[Met_help] [rt.rap.ucar.edu #66333] History for point-stat question

[John Halley Gotway via RT]

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

Dear Professor,
Nice hear from you ! Thank you very much for your help !

Now I have another problem about 'Point_stat' tool , I use the obs output from "ascii2nc" tool and fcst output from "pcp_combine" tool, the scripts can be run, but the problem is "Processing APCP_03(*,*) versus APCP/A03, for observation type ADPSFC, over region FULL, for interpolation method UW_MEAN(9), using 0 pairs." It is very strange there is no pairs matched. I am wandering why.  
I have posted the data I am using to the ftp site.The folder is  /met_help/qiu. Please help me.
Very thanks!
Crystal


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

Subject: point-stat question
From: John Halley Gotway
Time: Fri Apr 18 10:16:47 2014

Crystal,

Since this question about point_stat is a new question, I created a
separate help desk ticket for it.

Your question is basically - why am I getting 0 matched pairs from
Point-Stat?  That's a question that comes up a lot and there are
several possible reasons.  The first step is to turn up the verbosity
level (-v 3) to see more diagnostic output.  When I ran your data at
verbosity level 3, I saw the following:

DEBUG 2: Processing APCP_03(*,*) versus APCP/A03, for observation type
ADPSFC, over region FULL, for interpolation method UW_MEAN(9), using 0
pairs.
DEBUG 3: Number of matched pairs  = 0
DEBUG 3: Observations processed   = 1751
DEBUG 3: Rejected: GRIB code      = 0
DEBUG 3: Rejected: valid time     = 0
DEBUG 3: Rejected: bad obs value  = 0
DEBUG 3: Rejected: off the grid   = 0
DEBUG 3: Rejected: level mismatch = 1751
DEBUG 3: Rejected: quality marker = 0
DEBUG 3: Rejected: message type   = 0
DEBUG 3: Rejected: masking region = 0
DEBUG 3: Rejected: bad fcst value = 0
DEBUG 3: Rejected: duplicates     = 0

This output list counts of reasons for why observations were rejected.
The order of the reason counts is the order the filtering logic is
applied in Point-Stat.  In your case, all the observations were
rejected because of a "level mismatch".  For precip, a level mismatch
means that the observation accumulation interval doesn't match what
you requested in the config file.

Looking at your observations, I see that their accumulation interval
is 1 hour (3600 seconds).  In order to make sure there aren't any
other problems, I modified the config file to look for hourly
observations:
      {
        name       = "APCP";
        level      = [ "A01" ];
        cat_thresh = [ >=0.0 ];
      }

When I reran Point-Stat with this change, I got 1751 matched pairs:

DEBUG 2: Processing APCP_03(*,*) versus APCP/A01, for observation type
ADPSFC, over region FULL, for interpolation method UW_MEAN(9), using
1751 pairs.

So you just need to address the accumulation interval problem.  You'll
either need to sum up your hourly observations into 3-hourly bins, or
use these observations to verify hourly forecasts of precip.

Thanks,
John Halley Gotway
met_help at ucar.edu

------------------------------------------------
Subject: point-stat question
From: Crystal
Time: Sun Apr 20 20:11:03 2014

Dear Professor,My question is resolved according to your e-mail.Thank
you very much !crystal







At 2014-04-19 00:16:47,"John Halley Gotway via RT" <met_help at ucar.edu>
wrote:
>Crystal,
>
>Since this question about point_stat is a new question, I created a
separate help desk ticket for it.
>
>Your question is basically - why am I getting 0 matched pairs from
Point-Stat?  That's a question that comes up a lot and there are
several possible reasons.  The first step is to turn up the verbosity
level (-v 3) to see more diagnostic output.  When I ran your data at
verbosity level 3, I saw the following:
>
>DEBUG 2: Processing APCP_03(*,*) versus APCP/A03, for observation
type ADPSFC, over region FULL, for interpolation method UW_MEAN(9),
using 0 pairs.
>DEBUG 3: Number of matched pairs  = 0
>DEBUG 3: Observations processed   = 1751
>DEBUG 3: Rejected: GRIB code      = 0
>DEBUG 3: Rejected: valid time     = 0
>DEBUG 3: Rejected: bad obs value  = 0
>DEBUG 3: Rejected: off the grid   = 0
>DEBUG 3: Rejected: level mismatch = 1751
>DEBUG 3: Rejected: quality marker = 0
>DEBUG 3: Rejected: message type   = 0
>DEBUG 3: Rejected: masking region = 0
>DEBUG 3: Rejected: bad fcst value = 0
>DEBUG 3: Rejected: duplicates     = 0
>
>This output list counts of reasons for why observations were
rejected.  The order of the reason counts is the order the filtering
logic is applied in Point-Stat.  In your case, all the observations
were rejected because of a "level mismatch".  For precip, a level
mismatch means that the observation accumulation interval doesn't
match what you requested in the config file.
>
>Looking at your observations, I see that their accumulation interval
is 1 hour (3600 seconds).  In order to make sure there aren't any
other problems, I modified the config file to look for hourly
observations:
>      {
>        name       = "APCP";
>        level      = [ "A01" ];
>        cat_thresh = [ >=0.0 ];
>      }
>
>When I reran Point-Stat with this change, I got 1751 matched pairs:
>
>DEBUG 2: Processing APCP_03(*,*) versus APCP/A01, for observation
type ADPSFC, over region FULL, for interpolation method UW_MEAN(9),
using 1751 pairs.
>
>So you just need to address the accumulation interval problem.
You'll either need to sum up your hourly observations into 3-hourly
bins, or use these observations to verify hourly forecasts of precip.
>
>Thanks,
>John Halley Gotway
>met_help at ucar.edu
>

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