[Met_help] [rt.rap.ucar.edu #78196] History for why the point_met 's output (point_stat****$fcsthour_WINDS**.stat is void

[John Halley Gotway via RT]

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Dear Mr. or Mrs.,
I had two experiments for the same domain, though their wgrib2 format and
grids maybe are not exactly the same.
However, it is still confusing to me why one of experiments will get a void
Met_Point's output like
point_stat_namrr_hi_impnest_F18_WINDS_180000L_20151031_000000V.stat, in
which only the head line was there. At the same time, the file like
point_stat_namrr_hi_impnest_F18_ADPUPA_180000L_20151031_000000V.stat was
normal with the verification results.
My first question is , what the WIND*stat files are for?    If the same
configure files were used while one experiment got the expectied results
and the other got an exmpty files,  what is the possible reason for a void
output if I know there were observations in that domain?
So,I hope I made my question clear.  Maybe it is not an easy explanation
from what I given. Then, that will be also great if you can first let me
know what the Point_stat*WINDS*stat files are?
Your help is appreciated.
Best Regards,
Ting


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: why the point_met 's output (point_stat****$fcsthour_WINDS**.stat is void
From: John Halley Gotway
Time: Fri Oct 07 16:50:34 2016

Hello Ting,

>From your message, I understand that you ran Point-Stat twice using
two
similar but slightly different datasets.  For one, you got the output
from
Point-Stat that you expected, but for the other, you got empty output
files
(i.e. just the header line).

The most likely cause for getting no output is that Point-Stat was not
able
to produce any matched forecast/observation pairs.  So I believe
you're
question is really, why are you getting 0 matched pairs from one of
your
Point-Stat runs?  There are several possible reasons... a mismatch
between
the forecast data and observation points in... space, time, vertical
level,
variable type, quality control level, and so on.

The easiest way to diagnose this behavior is to rerun the problematic
configuration of Point-Stat but use "-v" command line option to
increase
the verbosity level.  Setting the verbosity to 3 by using "-v 3" will
cause
the following type of information to be printed to the screen:

DEBUG 3: Number of matched pairs  = 364
DEBUG 3: Observations processed   = 89893
DEBUG 3: Rejected: SID exclusion  = 0
DEBUG 3: Rejected: GRIB code      = 77294
DEBUG 3: Rejected: valid time     = 0
DEBUG 3: Rejected: bad obs value  = 0
DEBUG 3: Rejected: off the grid   = 6
DEBUG 3: Rejected: level mismatch = 8077
DEBUG 3: Rejected: quality marker = 0
DEBUG 3: Rejected: message type   = 332
DEBUG 3: Rejected: masking region = 3820
DEBUG 3: Rejected: bad fcst value = 0
DEBUG 3: Rejected: duplicates     = 0

This info is printed for each verification task and list the number of
observations used for the each task followed by the number that were
not
used for various reasons.  Looking at these counts will give us a very
good
idea why you are getting 0 matched pairs.  For example, if all the
observations are discarded because of "valid time" then there is a
time
offset between your forecast and observation data.

Please try rerunning Point-Stat with "-v 3" and let me know what you
find
out.

Thanks,
John Halley Gotway




On Thu, Oct 6, 2016 at 6:38 PM, Ting Lei - NOAA Affiliate via RT <
met_help at ucar.edu> wrote:

>
> Thu Oct 06 18:38:09 2016: Request 78196 was acted upon.
> Transaction: Ticket created by ting.lei at noaa.gov
>        Queue: met_help
>      Subject: why the point_met 's output
(point_stat****$fcsthour_WINDS**.stat
> is void
>        Owner: Nobody
>   Requestors: ting.lei at noaa.gov
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78196 >
>
>
> Dear Mr. or Mrs.,
> I had two experiments for the same domain, though their wgrib2
format and
> grids maybe are not exactly the same.
> However, it is still confusing to me why one of experiments will get
a void
> Met_Point's output like
> point_stat_namrr_hi_impnest_F18_WINDS_180000L_20151031_000000V.stat,
in
> which only the head line was there. At the same time, the file like
> point_stat_namrr_hi_impnest_F18_ADPUPA_180000L_20151031_000000V.stat
was
> normal with the verification results.
> My first question is , what the WIND*stat files are for?    If the
same
> configure files were used while one experiment got the expectied
results
> and the other got an exmpty files,  what is the possible reason for
a void
> output if I know there were observations in that domain?
> So,I hope I made my question clear.  Maybe it is not an easy
explanation
> from what I given. Then, that will be also great if you can first
let me
> know what the Point_stat*WINDS*stat files are?
> Your help is appreciated.
> Best Regards,
> Ting
>
>

------------------------------------------------
Subject: why the point_met 's output (point_stat****$fcsthour_WINDS**.stat is void
From: Ting Lei - NOAA Affiliate
Time: Fri Oct 14 08:09:42 2016

Hi, John,
Sorry for the late reply.
I had just come back on this issue for last couple of days.
Thank you so much. Yes, using the "-v 3 " option as you suggested gave
more
info. Seems most of the 0 pairs compared with another experiment came
from
the "Rejected: level mismatch".
I will dig to see why this happened.
Your help is really appreciated.
Best Wishes,
Ting




On Fri, Oct 7, 2016 at 6:50 PM, John Halley Gotway via RT
<met_help at ucar.edu
> wrote:

> Hello Ting,
>
> From your message, I understand that you ran Point-Stat twice using
two
> similar but slightly different datasets.  For one, you got the
output from
> Point-Stat that you expected, but for the other, you got empty
output files
> (i.e. just the header line).
>
> The most likely cause for getting no output is that Point-Stat was
not able
> to produce any matched forecast/observation pairs.  So I believe
you're
> question is really, why are you getting 0 matched pairs from one of
your
> Point-Stat runs?  There are several possible reasons... a mismatch
between
> the forecast data and observation points in... space, time, vertical
level,
> variable type, quality control level, and so on.
>
> The easiest way to diagnose this behavior is to rerun the
problematic
> configuration of Point-Stat but use "-v" command line option to
increase
> the verbosity level.  Setting the verbosity to 3 by using "-v 3"
will cause
> the following type of information to be printed to the screen:
>
> DEBUG 3: Number of matched pairs  = 364
> DEBUG 3: Observations processed   = 89893
> DEBUG 3: Rejected: SID exclusion  = 0
> DEBUG 3: Rejected: GRIB code      = 77294
> DEBUG 3: Rejected: valid time     = 0
> DEBUG 3: Rejected: bad obs value  = 0
> DEBUG 3: Rejected: off the grid   = 6
> DEBUG 3: Rejected: level mismatch = 8077
> DEBUG 3: Rejected: quality marker = 0
> DEBUG 3: Rejected: message type   = 332
> DEBUG 3: Rejected: masking region = 3820
> DEBUG 3: Rejected: bad fcst value = 0
> DEBUG 3: Rejected: duplicates     = 0
>
> This info is printed for each verification task and list the number
of
> observations used for the each task followed by the number that were
not
> used for various reasons.  Looking at these counts will give us a
very good
> idea why you are getting 0 matched pairs.  For example, if all the
> observations are discarded because of "valid time" then there is a
time
> offset between your forecast and observation data.
>
> Please try rerunning Point-Stat with "-v 3" and let me know what you
find
> out.
>
> Thanks,
> John Halley Gotway
>
>
>
>
> On Thu, Oct 6, 2016 at 6:38 PM, Ting Lei - NOAA Affiliate via RT <
> met_help at ucar.edu> wrote:
>
> >
> > Thu Oct 06 18:38:09 2016: Request 78196 was acted upon.
> > Transaction: Ticket created by ting.lei at noaa.gov
> >        Queue: met_help
> >      Subject: why the point_met 's output
(point_stat****$fcsthour_
> WINDS**.stat
> > is void
> >        Owner: Nobody
> >   Requestors: ting.lei at noaa.gov
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78196 >
> >
> >
> > Dear Mr. or Mrs.,
> > I had two experiments for the same domain, though their wgrib2
format and
> > grids maybe are not exactly the same.
> > However, it is still confusing to me why one of experiments will
get a
> void
> > Met_Point's output like
> >
point_stat_namrr_hi_impnest_F18_WINDS_180000L_20151031_000000V.stat,
in
> > which only the head line was there. At the same time, the file
like
> >
point_stat_namrr_hi_impnest_F18_ADPUPA_180000L_20151031_000000V.stat
was
> > normal with the verification results.
> > My first question is , what the WIND*stat files are for?    If the
same
> > configure files were used while one experiment got the expectied
results
> > and the other got an exmpty files,  what is the possible reason
for a
> void
> > output if I know there were observations in that domain?
> > So,I hope I made my question clear.  Maybe it is not an easy
explanation
> > from what I given. Then, that will be also great if you can first
let me
> > know what the Point_stat*WINDS*stat files are?
> > Your help is appreciated.
> > Best Regards,
> > Ting
> >
> >
>
>

------------------------------------------------