Richard Dixon wrote:

Thanks all for your comments. The reason why I ask is that I'm
looking at some data from a remarkable 6000 year simulation (100 lots
of 1951-2011) but we only have the 10 metre wind (not gust) from the
model to look at. It's not at a resolution that I expect it to
resolve the really, really intense narrow wind events but enough to
get some fairly OK structure and wanted to sense-check against what
may have been observed in the real world.

Richard

Most model wind speeds are probably best considered as 1-hour mean
winds as it is unlikely that the model will be able to resolve
variations in wind speed over shorter durations. Any individual
measured 10-min mean may be higher or lower than the 1-hour mean within
which it is embedded.

--
Norman Lynagh
Tideswell, Derbyshire
303m a.s.l.
https://peakdistrictweather.org
Twitter: @TideswellWeathr

On Tuesday, 23 January 2018 19:41:57 UTC, Norman Lynagh wrote:

Most model wind speeds are probably best considered as 1-hour mean
winds as it is unlikely that the model will be able to resolve
variations in wind speed over shorter durations. Any individual
measured 10-min mean may be higher or lower than the 1-hour mean within
which it is embedded.

We're discussing this at the moment and the meaningfulness of a model "number". For starters the timestep is 20 minutes for this model which could be a starting point for averaging: but is an instantaneous model output an average...?!

Richard

Richard Dixon wrote:

On Tuesday, 23 January 2018 19:41:57 UTC, Norman Lynagh wrote:

Most model wind speeds are probably best considered as 1-hour mean
winds as it is unlikely that the model will be able to resolve
variations in wind speed over shorter durations. Any individual
measured 10-min mean may be higher or lower than the 1-hour mean
within which it is embedded.

We're discussing this at the moment and the meaningfulness of a model
"number". For starters the timestep is 20 minutes for this model
which could be a starting point for averaging: but is an
instantaneous model output an average...?!

Richard

I'm not sure that the time-step really matters. Surely what the model
is producing is a mean wind speed, albeit an instantaneous one, without
really trying to resolve the inevitable variations that occur over
short periods of time from tens of minutes downwards. Perhaps best to
consider the values as overlapping 1-hour means generated every 20
minutes.

For example, I doubt if the model will show much change from one
time-step to the next at any individual grid-point whereas the actual
instantaneous wind at the same grid-point may vary greatly between the
same time-steps.

--
Norman Lynagh
Tideswell, Derbyshire
303m a.s.l.
https://peakdistrictweather.org
Twitter: @TideswellWeathr

On Tuesday, 23 January 2018 17:19:19 UTC, wrote:
On Tue, 23 Jan 2018 07:46:39 -0800 (PST)
Richard Dixon wrote:

Bit of random question for you all on here (which I'll also post the Weather
and Climate group).

Just interested to know what the highest *sustained* (i.e. mean) wind you've
seen is in and around the UK/Ireland. Doesn't have to be on land...


I've experienced 1 knot higher than what Graham says because my weather is
always better than his :-)

Seriously though, I don't know. However, I do know that on my first observing
shift on my own at RAF Finningley in 1970 I recorded a land gale with a mean
wind of 34 knots. Had to put a beaufort letter "g" in the remarks column. I
remember thinking at the time "my first gale", never officially recorded one
since. Strongest wind experienced here in Haytor was on Hay Tor summit in 2004
when I couldn't stand upright. Also on Mam Tor in the Peak District 2 years ago
when we were on our hands and knees walking on the ridge, so at least a force
10. I also remember Lerwick observatory in 1980 a severe gale force 9 meaning
circa 45 knots, locals called it a "wee breeze"!

How to wake up a sleeper?

Anyone going to mention that the gale warning extends from the coast of Labrador in the Davies Straight to within 5 degrees of Ireland?

On Tuesday, 23 January 2018 21:08:13 UTC, Norman Lynagh wrote:

I'm not sure that the time-step really matters. Surely what the model
is producing is a mean wind speed, albeit an instantaneous one, without
really trying to resolve the inevitable variations that occur over
short periods of time from tens of minutes downwards. Perhaps best to
consider the values as overlapping 1-hour means generated every 20
minutes.

My only issue here is why you chosen one-hour means here as the likely equivalent output that the model is creating? Why no 10-minute means, for example?

Richard

On Tuesday, 23 January 2018 21:53:13 UTC, Weatherlawyer wrote:

Anyone going to mention that the gale warning extends from the coast of Labrador in the Davies Straight to within 5 degrees of Ireland?

Err, no-one, because we're talking about something different?

Richard

Richard Dixon wrote:

On Tuesday, 23 January 2018 21:08:13 UTC, Norman Lynagh wrote:

I'm not sure that the time-step really matters. Surely what the
model is producing is a mean wind speed, albeit an instantaneous
one, without really trying to resolve the inevitable variations
that occur over short periods of time from tens of minutes
downwards. Perhaps best to consider the values as overlapping
1-hour means generated every 20 minutes.

My only issue here is why you chosen one-hour means here as the
likely equivalent output that the model is creating? Why no 10-minute
means, for example?

Richard

As you suggested earlier in the thread the model is actually producing
instantaneous values of mean wind speed. If the model was configured to
produce values every 5 seconds then at each grid-point there would be
720 values produced in an hour. At any individual grid-point, in most
cases, the 720 values are all likely to be very similar. On the other
hand, if measured data are available for the location of the same
grid-point and if instantaneous values are measured every 5 seconds
there would almost certainly be great variablity amongst the 720
individual values because of the random effects of mechanical
turbulence. If the measured data were distilled into 1-min means and
10-min means there would still be a lot of variability, albeit more so
with the 1-min means than the 10-min means but significant random
variability is still present in 10-min means. By the time you get to
the 1-hour mean much of the variability is removed from the measured
data and the measured 1-hour mean is more directly comparable with the
values from the model i.e. the model is likely to be producing values
that are representative of the measured 1-hour mean.

The 1-hour mean is a robust statistic as it removes all of the random
short-term fluctuatiuon caused by mecanical turbulence. It is the
1-hour mean that is used for climatological purposes.

I was recently involved in a matter in which downtime statistics had
been generated based on winds produced by a numerical model.
Operationally, it was found that the actual downtime was much greater
than the statistics had predicted. When I looked into it, I found that
operations were sensitive to winds exceeding the downtime threshold for
only a few minutes at a time. It was, of course, a case of comparing
apples with oranges. The model winds were not representative of the
short duration wind increases to which the operation was sensitive.

I accept that there's a fair amount of subjectivity in this but
practical experience suggests that model winds are best assumed to be
representative of 1-hour means.

--
Norman Lynagh
Tideswell, Derbyshire
303m a.s.l.
https://peakdistrictweather.org
Twitter: @TideswellWeathr

In message ,
Richard Dixon writes
Just interested to know what the highest *sustained* (i.e. mean) wind
you've seen is in and around the UK/Ireland. Doesn't have to be on
land...

I'm not sure whether lying in bed in the early hours while the wind was
raging outside counts as having "seen" it. If so, then it would be the
Great Storm of October 1987. The Burns' Day Storm of January 1990 would
come close.
--
John Hall
"Hegel was right when he said that we learn from history
that man can never learn anything from history."
George Bernard Shaw (1856-1950)