[Met_help] [rt.rap.ucar.edu #46697] History for wind direction (point-stat?)
RAL HelpDesk {for John Halley Gotway}
met_help at ucar.edu
Mon May 16 10:41:59 MDT 2011
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hi,
I'm trying to verify some wind speed an direction forecasts, using observed
(surface) windspeeds.
The forecasts contain windspeed and the U and V components (and hence
direction can be derived). However, the observations only contain a
windspeed and wind direction (degrees).
When I try running point-stat for wind direction I get:
"wind direction field may not be verified using point_stat"
I understand (from the tutorial and web pages etc) that wind direction can
be verified using STAT-Analysis (with VL1L2 from the point-stat tool, which
is run on the U and V winds). However, my obs don't have a U and V
windspeeds, so is it not possible to verify the wind direction?
Of course it would be possible to convert the observed windspeed into U and
V components (with knowledge of the wind direction), and then run point-sta
and STAT-Analysis on those, but I was wondering if there was another way
around?
Thanks,
Jason
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46697] wind direction (point-stat?)
From: John Halley Gotway
Time: Wed May 04 21:51:14 2011
Jason,
Unfortunately, no, there is currently no way in MET for you to compare
forecast wind directions directly to observed wind directions. You're
correct, that you could back out the U and V values from the wind
speed
and direction in your observations, but I do see why that's not very
desirable.
I do think our support for verifying wind direction is a little
lacking,
but we've had trouble defining what statistics are most appropriate
for
wind direction.
Thanks,
John Halley Gotway
met_help at ucar.edu
>
> Wed May 04 17:11:38 2011: Request 46697 was acted upon.
> Transaction: Ticket created by marshj77 at gmail.com
> Queue: met_help
> Subject: wind direction (point-stat?)
> Owner: Nobody
> Requestors: marshj77 at gmail.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
>
>
> Hi,
> I'm trying to verify some wind speed an direction forecasts, using
> observed
> (surface) windspeeds.
> The forecasts contain windspeed and the U and V components (and
hence
> direction can be derived). However, the observations only contain a
> windspeed and wind direction (degrees).
> When I try running point-stat for wind direction I get:
> "wind direction field may not be verified using point_stat"
>
> I understand (from the tutorial and web pages etc) that wind
direction can
> be verified using STAT-Analysis (with VL1L2 from the point-stat
tool,
> which
> is run on the U and V winds). However, my obs don't have a U and V
> windspeeds, so is it not possible to verify the wind direction?
>
> Of course it would be possible to convert the observed windspeed
into U
> and
> V components (with knowledge of the wind direction), and then run
> point-sta
> and STAT-Analysis on those, but I was wondering if there was another
way
> around?
> Thanks,
> Jason
>
------------------------------------------------
Subject: wind direction (point-stat?)
From: jason marsh
Time: Thu May 05 13:44:34 2011
Hi John,
Thanks for your response.
I'll maybe go down the route of splitting into U and V components
then.
Thanks for you continued support of MET.
Jason
On Wed, May 4, 2011 at 5:51 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Jason,
>
> Unfortunately, no, there is currently no way in MET for you to
compare
> forecast wind directions directly to observed wind directions.
You're
> correct, that you could back out the U and V values from the wind
speed
> and direction in your observations, but I do see why that's not very
> desirable.
>
> I do think our support for verifying wind direction is a little
lacking,
> but we've had trouble defining what statistics are most appropriate
for
> wind direction.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
> >
> > Wed May 04 17:11:38 2011: Request 46697 was acted upon.
> > Transaction: Ticket created by marshj77 at gmail.com
> > Queue: met_help
> > Subject: wind direction (point-stat?)
> > Owner: Nobody
> > Requestors: marshj77 at gmail.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
> >
> >
> > Hi,
> > I'm trying to verify some wind speed an direction forecasts, using
> > observed
> > (surface) windspeeds.
> > The forecasts contain windspeed and the U and V components (and
hence
> > direction can be derived). However, the observations only contain
a
> > windspeed and wind direction (degrees).
> > When I try running point-stat for wind direction I get:
> > "wind direction field may not be verified using point_stat"
> >
> > I understand (from the tutorial and web pages etc) that wind
direction
> can
> > be verified using STAT-Analysis (with VL1L2 from the point-stat
tool,
> > which
> > is run on the U and V winds). However, my obs don't have a U and V
> > windspeeds, so is it not possible to verify the wind direction?
> >
> > Of course it would be possible to convert the observed windspeed
into U
> > and
> > V components (with knowledge of the wind direction), and then run
> > point-sta
> > and STAT-Analysis on those, but I was wondering if there was
another way
> > around?
> > Thanks,
> > Jason
> >
>
>
>
>
------------------------------------------------
Subject: wind direction (point-stat?)
From: jason marsh
Time: Wed May 11 18:26:56 2011
Hi John,
I've managed to produce the U and V components from windspeed and
direction,
and run them through the ascii2nc tool with appropriate grib codes.
When I try to run point stat, it fails with:
ERROR: process_obs_file() -> when computing vector winds, each UGRD
observation must be followed by a VGRD observation with the same
header and
at the same level
I tried switching the order of U and W in the point-stat config file,
but
then it does not produce the vector partial sums file (VL1L2) required
for
wind direction verification in stat-analysis.
Is there a way around this?
Thanks,
Jason
On Thu, May 5, 2011 at 9:44 AM, RAL HelpDesk {for jason marsh} <
met_help at ucar.edu> wrote:
> Hi John,
> Thanks for your response.
> I'll maybe go down the route of splitting into U and V components
then.
> Thanks for you continued support of MET.
>
> Jason
>
> On Wed, May 4, 2011 at 5:51 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
> > Jason,
> >
> > Unfortunately, no, there is currently no way in MET for you to
compare
> > forecast wind directions directly to observed wind directions.
You're
> > correct, that you could back out the U and V values from the wind
speed
> > and direction in your observations, but I do see why that's not
very
> > desirable.
> >
> > I do think our support for verifying wind direction is a little
lacking,
> > but we've had trouble defining what statistics are most
appropriate for
> > wind direction.
> >
> > Thanks,
> > John Halley Gotway
> > met_help at ucar.edu
> >
> > >
> > > Wed May 04 17:11:38 2011: Request 46697 was acted upon.
> > > Transaction: Ticket created by marshj77 at gmail.com
> > > Queue: met_help
> > > Subject: wind direction (point-stat?)
> > > Owner: Nobody
> > > Requestors: marshj77 at gmail.com
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697>
> > >
> > >
> > > Hi,
> > > I'm trying to verify some wind speed an direction forecasts,
using
> > > observed
> > > (surface) windspeeds.
> > > The forecasts contain windspeed and the U and V components (and
hence
> > > direction can be derived). However, the observations only
contain a
> > > windspeed and wind direction (degrees).
> > > When I try running point-stat for wind direction I get:
> > > "wind direction field may not be verified using point_stat"
> > >
> > > I understand (from the tutorial and web pages etc) that wind
direction
> > can
> > > be verified using STAT-Analysis (with VL1L2 from the point-stat
tool,
> > > which
> > > is run on the U and V winds). However, my obs don't have a U and
V
> > > windspeeds, so is it not possible to verify the wind direction?
> > >
> > > Of course it would be possible to convert the observed windspeed
into U
> > > and
> > > V components (with knowledge of the wind direction), and then
run
> > > point-sta
> > > and STAT-Analysis on those, but I was wondering if there was
another
> way
> > > around?
> > > Thanks,
> > > Jason
> > >
> >
> >
> >
> >
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46697] wind direction (point-stat?)
From: John Halley Gotway
Time: Wed May 11 22:22:27 2011
Jason,
When processing the vector components of wind in Point-Stat, we
adopted a
convention requiring that the U and V pairs that "go together" have to
be
listed sequentially, U followed by V. If we didn't require this, the
code
would waste a lot of time searching around looking for corresponding
U/V
pairs.
Hopefully this won't be too difficult to address. Take a look at your
ASCII input to the ascii2nc tool. Are the U/V pairs that go together
listed sequentially in that file? If not, please try ordering them in
that way, and rerunning through ascii2nc.
If they already are in order, I'll need to take a closer look. Could
you
send me a sample ASCII input file for ascii2nc, and I'll take a look?
Thanks,
John
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
>
> Hi John,
> I've managed to produce the U and V components from windspeed and
> direction,
> and run them through the ascii2nc tool with appropriate grib codes.
> When I try to run point stat, it fails with:
>
> ERROR: process_obs_file() -> when computing vector winds, each UGRD
> observation must be followed by a VGRD observation with the same
header
> and
> at the same level
>
> I tried switching the order of U and W in the point-stat config
file, but
> then it does not produce the vector partial sums file (VL1L2)
required for
> wind direction verification in stat-analysis.
>
> Is there a way around this?
> Thanks,
> Jason
>
> On Thu, May 5, 2011 at 9:44 AM, RAL HelpDesk {for jason marsh} <
> met_help at ucar.edu> wrote:
>
>> Hi John,
>> Thanks for your response.
>> I'll maybe go down the route of splitting into U and V components
then.
>> Thanks for you continued support of MET.
>>
>> Jason
>>
>> On Wed, May 4, 2011 at 5:51 PM, RAL HelpDesk {for John Halley
Gotway} <
>> met_help at ucar.edu> wrote:
>>
>> > Jason,
>> >
>> > Unfortunately, no, there is currently no way in MET for you to
compare
>> > forecast wind directions directly to observed wind directions.
You're
>> > correct, that you could back out the U and V values from the wind
>> speed
>> > and direction in your observations, but I do see why that's not
very
>> > desirable.
>> >
>> > I do think our support for verifying wind direction is a little
>> lacking,
>> > but we've had trouble defining what statistics are most
appropriate
>> for
>> > wind direction.
>> >
>> > Thanks,
>> > John Halley Gotway
>> > met_help at ucar.edu
>> >
>> > >
>> > > Wed May 04 17:11:38 2011: Request 46697 was acted upon.
>> > > Transaction: Ticket created by marshj77 at gmail.com
>> > > Queue: met_help
>> > > Subject: wind direction (point-stat?)
>> > > Owner: Nobody
>> > > Requestors: marshj77 at gmail.com
>> > > Status: new
>> > > Ticket <URL:
>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697>
>> > >
>> > >
>> > > Hi,
>> > > I'm trying to verify some wind speed an direction forecasts,
using
>> > > observed
>> > > (surface) windspeeds.
>> > > The forecasts contain windspeed and the U and V components (and
>> hence
>> > > direction can be derived). However, the observations only
contain a
>> > > windspeed and wind direction (degrees).
>> > > When I try running point-stat for wind direction I get:
>> > > "wind direction field may not be verified using point_stat"
>> > >
>> > > I understand (from the tutorial and web pages etc) that wind
>> direction
>> > can
>> > > be verified using STAT-Analysis (with VL1L2 from the point-stat
>> tool,
>> > > which
>> > > is run on the U and V winds). However, my obs don't have a U
and V
>> > > windspeeds, so is it not possible to verify the wind direction?
>> > >
>> > > Of course it would be possible to convert the observed
windspeed
>> into U
>> > > and
>> > > V components (with knowledge of the wind direction), and then
run
>> > > point-sta
>> > > and STAT-Analysis on those, but I was wondering if there was
another
>> way
>> > > around?
>> > > Thanks,
>> > > Jason
>> > >
>> >
>> >
>> >
>> >
>>
>>
>
------------------------------------------------
Subject: wind direction (point-stat?)
From: jason marsh
Time: Thu May 12 15:28:41 2011
Hi John,
As you suspected the problem was that the input into ascii2nc was
organized
by variable then time. I sorted by time, so U and V components were
sequential, and ran it through ascii2nc and point-stat and it worked
fine.
Thanks again for you help.
Jason
On Wed, May 11, 2011 at 6:22 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Jason,
>
> When processing the vector components of wind in Point-Stat, we
adopted a
> convention requiring that the U and V pairs that "go together" have
to be
> listed sequentially, U followed by V. If we didn't require this,
the code
> would waste a lot of time searching around looking for corresponding
U/V
> pairs.
>
> Hopefully this won't be too difficult to address. Take a look at
your
> ASCII input to the ascii2nc tool. Are the U/V pairs that go
together
> listed sequentially in that file? If not, please try ordering them
in
> that way, and rerunning through ascii2nc.
>
> If they already are in order, I'll need to take a closer look.
Could you
> send me a sample ASCII input file for ascii2nc, and I'll take a
look?
>
> Thanks,
> John
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
> >
> > Hi John,
> > I've managed to produce the U and V components from windspeed and
> > direction,
> > and run them through the ascii2nc tool with appropriate grib
codes.
> > When I try to run point stat, it fails with:
> >
> > ERROR: process_obs_file() -> when computing vector winds, each
UGRD
> > observation must be followed by a VGRD observation with the same
header
> > and
> > at the same level
> >
> > I tried switching the order of U and W in the point-stat config
file, but
> > then it does not produce the vector partial sums file (VL1L2)
required
> for
> > wind direction verification in stat-analysis.
> >
> > Is there a way around this?
> > Thanks,
> > Jason
> >
> > On Thu, May 5, 2011 at 9:44 AM, RAL HelpDesk {for jason marsh} <
> > met_help at ucar.edu> wrote:
> >
> >> Hi John,
> >> Thanks for your response.
> >> I'll maybe go down the route of splitting into U and V components
then.
> >> Thanks for you continued support of MET.
> >>
> >> Jason
> >>
> >> On Wed, May 4, 2011 at 5:51 PM, RAL HelpDesk {for John Halley
Gotway} <
> >> met_help at ucar.edu> wrote:
> >>
> >> > Jason,
> >> >
> >> > Unfortunately, no, there is currently no way in MET for you to
compare
> >> > forecast wind directions directly to observed wind directions.
You're
> >> > correct, that you could back out the U and V values from the
wind
> >> speed
> >> > and direction in your observations, but I do see why that's not
very
> >> > desirable.
> >> >
> >> > I do think our support for verifying wind direction is a little
> >> lacking,
> >> > but we've had trouble defining what statistics are most
appropriate
> >> for
> >> > wind direction.
> >> >
> >> > Thanks,
> >> > John Halley Gotway
> >> > met_help at ucar.edu
> >> >
> >> > >
> >> > > Wed May 04 17:11:38 2011: Request 46697 was acted upon.
> >> > > Transaction: Ticket created by marshj77 at gmail.com
> >> > > Queue: met_help
> >> > > Subject: wind direction (point-stat?)
> >> > > Owner: Nobody
> >> > > Requestors: marshj77 at gmail.com
> >> > > Status: new
> >> > > Ticket <URL:
> >> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697>
> >> > >
> >> > >
> >> > > Hi,
> >> > > I'm trying to verify some wind speed an direction forecasts,
using
> >> > > observed
> >> > > (surface) windspeeds.
> >> > > The forecasts contain windspeed and the U and V components
(and
> >> hence
> >> > > direction can be derived). However, the observations only
contain a
> >> > > windspeed and wind direction (degrees).
> >> > > When I try running point-stat for wind direction I get:
> >> > > "wind direction field may not be verified using point_stat"
> >> > >
> >> > > I understand (from the tutorial and web pages etc) that wind
> >> direction
> >> > can
> >> > > be verified using STAT-Analysis (with VL1L2 from the point-
stat
> >> tool,
> >> > > which
> >> > > is run on the U and V winds). However, my obs don't have a U
and V
> >> > > windspeeds, so is it not possible to verify the wind
direction?
> >> > >
> >> > > Of course it would be possible to convert the observed
windspeed
> >> into U
> >> > > and
> >> > > V components (with knowledge of the wind direction), and then
run
> >> > > point-sta
> >> > > and STAT-Analysis on those, but I was wondering if there was
another
> >> way
> >> > > around?
> >> > > Thanks,
> >> > > Jason
> >> > >
> >> >
> >> >
> >> >
> >> >
> >>
> >>
> >
>
>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #46697] wind direction (point-stat?)
From: John Halley Gotway
Time: Mon May 16 10:41:42 2011
Jason,
Glad that did the trick. I'll go ahead and resolve this ticket.
Thanks,
John
On 05/12/2011 03:28 PM, RAL HelpDesk {for jason marsh} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
>
> Hi John,
> As you suspected the problem was that the input into ascii2nc was
organized
> by variable then time. I sorted by time, so U and V components were
> sequential, and ran it through ascii2nc and point-stat and it worked
fine.
> Thanks again for you help.
> Jason
>
>
> On Wed, May 11, 2011 at 6:22 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Jason,
>>
>> When processing the vector components of wind in Point-Stat, we
adopted a
>> convention requiring that the U and V pairs that "go together" have
to be
>> listed sequentially, U followed by V. If we didn't require this,
the code
>> would waste a lot of time searching around looking for
corresponding U/V
>> pairs.
>>
>> Hopefully this won't be too difficult to address. Take a look at
your
>> ASCII input to the ascii2nc tool. Are the U/V pairs that go
together
>> listed sequentially in that file? If not, please try ordering them
in
>> that way, and rerunning through ascii2nc.
>>
>> If they already are in order, I'll need to take a closer look.
Could you
>> send me a sample ASCII input file for ascii2nc, and I'll take a
look?
>>
>> Thanks,
>> John
>>
>>>
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697 >
>>>
>>> Hi John,
>>> I've managed to produce the U and V components from windspeed and
>>> direction,
>>> and run them through the ascii2nc tool with appropriate grib
codes.
>>> When I try to run point stat, it fails with:
>>>
>>> ERROR: process_obs_file() -> when computing vector winds, each
UGRD
>>> observation must be followed by a VGRD observation with the same
header
>>> and
>>> at the same level
>>>
>>> I tried switching the order of U and W in the point-stat config
file, but
>>> then it does not produce the vector partial sums file (VL1L2)
required
>> for
>>> wind direction verification in stat-analysis.
>>>
>>> Is there a way around this?
>>> Thanks,
>>> Jason
>>>
>>> On Thu, May 5, 2011 at 9:44 AM, RAL HelpDesk {for jason marsh} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Hi John,
>>>> Thanks for your response.
>>>> I'll maybe go down the route of splitting into U and V components
then.
>>>> Thanks for you continued support of MET.
>>>>
>>>> Jason
>>>>
>>>> On Wed, May 4, 2011 at 5:51 PM, RAL HelpDesk {for John Halley
Gotway} <
>>>> met_help at ucar.edu> wrote:
>>>>
>>>>> Jason,
>>>>>
>>>>> Unfortunately, no, there is currently no way in MET for you to
compare
>>>>> forecast wind directions directly to observed wind directions.
You're
>>>>> correct, that you could back out the U and V values from the
wind
>>>> speed
>>>>> and direction in your observations, but I do see why that's not
very
>>>>> desirable.
>>>>>
>>>>> I do think our support for verifying wind direction is a little
>>>> lacking,
>>>>> but we've had trouble defining what statistics are most
appropriate
>>>> for
>>>>> wind direction.
>>>>>
>>>>> Thanks,
>>>>> John Halley Gotway
>>>>> met_help at ucar.edu
>>>>>
>>>>>>
>>>>>> Wed May 04 17:11:38 2011: Request 46697 was acted upon.
>>>>>> Transaction: Ticket created by marshj77 at gmail.com
>>>>>> Queue: met_help
>>>>>> Subject: wind direction (point-stat?)
>>>>>> Owner: Nobody
>>>>>> Requestors: marshj77 at gmail.com
>>>>>> Status: new
>>>>>> Ticket <URL:
>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46697>
>>>>>>
>>>>>>
>>>>>> Hi,
>>>>>> I'm trying to verify some wind speed an direction forecasts,
using
>>>>>> observed
>>>>>> (surface) windspeeds.
>>>>>> The forecasts contain windspeed and the U and V components (and
>>>> hence
>>>>>> direction can be derived). However, the observations only
contain a
>>>>>> windspeed and wind direction (degrees).
>>>>>> When I try running point-stat for wind direction I get:
>>>>>> "wind direction field may not be verified using point_stat"
>>>>>>
>>>>>> I understand (from the tutorial and web pages etc) that wind
>>>> direction
>>>>> can
>>>>>> be verified using STAT-Analysis (with VL1L2 from the point-stat
>>>> tool,
>>>>>> which
>>>>>> is run on the U and V winds). However, my obs don't have a U
and V
>>>>>> windspeeds, so is it not possible to verify the wind direction?
>>>>>>
>>>>>> Of course it would be possible to convert the observed
windspeed
>>>> into U
>>>>>> and
>>>>>> V components (with knowledge of the wind direction), and then
run
>>>>>> point-sta
>>>>>> and STAT-Analysis on those, but I was wondering if there was
another
>>>> way
>>>>>> around?
>>>>>> Thanks,
>>>>>> Jason
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>>
>>
>>
------------------------------------------------
More information about the Met_help
mailing list