[Met_help] [rt.rap.ucar.edu #40061] History for 2 quick ascii2nc format questions
RAL HelpDesk {for Paul Oldenburg}
met_help at ucar.edu
Wed Aug 18 14:38:41 MDT 2010
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
I am writing a program to convert madis observations to the ascii file
format required by the "ascii2nc" program. I have 2 brief questions:
1) How do I write an entries for 2m-temperature or 10m-wind surface
observations where the station pressure is unknown/un-reported. Looking at
the ascii2nc documentation it appears as though the "Elev" & "Hgt" columns
are not used (ie -9999). Furthermore, the "Lvl" column expects only pressure
or accumulation intervals, but not height. Without a way to include 2m or
10m above ground heights, is there any way to distinguish a difference in
these observations? Without an observation pressure can I set "Lvl" to -9999
without point_stat rejecting these?
2) If I do not have a WMO id for an observation, can I set this field to
-999 without the observation being rejected?
I would greatly appreciate advice on how to proceed with these two issues.
Thanks,
Scott
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #40061] 2 quick ascii2nc format questions
From: John Halley Gotway
Time: Fri Aug 13 10:31:25 2010
Scott,
Good questions.
I'll answer the second one first.
(2) You're welcome to put in any text string you'd like in the station
name column. Personally, I'd think a string like "NA" or "MISSING"
would make more sense. Point-Stat shouldn't reject
observations based on the station name unless you use the "mask_sid"
option in the Point-Stat config file to list out the names of the
stations you want. But as long as you're just looking over an
area, using a masking grid or polyline, it'll behave as you expect.
And now to the first question.
(1) I see that you want to verify 2-m tmp and 10-m winds. Point-Stat
treats these surface observations in a special way we've inherited
from the NCEP verification software upon which MET was
initially based. When verifying surface variables, the actual
pressure or level values of the observation are NOT used. Instead,
the matching is only done using message types. Specifically,
observations with message types of "ADPSFC" or "SFCSHP" are used to
verify surface variables. So in your reformatting step, be sure to
encode 2-m tmp and 10-m winds with the "ADPSFC" message type.
Just let us know if more questions arise in your use of MET.
Thanks,
John Halley Gotway
met_help at ucar.edu
RAL HelpDesk {for Scott Rabenhorst} wrote:
> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> Transaction: Ticket created by scott.rabenhorst at gmail.com
> Queue: met_help
> Subject: 2 quick ascii2nc format questions
> Owner: Nobody
> Requestors: scott.rabenhorst at gmail.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>
>
> Hello,
>
> I am writing a program to convert madis observations to the ascii
file
> format required by the "ascii2nc" program. I have 2 brief questions:
>
> 1) How do I write an entries for 2m-temperature or 10m-wind surface
> observations where the station pressure is unknown/un-reported.
Looking at
> the ascii2nc documentation it appears as though the "Elev" & "Hgt"
columns
> are not used (ie -9999). Furthermore, the "Lvl" column expects only
pressure
> or accumulation intervals, but not height. Without a way to include
2m or
> 10m above ground heights, is there any way to distinguish a
difference in
> these observations? Without an observation pressure can I set "Lvl"
to -9999
> without point_stat rejecting these?
>
> 2) If I do not have a WMO id for an observation, can I set this
field to
> -999 without the observation being rejected?
>
> I would greatly appreciate advice on how to proceed with these two
issues.
>
> Thanks,
> Scott
------------------------------------------------
Subject: 2 quick ascii2nc format questions
From: Scott Rabenhorst
Time: Fri Aug 13 12:12:20 2010
Hi John,
Thanks for your time and helpful information answering my questions. A
special thanks for such a prompt response and insight into how MET
processes
observations.
Take care,
Scott
On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
Gotway} <
met_help at ucar.edu> wrote:
> Scott,
>
> Good questions.
>
> I'll answer the second one first.
>
> (2) You're welcome to put in any text string you'd like in the
station name
> column. Personally, I'd think a string like "NA" or "MISSING" would
make
> more sense. Point-Stat shouldn't reject
> observations based on the station name unless you use the "mask_sid"
option
> in the Point-Stat config file to list out the names of the stations
you
> want. But as long as you're just looking over an
> area, using a masking grid or polyline, it'll behave as you expect.
>
> And now to the first question.
>
> (1) I see that you want to verify 2-m tmp and 10-m winds. Point-
Stat
> treats these surface observations in a special way we've inherited
from the
> NCEP verification software upon which MET was
> initially based. When verifying surface variables, the actual
pressure or
> level values of the observation are NOT used. Instead, the matching
is only
> done using message types. Specifically,
> observations with message types of "ADPSFC" or "SFCSHP" are used to
verify
> surface variables. So in your reformatting step, be sure to encode
2-m tmp
> and 10-m winds with the "ADPSFC" message type.
>
> Just let us know if more questions arise in your use of MET.
>
> Thanks,
> John Halley Gotway
> met_help at ucar.edu
>
> RAL HelpDesk {for Scott Rabenhorst} wrote:
> > Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> > Transaction: Ticket created by scott.rabenhorst at gmail.com
> > Queue: met_help
> > Subject: 2 quick ascii2nc format questions
> > Owner: Nobody
> > Requestors: scott.rabenhorst at gmail.com
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >
> >
> > Hello,
> >
> > I am writing a program to convert madis observations to the ascii
file
> > format required by the "ascii2nc" program. I have 2 brief
questions:
> >
> > 1) How do I write an entries for 2m-temperature or 10m-wind
surface
> > observations where the station pressure is unknown/un-reported.
Looking
> at
> > the ascii2nc documentation it appears as though the "Elev" & "Hgt"
> columns
> > are not used (ie -9999). Furthermore, the "Lvl" column expects
only
> pressure
> > or accumulation intervals, but not height. Without a way to
include 2m or
> > 10m above ground heights, is there any way to distinguish a
difference in
> > these observations? Without an observation pressure can I set
"Lvl" to
> -9999
> > without point_stat rejecting these?
> >
> > 2) If I do not have a WMO id for an observation, can I set this
field to
> > -999 without the observation being rejected?
> >
> > I would greatly appreciate advice on how to proceed with these two
> issues.
> >
> > Thanks,
> > Scott
>
>
------------------------------------------------
Subject: 2 quick ascii2nc format questions
From: John Halley Gotway
Time: Fri Aug 13 13:02:01 2010
Sure. Happy to help. I'll go ahead and resolve this ticket.
Thanks,
John
On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
> Hi John,
> Thanks for your time and helpful information answering my questions.
A
> special thanks for such a prompt response and insight into how MET
> processes
> observations.
> Take care,
> Scott
>
>
>
> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
> Gotway} <
> met_help at ucar.edu> wrote:
>
> > Scott,
> >
> > Good questions.
> >
> > I'll answer the second one first.
> >
> > (2) You're welcome to put in any text string you'd like in the
> station name
> > column. Personally, I'd think a string like "NA" or "MISSING"
would
> make
> > more sense. Point-Stat shouldn't reject
> > observations based on the station name unless you use the
"mask_sid"
> option
> > in the Point-Stat config file to list out the names of the
stations
> you
> > want. But as long as you're just looking over an
> > area, using a masking grid or polyline, it'll behave as you
expect.
> >
> > And now to the first question.
> >
> > (1) I see that you want to verify 2-m tmp and 10-m winds. Point-
> Stat
> > treats these surface observations in a special way we've inherited
> from the
> > NCEP verification software upon which MET was
> > initially based. When verifying surface variables, the actual
> pressure or
> > level values of the observation are NOT used. Instead, the
matching
> is only
> > done using message types. Specifically,
> > observations with message types of "ADPSFC" or "SFCSHP" are used
to
> verify
> > surface variables. So in your reformatting step, be sure to
encode
> 2-m tmp
> > and 10-m winds with the "ADPSFC" message type.
> >
> > Just let us know if more questions arise in your use of MET.
> >
> > Thanks,
> > John Halley Gotway
> > met_help at ucar.edu
> >
> > RAL HelpDesk {for Scott Rabenhorst} wrote:
> > > Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> > > Transaction: Ticket created by scott.rabenhorst at gmail.com
> > > Queue: met_help
> > > Subject: 2 quick ascii2nc format questions
> > > Owner: Nobody
> > > Requestors: scott.rabenhorst at gmail.com
> > > Status: new
> > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> > >
> > >
> > > Hello,
> > >
> > > I am writing a program to convert madis observations to the
ascii
> file
> > > format required by the "ascii2nc" program. I have 2 brief
> questions:
> > >
> > > 1) How do I write an entries for 2m-temperature or 10m-wind
> surface
> > > observations where the station pressure is unknown/un-reported.
> Looking
> > at
> > > the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
> > columns
> > > are not used (ie -9999). Furthermore, the "Lvl" column expects
> only
> > pressure
> > > or accumulation intervals, but not height. Without a way to
> include 2m or
> > > 10m above ground heights, is there any way to distinguish a
> difference in
> > > these observations? Without an observation pressure can I set
> "Lvl" to
> > -9999
> > > without point_stat rejecting these?
> > >
> > > 2) If I do not have a WMO id for an observation, can I set this
> field to
> > > -999 without the observation being rejected?
> > >
> > > I would greatly appreciate advice on how to proceed with these
two
> > issues.
> > >
> > > Thanks,
> > > Scott
> >
> >
------------------------------------------------
Subject: 2 quick ascii2nc format questions
From: Scott Rabenhorst
Time: Wed Aug 18 11:37:17 2010
Hello again,
Sorry, I have one more question regarding this topic. What
"fcst_field"
syntax should I use to select verification of only surface variables
only?
When I use "TMP/Z1", point_stat displays a message that only levels 2
and 10
m are supported. However, when I use TMP/Z2 or TMP/Z10 there are 0
matched
pairs...I don't think madis observations use these levels. When I try
TMP/R11 to select all/only grib11 fields, it I receive an error
because
point_stat tries to literally find level R11. I am not sure how this
syntax
is supposed to be used. Additionally, does GC/LNNN only correspond to
eta/model levels? Can Z, L, and P levels all have ranges or just
pressure? I
am a bit confused here. Any explaination would be much appreciated.
Thanks,
Scott
On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley Gotway}
<
met_help at ucar.edu> wrote:
> Sure. Happy to help. I'll go ahead and resolve this ticket.
>
> Thanks,
> John
>
> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
> > Hi John,
> > Thanks for your time and helpful information answering my
questions. A
> > special thanks for such a prompt response and insight into how MET
> > processes
> > observations.
> > Take care,
> > Scott
> >
> >
> >
> > On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
> > Gotway} <
> > met_help at ucar.edu> wrote:
> >
> > > Scott,
> > >
> > > Good questions.
> > >
> > > I'll answer the second one first.
> > >
> > > (2) You're welcome to put in any text string you'd like in the
> > station name
> > > column. Personally, I'd think a string like "NA" or "MISSING"
would
> > make
> > > more sense. Point-Stat shouldn't reject
> > > observations based on the station name unless you use the
"mask_sid"
> > option
> > > in the Point-Stat config file to list out the names of the
stations
> > you
> > > want. But as long as you're just looking over an
> > > area, using a masking grid or polyline, it'll behave as you
expect.
> > >
> > > And now to the first question.
> > >
> > > (1) I see that you want to verify 2-m tmp and 10-m winds.
Point-
> > Stat
> > > treats these surface observations in a special way we've
inherited
> > from the
> > > NCEP verification software upon which MET was
> > > initially based. When verifying surface variables, the actual
> > pressure or
> > > level values of the observation are NOT used. Instead, the
matching
> > is only
> > > done using message types. Specifically,
> > > observations with message types of "ADPSFC" or "SFCSHP" are used
to
> > verify
> > > surface variables. So in your reformatting step, be sure to
encode
> > 2-m tmp
> > > and 10-m winds with the "ADPSFC" message type.
> > >
> > > Just let us know if more questions arise in your use of MET.
> > >
> > > Thanks,
> > > John Halley Gotway
> > > met_help at ucar.edu
> > >
> > > RAL HelpDesk {for Scott Rabenhorst} wrote:
> > > > Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> > > > Transaction: Ticket created by scott.rabenhorst at gmail.com
> > > > Queue: met_help
> > > > Subject: 2 quick ascii2nc format questions
> > > > Owner: Nobody
> > > > Requestors: scott.rabenhorst at gmail.com
> > > > Status: new
> > > > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> > > >
> > > >
> > > > Hello,
> > > >
> > > > I am writing a program to convert madis observations to the
ascii
> > file
> > > > format required by the "ascii2nc" program. I have 2 brief
> > questions:
> > > >
> > > > 1) How do I write an entries for 2m-temperature or 10m-wind
> > surface
> > > > observations where the station pressure is unknown/un-
reported.
> > Looking
> > > at
> > > > the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
> > > columns
> > > > are not used (ie -9999). Furthermore, the "Lvl" column expects
> > only
> > > pressure
> > > > or accumulation intervals, but not height. Without a way to
> > include 2m or
> > > > 10m above ground heights, is there any way to distinguish a
> > difference in
> > > > these observations? Without an observation pressure can I set
> > "Lvl" to
> > > -9999
> > > > without point_stat rejecting these?
> > > >
> > > > 2) If I do not have a WMO id for an observation, can I set
this
> > field to
> > > > -999 without the observation being rejected?
> > > >
> > > > I would greatly appreciate advice on how to proceed with these
two
> > > issues.
> > > >
> > > > Thanks,
> > > > Scott
> > >
> > >
>
>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #40061] 2 quick ascii2nc format questions
From: Paul Oldenburg
Time: Wed Aug 18 13:09:06 2010
Scott,
In the madis obs text file you sent us, there are no surface
temperature observations (type ADPSFC with grib code 11).
There are upper air observations (type ADPUPA) that you could use to
verify forecast temperatures at pressure levels.
The notation for specifying such levels is: TMP/PNNN or TMP/PNNN-NNN.
We recommend that you avoid using the GRIB record
number notation. If you find some temperature surface observations,
use the notation TMP/ZNNN to specify the level for
those.
In general, point_stat is usually used with either an accumulation
interval (ANNN) or vertical/pressure levels (ZNNN,
PNNN or PNNN-NNN). If you have any other questions, please let us
know.
Thanks,
Paul
RAL HelpDesk {for Scott Rabenhorst} wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>
> Hello again,
>
> Sorry, I have one more question regarding this topic. What
"fcst_field"
> syntax should I use to select verification of only surface variables
only?
> When I use "TMP/Z1", point_stat displays a message that only levels
2 and 10
> m are supported. However, when I use TMP/Z2 or TMP/Z10 there are 0
matched
> pairs...I don't think madis observations use these levels. When I
try
> TMP/R11 to select all/only grib11 fields, it I receive an error
because
> point_stat tries to literally find level R11. I am not sure how this
syntax
> is supposed to be used. Additionally, does GC/LNNN only correspond
to
> eta/model levels? Can Z, L, and P levels all have ranges or just
pressure? I
> am a bit confused here. Any explaination would be much appreciated.
>
> Thanks,
> Scott
>
>
> On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley
Gotway} <
> met_help at ucar.edu> wrote:
>
>> Sure. Happy to help. I'll go ahead and resolve this ticket.
>>
>> Thanks,
>> John
>>
>> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
>>> Hi John,
>>> Thanks for your time and helpful information answering my
questions. A
>>> special thanks for such a prompt response and insight into how MET
>>> processes
>>> observations.
>>> Take care,
>>> Scott
>>>
>>>
>>>
>>> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
>>> Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Scott,
>>>>
>>>> Good questions.
>>>>
>>>> I'll answer the second one first.
>>>>
>>>> (2) You're welcome to put in any text string you'd like in the
>>> station name
>>>> column. Personally, I'd think a string like "NA" or "MISSING"
would
>>> make
>>>> more sense. Point-Stat shouldn't reject
>>>> observations based on the station name unless you use the
"mask_sid"
>>> option
>>>> in the Point-Stat config file to list out the names of the
stations
>>> you
>>>> want. But as long as you're just looking over an
>>>> area, using a masking grid or polyline, it'll behave as you
expect.
>>>>
>>>> And now to the first question.
>>>>
>>>> (1) I see that you want to verify 2-m tmp and 10-m winds. Point-
>>> Stat
>>>> treats these surface observations in a special way we've
inherited
>>> from the
>>>> NCEP verification software upon which MET was
>>>> initially based. When verifying surface variables, the actual
>>> pressure or
>>>> level values of the observation are NOT used. Instead, the
matching
>>> is only
>>>> done using message types. Specifically,
>>>> observations with message types of "ADPSFC" or "SFCSHP" are used
to
>>> verify
>>>> surface variables. So in your reformatting step, be sure to
encode
>>> 2-m tmp
>>>> and 10-m winds with the "ADPSFC" message type.
>>>>
>>>> Just let us know if more questions arise in your use of MET.
>>>>
>>>> Thanks,
>>>> John Halley Gotway
>>>> met_help at ucar.edu
>>>>
>>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>>>> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
>>>>> Transaction: Ticket created by scott.rabenhorst at gmail.com
>>>>> Queue: met_help
>>>>> Subject: 2 quick ascii2nc format questions
>>>>> Owner: Nobody
>>>>> Requestors: scott.rabenhorst at gmail.com
>>>>> Status: new
>>>>> Ticket <URL:
>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>>>
>>>>> Hello,
>>>>>
>>>>> I am writing a program to convert madis observations to the
ascii
>>> file
>>>>> format required by the "ascii2nc" program. I have 2 brief
>>> questions:
>>>>> 1) How do I write an entries for 2m-temperature or 10m-wind
>>> surface
>>>>> observations where the station pressure is unknown/un-reported.
>>> Looking
>>>> at
>>>>> the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
>>>> columns
>>>>> are not used (ie -9999). Furthermore, the "Lvl" column expects
>>> only
>>>> pressure
>>>>> or accumulation intervals, but not height. Without a way to
>>> include 2m or
>>>>> 10m above ground heights, is there any way to distinguish a
>>> difference in
>>>>> these observations? Without an observation pressure can I set
>>> "Lvl" to
>>>> -9999
>>>>> without point_stat rejecting these?
>>>>>
>>>>> 2) If I do not have a WMO id for an observation, can I set this
>>> field to
>>>>> -999 without the observation being rejected?
>>>>>
>>>>> I would greatly appreciate advice on how to proceed with these
two
>>>> issues.
>>>>> Thanks,
>>>>> Scott
>>>>
>>
>>
>>
------------------------------------------------
Subject: 2 quick ascii2nc format questions
From: Scott Rabenhorst
Time: Wed Aug 18 14:07:31 2010
Hi Paul,
Actually, I have several observation files for different verification
runs.
I believe the attached file has type ADPSFC, and this is the file I am
having problems with. When I try to select only surface obs using the
TMP/ZNNN notation, I have zero matched pairs. As you can see, there
should
be some surface comparisons produced. Any ideas which level(s) the
ascii2nc
program places these surface observations on?
Thanks,
Scott
On Wed, Aug 18, 2010 at 3:09 PM, RAL HelpDesk {for Paul Oldenburg} <
met_help at ucar.edu> wrote:
> Scott,
>
> In the madis obs text file you sent us, there are no surface
temperature
> observations (type ADPSFC with grib code 11).
> There are upper air observations (type ADPUPA) that you could use to
verify
> forecast temperatures at pressure levels.
> The notation for specifying such levels is: TMP/PNNN or TMP/PNNN-
NNN. We
> recommend that you avoid using the GRIB record
> number notation. If you find some temperature surface observations,
use
> the notation TMP/ZNNN to specify the level for
> those.
>
> In general, point_stat is usually used with either an accumulation
interval
> (ANNN) or vertical/pressure levels (ZNNN,
> PNNN or PNNN-NNN). If you have any other questions, please let us
know.
>
> Thanks,
>
> Paul
>
>
> RAL HelpDesk {for Scott Rabenhorst} wrote:
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >
> > Hello again,
> >
> > Sorry, I have one more question regarding this topic. What
"fcst_field"
> > syntax should I use to select verification of only surface
variables
> only?
> > When I use "TMP/Z1", point_stat displays a message that only
levels 2 and
> 10
> > m are supported. However, when I use TMP/Z2 or TMP/Z10 there are 0
> matched
> > pairs...I don't think madis observations use these levels. When I
try
> > TMP/R11 to select all/only grib11 fields, it I receive an error
because
> > point_stat tries to literally find level R11. I am not sure how
this
> syntax
> > is supposed to be used. Additionally, does GC/LNNN only correspond
to
> > eta/model levels? Can Z, L, and P levels all have ranges or just
> pressure? I
> > am a bit confused here. Any explaination would be much
appreciated.
> >
> > Thanks,
> > Scott
> >
> >
> > On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley
Gotway} <
> > met_help at ucar.edu> wrote:
> >
> >> Sure. Happy to help. I'll go ahead and resolve this ticket.
> >>
> >> Thanks,
> >> John
> >>
> >> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
> >>> Hi John,
> >>> Thanks for your time and helpful information answering my
questions. A
> >>> special thanks for such a prompt response and insight into how
MET
> >>> processes
> >>> observations.
> >>> Take care,
> >>> Scott
> >>>
> >>>
> >>>
> >>> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
> >>> Gotway} <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Scott,
> >>>>
> >>>> Good questions.
> >>>>
> >>>> I'll answer the second one first.
> >>>>
> >>>> (2) You're welcome to put in any text string you'd like in the
> >>> station name
> >>>> column. Personally, I'd think a string like "NA" or "MISSING"
would
> >>> make
> >>>> more sense. Point-Stat shouldn't reject
> >>>> observations based on the station name unless you use the
"mask_sid"
> >>> option
> >>>> in the Point-Stat config file to list out the names of the
stations
> >>> you
> >>>> want. But as long as you're just looking over an
> >>>> area, using a masking grid or polyline, it'll behave as you
expect.
> >>>>
> >>>> And now to the first question.
> >>>>
> >>>> (1) I see that you want to verify 2-m tmp and 10-m winds.
Point-
> >>> Stat
> >>>> treats these surface observations in a special way we've
inherited
> >>> from the
> >>>> NCEP verification software upon which MET was
> >>>> initially based. When verifying surface variables, the actual
> >>> pressure or
> >>>> level values of the observation are NOT used. Instead, the
matching
> >>> is only
> >>>> done using message types. Specifically,
> >>>> observations with message types of "ADPSFC" or "SFCSHP" are
used to
> >>> verify
> >>>> surface variables. So in your reformatting step, be sure to
encode
> >>> 2-m tmp
> >>>> and 10-m winds with the "ADPSFC" message type.
> >>>>
> >>>> Just let us know if more questions arise in your use of MET.
> >>>>
> >>>> Thanks,
> >>>> John Halley Gotway
> >>>> met_help at ucar.edu
> >>>>
> >>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
> >>>>> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> >>>>> Transaction: Ticket created by scott.rabenhorst at gmail.com
> >>>>> Queue: met_help
> >>>>> Subject: 2 quick ascii2nc format questions
> >>>>> Owner: Nobody
> >>>>> Requestors: scott.rabenhorst at gmail.com
> >>>>> Status: new
> >>>>> Ticket <URL:
> >>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >>>>>
> >>>>> Hello,
> >>>>>
> >>>>> I am writing a program to convert madis observations to the
ascii
> >>> file
> >>>>> format required by the "ascii2nc" program. I have 2 brief
> >>> questions:
> >>>>> 1) How do I write an entries for 2m-temperature or 10m-wind
> >>> surface
> >>>>> observations where the station pressure is unknown/un-
reported.
> >>> Looking
> >>>> at
> >>>>> the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
> >>>> columns
> >>>>> are not used (ie -9999). Furthermore, the "Lvl" column expects
> >>> only
> >>>> pressure
> >>>>> or accumulation intervals, but not height. Without a way to
> >>> include 2m or
> >>>>> 10m above ground heights, is there any way to distinguish a
> >>> difference in
> >>>>> these observations? Without an observation pressure can I set
> >>> "Lvl" to
> >>>> -9999
> >>>>> without point_stat rejecting these?
> >>>>>
> >>>>> 2) If I do not have a WMO id for an observation, can I set
this
> >>> field to
> >>>>> -999 without the observation being rejected?
> >>>>>
> >>>>> I would greatly appreciate advice on how to proceed with these
two
> >>>> issues.
> >>>>> Thanks,
> >>>>> Scott
> >>>>
> >>
> >>
> >>
>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #40061] 2 quick ascii2nc format questions
From: Paul Oldenburg
Time: Wed Aug 18 14:16:59 2010
Scott,
Both of the text files that you sent me have ADPSFC records, but
neither have such records for temperature, which is
grib code 11. To see this, try the following command:
cat madisObs_20060802_010000.txt | grep 'ADPSFC' | grep ' 11 '
This command will print out all lines that have both an observation
type of ADPSFC and a grib code of 11. Neither file
that you sent me has lines that satisfy this criteria. Does this make
sense? That is why you are not able to find any
matched pairs for surface temperature - because you have no
observations at the surface for temperature. Please let me
know if you have any questions.
Thanks,
Paul
RAL HelpDesk {for Scott Rabenhorst} wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>
> Hi Paul,
>
> Actually, I have several observation files for different
verification runs.
> I believe the attached file has type ADPSFC, and this is the file I
am
> having problems with. When I try to select only surface obs using
the
> TMP/ZNNN notation, I have zero matched pairs. As you can see, there
should
> be some surface comparisons produced. Any ideas which level(s) the
ascii2nc
> program places these surface observations on?
>
> Thanks,
> Scott
>
>
>
> On Wed, Aug 18, 2010 at 3:09 PM, RAL HelpDesk {for Paul Oldenburg} <
> met_help at ucar.edu> wrote:
>
>> Scott,
>>
>> In the madis obs text file you sent us, there are no surface
temperature
>> observations (type ADPSFC with grib code 11).
>> There are upper air observations (type ADPUPA) that you could use
to verify
>> forecast temperatures at pressure levels.
>> The notation for specifying such levels is: TMP/PNNN or TMP/PNNN-
NNN. We
>> recommend that you avoid using the GRIB record
>> number notation. If you find some temperature surface
observations, use
>> the notation TMP/ZNNN to specify the level for
>> those.
>>
>> In general, point_stat is usually used with either an accumulation
interval
>> (ANNN) or vertical/pressure levels (ZNNN,
>> PNNN or PNNN-NNN). If you have any other questions, please let us
know.
>>
>> Thanks,
>>
>> Paul
>>
>>
>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>
>>> Hello again,
>>>
>>> Sorry, I have one more question regarding this topic. What
"fcst_field"
>>> syntax should I use to select verification of only surface
variables
>> only?
>>> When I use "TMP/Z1", point_stat displays a message that only
levels 2 and
>> 10
>>> m are supported. However, when I use TMP/Z2 or TMP/Z10 there are 0
>> matched
>>> pairs...I don't think madis observations use these levels. When I
try
>>> TMP/R11 to select all/only grib11 fields, it I receive an error
because
>>> point_stat tries to literally find level R11. I am not sure how
this
>> syntax
>>> is supposed to be used. Additionally, does GC/LNNN only correspond
to
>>> eta/model levels? Can Z, L, and P levels all have ranges or just
>> pressure? I
>>> am a bit confused here. Any explaination would be much
appreciated.
>>>
>>> Thanks,
>>> Scott
>>>
>>>
>>> On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley
Gotway} <
>>> met_help at ucar.edu> wrote:
>>>
>>>> Sure. Happy to help. I'll go ahead and resolve this ticket.
>>>>
>>>> Thanks,
>>>> John
>>>>
>>>> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
>>>>> Hi John,
>>>>> Thanks for your time and helpful information answering my
questions. A
>>>>> special thanks for such a prompt response and insight into how
MET
>>>>> processes
>>>>> observations.
>>>>> Take care,
>>>>> Scott
>>>>>
>>>>>
>>>>>
>>>>> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John Halley
>>>>> Gotway} <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Scott,
>>>>>>
>>>>>> Good questions.
>>>>>>
>>>>>> I'll answer the second one first.
>>>>>>
>>>>>> (2) You're welcome to put in any text string you'd like in the
>>>>> station name
>>>>>> column. Personally, I'd think a string like "NA" or "MISSING"
would
>>>>> make
>>>>>> more sense. Point-Stat shouldn't reject
>>>>>> observations based on the station name unless you use the
"mask_sid"
>>>>> option
>>>>>> in the Point-Stat config file to list out the names of the
stations
>>>>> you
>>>>>> want. But as long as you're just looking over an
>>>>>> area, using a masking grid or polyline, it'll behave as you
expect.
>>>>>>
>>>>>> And now to the first question.
>>>>>>
>>>>>> (1) I see that you want to verify 2-m tmp and 10-m winds.
Point-
>>>>> Stat
>>>>>> treats these surface observations in a special way we've
inherited
>>>>> from the
>>>>>> NCEP verification software upon which MET was
>>>>>> initially based. When verifying surface variables, the actual
>>>>> pressure or
>>>>>> level values of the observation are NOT used. Instead, the
matching
>>>>> is only
>>>>>> done using message types. Specifically,
>>>>>> observations with message types of "ADPSFC" or "SFCSHP" are
used to
>>>>> verify
>>>>>> surface variables. So in your reformatting step, be sure to
encode
>>>>> 2-m tmp
>>>>>> and 10-m winds with the "ADPSFC" message type.
>>>>>>
>>>>>> Just let us know if more questions arise in your use of MET.
>>>>>>
>>>>>> Thanks,
>>>>>> John Halley Gotway
>>>>>> met_help at ucar.edu
>>>>>>
>>>>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>>>>>> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
>>>>>>> Transaction: Ticket created by scott.rabenhorst at gmail.com
>>>>>>> Queue: met_help
>>>>>>> Subject: 2 quick ascii2nc format questions
>>>>>>> Owner: Nobody
>>>>>>> Requestors: scott.rabenhorst at gmail.com
>>>>>>> Status: new
>>>>>>> Ticket <URL:
>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>>>>> Hello,
>>>>>>>
>>>>>>> I am writing a program to convert madis observations to the
ascii
>>>>> file
>>>>>>> format required by the "ascii2nc" program. I have 2 brief
>>>>> questions:
>>>>>>> 1) How do I write an entries for 2m-temperature or 10m-wind
>>>>> surface
>>>>>>> observations where the station pressure is unknown/un-
reported.
>>>>> Looking
>>>>>> at
>>>>>>> the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
>>>>>> columns
>>>>>>> are not used (ie -9999). Furthermore, the "Lvl" column expects
>>>>> only
>>>>>> pressure
>>>>>>> or accumulation intervals, but not height. Without a way to
>>>>> include 2m or
>>>>>>> 10m above ground heights, is there any way to distinguish a
>>>>> difference in
>>>>>>> these observations? Without an observation pressure can I set
>>>>> "Lvl" to
>>>>>> -9999
>>>>>>> without point_stat rejecting these?
>>>>>>>
>>>>>>> 2) If I do not have a WMO id for an observation, can I set
this
>>>>> field to
>>>>>>> -999 without the observation being rejected?
>>>>>>>
>>>>>>> I would greatly appreciate advice on how to proceed with these
two
>>>>>> issues.
>>>>>>> Thanks,
>>>>>>> Scott
>>>>
>>>>
>>
>>
>
------------------------------------------------
Subject: 2 quick ascii2nc format questions
From: Scott Rabenhorst
Time: Wed Aug 18 14:37:25 2010
Hi Paul,
I apologize for wasting your time. I mixed up the names of different
observation files. You were correct that the one I had been using did
not
have TMP data. I corrected the file mistake and grepped the new one to
be
sure. Sure enough, point_stat worked this time selecting TMP, DPT, and
RH at
2m, and winds at 10m. This was what I was originally expecting. Again,
I am
sorry for the confusion and thanks for your help. Things are working
just
fine now.
Thanks,
Scott
On Wed, Aug 18, 2010 at 4:17 PM, RAL HelpDesk {for Paul Oldenburg} <
met_help at ucar.edu> wrote:
> Scott,
>
> Both of the text files that you sent me have ADPSFC records, but
neither
> have such records for temperature, which is
> grib code 11. To see this, try the following command:
>
> cat madisObs_20060802_010000.txt | grep 'ADPSFC' | grep ' 11 '
>
> This command will print out all lines that have both an observation
type of
> ADPSFC and a grib code of 11. Neither file
> that you sent me has lines that satisfy this criteria. Does this
make
> sense? That is why you are not able to find any
> matched pairs for surface temperature - because you have no
observations at
> the surface for temperature. Please let me
> know if you have any questions.
>
> Thanks,
>
> Paul
>
>
> RAL HelpDesk {for Scott Rabenhorst} wrote:
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >
> > Hi Paul,
> >
> > Actually, I have several observation files for different
verification
> runs.
> > I believe the attached file has type ADPSFC, and this is the file
I am
> > having problems with. When I try to select only surface obs using
the
> > TMP/ZNNN notation, I have zero matched pairs. As you can see,
there
> should
> > be some surface comparisons produced. Any ideas which level(s) the
> ascii2nc
> > program places these surface observations on?
> >
> > Thanks,
> > Scott
> >
> >
> >
> > On Wed, Aug 18, 2010 at 3:09 PM, RAL HelpDesk {for Paul Oldenburg}
<
> > met_help at ucar.edu> wrote:
> >
> >> Scott,
> >>
> >> In the madis obs text file you sent us, there are no surface
temperature
> >> observations (type ADPSFC with grib code 11).
> >> There are upper air observations (type ADPUPA) that you could use
to
> verify
> >> forecast temperatures at pressure levels.
> >> The notation for specifying such levels is: TMP/PNNN or TMP/PNNN-
NNN.
> We
> >> recommend that you avoid using the GRIB record
> >> number notation. If you find some temperature surface
observations, use
> >> the notation TMP/ZNNN to specify the level for
> >> those.
> >>
> >> In general, point_stat is usually used with either an
accumulation
> interval
> >> (ANNN) or vertical/pressure levels (ZNNN,
> >> PNNN or PNNN-NNN). If you have any other questions, please let
us know.
> >>
> >> Thanks,
> >>
> >> Paul
> >>
> >>
> >> RAL HelpDesk {for Scott Rabenhorst} wrote:
> >>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >>>
> >>> Hello again,
> >>>
> >>> Sorry, I have one more question regarding this topic. What
"fcst_field"
> >>> syntax should I use to select verification of only surface
variables
> >> only?
> >>> When I use "TMP/Z1", point_stat displays a message that only
levels 2
> and
> >> 10
> >>> m are supported. However, when I use TMP/Z2 or TMP/Z10 there are
0
> >> matched
> >>> pairs...I don't think madis observations use these levels. When
I try
> >>> TMP/R11 to select all/only grib11 fields, it I receive an error
because
> >>> point_stat tries to literally find level R11. I am not sure how
this
> >> syntax
> >>> is supposed to be used. Additionally, does GC/LNNN only
correspond to
> >>> eta/model levels? Can Z, L, and P levels all have ranges or just
> >> pressure? I
> >>> am a bit confused here. Any explaination would be much
appreciated.
> >>>
> >>> Thanks,
> >>> Scott
> >>>
> >>>
> >>> On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley
Gotway}
> <
> >>> met_help at ucar.edu> wrote:
> >>>
> >>>> Sure. Happy to help. I'll go ahead and resolve this ticket.
> >>>>
> >>>> Thanks,
> >>>> John
> >>>>
> >>>> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
> >>>>> Hi John,
> >>>>> Thanks for your time and helpful information answering my
questions.
> A
> >>>>> special thanks for such a prompt response and insight into how
MET
> >>>>> processes
> >>>>> observations.
> >>>>> Take care,
> >>>>> Scott
> >>>>>
> >>>>>
> >>>>>
> >>>>> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John
Halley
> >>>>> Gotway} <
> >>>>> met_help at ucar.edu> wrote:
> >>>>>
> >>>>>> Scott,
> >>>>>>
> >>>>>> Good questions.
> >>>>>>
> >>>>>> I'll answer the second one first.
> >>>>>>
> >>>>>> (2) You're welcome to put in any text string you'd like in
the
> >>>>> station name
> >>>>>> column. Personally, I'd think a string like "NA" or
"MISSING" would
> >>>>> make
> >>>>>> more sense. Point-Stat shouldn't reject
> >>>>>> observations based on the station name unless you use the
"mask_sid"
> >>>>> option
> >>>>>> in the Point-Stat config file to list out the names of the
stations
> >>>>> you
> >>>>>> want. But as long as you're just looking over an
> >>>>>> area, using a masking grid or polyline, it'll behave as you
expect.
> >>>>>>
> >>>>>> And now to the first question.
> >>>>>>
> >>>>>> (1) I see that you want to verify 2-m tmp and 10-m winds.
Point-
> >>>>> Stat
> >>>>>> treats these surface observations in a special way we've
inherited
> >>>>> from the
> >>>>>> NCEP verification software upon which MET was
> >>>>>> initially based. When verifying surface variables, the
actual
> >>>>> pressure or
> >>>>>> level values of the observation are NOT used. Instead, the
matching
> >>>>> is only
> >>>>>> done using message types. Specifically,
> >>>>>> observations with message types of "ADPSFC" or "SFCSHP" are
used to
> >>>>> verify
> >>>>>> surface variables. So in your reformatting step, be sure to
encode
> >>>>> 2-m tmp
> >>>>>> and 10-m winds with the "ADPSFC" message type.
> >>>>>>
> >>>>>> Just let us know if more questions arise in your use of MET.
> >>>>>>
> >>>>>> Thanks,
> >>>>>> John Halley Gotway
> >>>>>> met_help at ucar.edu
> >>>>>>
> >>>>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
> >>>>>>> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
> >>>>>>> Transaction: Ticket created by scott.rabenhorst at gmail.com
> >>>>>>> Queue: met_help
> >>>>>>> Subject: 2 quick ascii2nc format questions
> >>>>>>> Owner: Nobody
> >>>>>>> Requestors: scott.rabenhorst at gmail.com
> >>>>>>> Status: new
> >>>>>>> Ticket <URL:
> >>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
> >>>>>>> Hello,
> >>>>>>>
> >>>>>>> I am writing a program to convert madis observations to the
ascii
> >>>>> file
> >>>>>>> format required by the "ascii2nc" program. I have 2 brief
> >>>>> questions:
> >>>>>>> 1) How do I write an entries for 2m-temperature or 10m-wind
> >>>>> surface
> >>>>>>> observations where the station pressure is unknown/un-
reported.
> >>>>> Looking
> >>>>>> at
> >>>>>>> the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
> >>>>>> columns
> >>>>>>> are not used (ie -9999). Furthermore, the "Lvl" column
expects
> >>>>> only
> >>>>>> pressure
> >>>>>>> or accumulation intervals, but not height. Without a way to
> >>>>> include 2m or
> >>>>>>> 10m above ground heights, is there any way to distinguish a
> >>>>> difference in
> >>>>>>> these observations? Without an observation pressure can I
set
> >>>>> "Lvl" to
> >>>>>> -9999
> >>>>>>> without point_stat rejecting these?
> >>>>>>>
> >>>>>>> 2) If I do not have a WMO id for an observation, can I set
this
> >>>>> field to
> >>>>>>> -999 without the observation being rejected?
> >>>>>>>
> >>>>>>> I would greatly appreciate advice on how to proceed with
these two
> >>>>>> issues.
> >>>>>>> Thanks,
> >>>>>>> Scott
> >>>>
> >>>>
> >>
> >>
> >
>
>
>
------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #40061] 2 quick ascii2nc format questions
From: Paul Oldenburg
Time: Wed Aug 18 14:38:18 2010
Scott,
No problem. I'm glad to hear that point_stat is working as expected.
Thanks,
Paul
RAL HelpDesk {for Scott Rabenhorst} wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>
> Hi Paul,
>
> I apologize for wasting your time. I mixed up the names of different
> observation files. You were correct that the one I had been using
did not
> have TMP data. I corrected the file mistake and grepped the new one
to be
> sure. Sure enough, point_stat worked this time selecting TMP, DPT,
and RH at
> 2m, and winds at 10m. This was what I was originally expecting.
Again, I am
> sorry for the confusion and thanks for your help. Things are working
just
> fine now.
>
> Thanks,
> Scott
>
>
>
> On Wed, Aug 18, 2010 at 4:17 PM, RAL HelpDesk {for Paul Oldenburg} <
> met_help at ucar.edu> wrote:
>
>> Scott,
>>
>> Both of the text files that you sent me have ADPSFC records, but
neither
>> have such records for temperature, which is
>> grib code 11. To see this, try the following command:
>>
>> cat madisObs_20060802_010000.txt | grep 'ADPSFC' | grep ' 11 '
>>
>> This command will print out all lines that have both an observation
type of
>> ADPSFC and a grib code of 11. Neither file
>> that you sent me has lines that satisfy this criteria. Does this
make
>> sense? That is why you are not able to find any
>> matched pairs for surface temperature - because you have no
observations at
>> the surface for temperature. Please let me
>> know if you have any questions.
>>
>> Thanks,
>>
>> Paul
>>
>>
>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>
>>> Hi Paul,
>>>
>>> Actually, I have several observation files for different
verification
>> runs.
>>> I believe the attached file has type ADPSFC, and this is the file
I am
>>> having problems with. When I try to select only surface obs using
the
>>> TMP/ZNNN notation, I have zero matched pairs. As you can see,
there
>> should
>>> be some surface comparisons produced. Any ideas which level(s) the
>> ascii2nc
>>> program places these surface observations on?
>>>
>>> Thanks,
>>> Scott
>>>
>>>
>>>
>>> On Wed, Aug 18, 2010 at 3:09 PM, RAL HelpDesk {for Paul Oldenburg}
<
>>> met_help at ucar.edu> wrote:
>>>
>>>> Scott,
>>>>
>>>> In the madis obs text file you sent us, there are no surface
temperature
>>>> observations (type ADPSFC with grib code 11).
>>>> There are upper air observations (type ADPUPA) that you could use
to
>> verify
>>>> forecast temperatures at pressure levels.
>>>> The notation for specifying such levels is: TMP/PNNN or TMP/PNNN-
NNN.
>> We
>>>> recommend that you avoid using the GRIB record
>>>> number notation. If you find some temperature surface
observations, use
>>>> the notation TMP/ZNNN to specify the level for
>>>> those.
>>>>
>>>> In general, point_stat is usually used with either an
accumulation
>> interval
>>>> (ANNN) or vertical/pressure levels (ZNNN,
>>>> PNNN or PNNN-NNN). If you have any other questions, please let
us know.
>>>>
>>>> Thanks,
>>>>
>>>> Paul
>>>>
>>>>
>>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>>>> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>>>
>>>>> Hello again,
>>>>>
>>>>> Sorry, I have one more question regarding this topic. What
"fcst_field"
>>>>> syntax should I use to select verification of only surface
variables
>>>> only?
>>>>> When I use "TMP/Z1", point_stat displays a message that only
levels 2
>> and
>>>> 10
>>>>> m are supported. However, when I use TMP/Z2 or TMP/Z10 there are
0
>>>> matched
>>>>> pairs...I don't think madis observations use these levels. When
I try
>>>>> TMP/R11 to select all/only grib11 fields, it I receive an error
because
>>>>> point_stat tries to literally find level R11. I am not sure how
this
>>>> syntax
>>>>> is supposed to be used. Additionally, does GC/LNNN only
correspond to
>>>>> eta/model levels? Can Z, L, and P levels all have ranges or just
>>>> pressure? I
>>>>> am a bit confused here. Any explaination would be much
appreciated.
>>>>>
>>>>> Thanks,
>>>>> Scott
>>>>>
>>>>>
>>>>> On Fri, Aug 13, 2010 at 3:02 PM, RAL HelpDesk {for John Halley
Gotway}
>> <
>>>>> met_help at ucar.edu> wrote:
>>>>>
>>>>>> Sure. Happy to help. I'll go ahead and resolve this ticket.
>>>>>>
>>>>>> Thanks,
>>>>>> John
>>>>>>
>>>>>> On Fri Aug 13 12:12:20 2010, scott.rabenhorst at gmail.com wrote:
>>>>>>> Hi John,
>>>>>>> Thanks for your time and helpful information answering my
questions.
>> A
>>>>>>> special thanks for such a prompt response and insight into how
MET
>>>>>>> processes
>>>>>>> observations.
>>>>>>> Take care,
>>>>>>> Scott
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Fri, Aug 13, 2010 at 12:31 PM, RAL HelpDesk {for John
Halley
>>>>>>> Gotway} <
>>>>>>> met_help at ucar.edu> wrote:
>>>>>>>
>>>>>>>> Scott,
>>>>>>>>
>>>>>>>> Good questions.
>>>>>>>>
>>>>>>>> I'll answer the second one first.
>>>>>>>>
>>>>>>>> (2) You're welcome to put in any text string you'd like in
the
>>>>>>> station name
>>>>>>>> column. Personally, I'd think a string like "NA" or
"MISSING" would
>>>>>>> make
>>>>>>>> more sense. Point-Stat shouldn't reject
>>>>>>>> observations based on the station name unless you use the
"mask_sid"
>>>>>>> option
>>>>>>>> in the Point-Stat config file to list out the names of the
stations
>>>>>>> you
>>>>>>>> want. But as long as you're just looking over an
>>>>>>>> area, using a masking grid or polyline, it'll behave as you
expect.
>>>>>>>>
>>>>>>>> And now to the first question.
>>>>>>>>
>>>>>>>> (1) I see that you want to verify 2-m tmp and 10-m winds.
Point-
>>>>>>> Stat
>>>>>>>> treats these surface observations in a special way we've
inherited
>>>>>>> from the
>>>>>>>> NCEP verification software upon which MET was
>>>>>>>> initially based. When verifying surface variables, the
actual
>>>>>>> pressure or
>>>>>>>> level values of the observation are NOT used. Instead, the
matching
>>>>>>> is only
>>>>>>>> done using message types. Specifically,
>>>>>>>> observations with message types of "ADPSFC" or "SFCSHP" are
used to
>>>>>>> verify
>>>>>>>> surface variables. So in your reformatting step, be sure to
encode
>>>>>>> 2-m tmp
>>>>>>>> and 10-m winds with the "ADPSFC" message type.
>>>>>>>>
>>>>>>>> Just let us know if more questions arise in your use of MET.
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> John Halley Gotway
>>>>>>>> met_help at ucar.edu
>>>>>>>>
>>>>>>>> RAL HelpDesk {for Scott Rabenhorst} wrote:
>>>>>>>>> Fri Aug 13 09:55:53 2010: Request 40061 was acted upon.
>>>>>>>>> Transaction: Ticket created by scott.rabenhorst at gmail.com
>>>>>>>>> Queue: met_help
>>>>>>>>> Subject: 2 quick ascii2nc format questions
>>>>>>>>> Owner: Nobody
>>>>>>>>> Requestors: scott.rabenhorst at gmail.com
>>>>>>>>> Status: new
>>>>>>>>> Ticket <URL:
>>>>>>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40061 >
>>>>>>>>> Hello,
>>>>>>>>>
>>>>>>>>> I am writing a program to convert madis observations to the
ascii
>>>>>>> file
>>>>>>>>> format required by the "ascii2nc" program. I have 2 brief
>>>>>>> questions:
>>>>>>>>> 1) How do I write an entries for 2m-temperature or 10m-wind
>>>>>>> surface
>>>>>>>>> observations where the station pressure is unknown/un-
reported.
>>>>>>> Looking
>>>>>>>> at
>>>>>>>>> the ascii2nc documentation it appears as though the "Elev" &
"Hgt"
>>>>>>>> columns
>>>>>>>>> are not used (ie -9999). Furthermore, the "Lvl" column
expects
>>>>>>> only
>>>>>>>> pressure
>>>>>>>>> or accumulation intervals, but not height. Without a way to
>>>>>>> include 2m or
>>>>>>>>> 10m above ground heights, is there any way to distinguish a
>>>>>>> difference in
>>>>>>>>> these observations? Without an observation pressure can I
set
>>>>>>> "Lvl" to
>>>>>>>> -9999
>>>>>>>>> without point_stat rejecting these?
>>>>>>>>>
>>>>>>>>> 2) If I do not have a WMO id for an observation, can I set
this
>>>>>>> field to
>>>>>>>>> -999 without the observation being rejected?
>>>>>>>>>
>>>>>>>>> I would greatly appreciate advice on how to proceed with
these two
>>>>>>>> issues.
>>>>>>>>> Thanks,
>>>>>>>>> Scott
>>>>>>
>>>>
>>
>>
------------------------------------------------
More information about the Met_help
mailing list