar thing to what
I'm doing with accumulations, they may have beat me to it, but they can't
match my tabulated list of 3,100 wettest places (although it may be in the
premium version) which as far as I know is unique, and all generated from
what is after all a very simple application.

What beats me is the Met Office have accumulations from the Weather Radar
that they generate all the time. I asked for them to be released so we could
display them in NAMIS and Nimbus but hit a blank wall. They may have been
available on Horace or as far as I know Visual Weather but one hopes that
must have changed by now. In my opinion they would make an invaluable
resource for the Wet Bench team. Just before I left I did some work for that
team and proposed that we combine estimated hourly accumulations gleaned
from weather radar, with observational SREW reports (either 60 minute or
even more frequent ones because they are available) from SAMOS sites,
climate stations and all the hundreds of automatic gauges the Environment
Agency have, it never happened. Perhaps it has now. If you've got the data
you have to visualise it to make the best sense of it.
====================

Thanks Bruce. I think the biggest problem for you is the poor time
resolution. 15 minutes will be OK for prolonged orographic rainfall but for
frontal, and especially convective, you are going to be a long way out at
times. There are ways of getting round it to some extent but prob. not worth
the effort given that 1 minute radar data (single site) are available and
certainly 5 minute composites. You are doing some excellent work for the
anorak community, but sadly the corporate world will continue to go its own
way where just one thing really matters - MONEY.

Will
--
http://www.lyneside.demon.co.uk/Hayt...antage_Pro.htm
Will Hand (Haytor, Devon, 1017 feet asl)
---------------------------------------------