[Met_help] [rt.rap.ucar.edu #92709] History for A question about writing NETCDF file comptable with MET

Tue Oct 13 10:58:36 MDT 2020

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

Dear sir or madam,

Hello. I have a question about MET.

I usually put my forecast data into NETCDF file to make it compactable with MET. I write the NETCDF file according to the output NETCDF file of PCP_combine.

However, I found that output of PCP_combine doesn't set the forecast lead correct. When passing these data to verification tools, such as, grid_stat, the output has wrong FCST_LEAD which is actually the accum_time. I wonder if it can be correted by adding an attribute setting forecast lead in input NETCDF file?

It is not a big problem, but correcting it will ease my programing when I am dealing with a large quantity of MET outputs. So would you help me with it please?

Thanks a lot!

Jiangshan Zhu

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

Subject: A question about writing NETCDF file comptable with MET
From: David Fillmore
Time: Fri Oct 18 12:11:55 2019

Hi Jiangshan -
Thank you for reporting this potential bug in PCP_combine.
We will have a look and let you know when a fix is available.
thanks,
David

On Fri, Oct 18, 2019 at 9:33 AM zhujiangshan via RT
<met_help at ucar.edu>
wrote:

>
> Fri Oct 18 09:33:47 2019: Request 92709 was acted upon.
> Transaction: Ticket created by zhujiangshan at mail.iap.ac.cn
>        Queue: met_help
>      Subject: A question about writing NETCDF file comptable with
MET
>        Owner: Nobody
>   Requestors: zhujiangshan at mail.iap.ac.cn
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=92709 >
>
>
> Dear sir or madam,
>
> Hello. I have a question about MET.
>
> I usually put my forecast data into NETCDF file to make it
compactable
> with MET. I write the NETCDF file according to the output NETCDF
file of
> PCP_combine.
>
> However, I found that output of PCP_combine doesn't set the forecast
lead
> correct. When passing these data to verification tools, such as,
grid_stat,
> the output has wrong FCST_LEAD which is actually the accum_time. I
wonder
> if it can be correted by adding an attribute setting forecast lead
in input
> NETCDF file?
>
> It is not a big problem, but correcting it will ease my programing
when I
> am dealing with a large quantity of MET outputs. So would you help
me with
> it please?
>
>
>
>
> Thanks a lot!
>
>
>
>
> Jiangshan Zhu
>

------------------------------------------------
Subject: A question about writing NETCDF file comptable with MET
From: David Fillmore
Time: Mon Dec 02 16:06:12 2019

Bug fix issue created:
https://github.com/NCAR/MET/issues/1222

------------------------------------------------
Subject: A question about writing NETCDF file comptable with MET
From: John Halley Gotway
Time: Mon Jul 27 20:03:57 2020

Hello Jiangshan,

You wrote this question about MET's pcp_combine tool to met-help way
back in October 2019. You indicated that there may be a problem with
pcp_combine's handling of forecast lead and accumulation times.

As a result, we created a GitHub issue to investigate:
https://github.com/NCAR/MET/issues/1222

As we are finishing up development for the next release of MET,
version 9.1, we'd like to resolve this issue.

When I look at the output from pcp_combine, it is not obvious to me
what the problem is.  For example, when I run pcp_combine to subtract
an 18-hour precip forecast minus a 15-hour precip forecast, I see the
following in the NetCDF output file.

	float APCP_03(lat, lon) ;
		APCP_03:name = "APCP_03" ;
		APCP_03:long_name = "Total precipitation" ;
		APCP_03:level = "A3" ;
		APCP_03:units = "kg/m^2" ;
		APCP_03:_FillValue = -9999.f ;
		APCP_03:init_time = "20120409_000000" ;
		APCP_03:init_time_ut = "1333929600" ;
		APCP_03:valid_time = "20120409_180000" ;
		APCP_03:valid_time_ut = "1333994400" ;
		APCP_03:accum_time = "030000" ;
		APCP_03:accum_time_sec = 10800 ;

This is a 3-hour accumulation (18-hour minus 15-hour), as indicated by
the accum_time variable attribute. The model initialization time is
20120409_000000 and the valid time is 20120409_180000. The forecast
lead time of 18 hours is derived by subtracting the valid time minus
the initialization time. I don't see any issues here.

If pcp_combine is not behaving the way you expect, can you please send
us an example, including input data and a description of the expected
behavior, to demonstrate the situation?

And my apologies for the very long resolution on this issue. But if
there is a problem, I'd like to fix it before our next release.

Thanks,
John Halley Gotway

On Fri Oct 18 09:33:47 2019, zhujiangshan at mail.iap.ac.cn wrote:
> Dear sir or madam,
>
> Hello. I have a question about MET.
>
> I usually put my forecast data into NETCDF file to make it
compactable
> with MET. I write the NETCDF file according to the output NETCDF
file
> of PCP_combine.
>
> However, I found that output of PCP_combine doesn't set the forecast
> lead correct. When passing these data to verification tools, such
as,
> grid_stat, the output has wrong FCST_LEAD which is actually the
> accum_time. I wonder if it can be correted by adding an attribute
> setting forecast lead in input NETCDF file?
>
> It is not a big problem, but correcting it will ease my programing
> when I am dealing with a large quantity of MET outputs. So would you
> help me with it please?
>
>
>
>
> Thanks a lot!
>
>
>
>
> Jiangshan Zhu

------------------------------------------------
Subject: A question about writing NETCDF file comptable with MET
From: David Fillmore
Time: Tue Oct 13 10:58:35 2020

Closed.

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