[Met_help] [rt.rap.ucar.edu #78757] History for pcp_combine -subtract question

[John Halley Gotway via RT]

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

Hi met_help,

I'd like to verify different thresholds of forecast temperature anomalies.
I've used regrid_data_plane to put my forecast, observation, and
climatology grids onto one common grid.

I am able to successfully create the observed temperature anomaly grid by
using pcp_combine to subtract the climatology grid from the observation
grid.  However, I receive an error when trying to subtract the climatology
grid from the forecast grid because the initialization times are not the
same:

ERROR  :
ERROR  : do_sub_command() -> init_time1 (20160204_000000) must be equal to
init_time2 (20160214_000000) for subtraction.
ERROR  :

I see in the User's Guide that there is a way to ignore the same initial
time requirement when using the sum command.  Is there a way to ignore the
requirement when using the subtract command?  I'm currently using V5.1.

Thanks,
Dan


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

Subject: pcp_combine -subtract question
From: John Halley Gotway
Time: Wed Nov 23 13:27:50 2016

Dan,

Well, we couldn't think of a reasonable use case for someone wanting
to do
a subtraction of data from two different initialization times, but
this
sounds like a pretty reasonable one to me.

I'd suggest that we change that logic from an ERROR to a WARNING
message.
But the question is what timing info should be written to the output
file.
In your case, I suspect you'd want the timing info from the FIRST
file, not
the second.

This would require a small code change and recompilation.  If that's
problematic, we could edit the init time of the climatology data to
get
past this error.  I'd just use the "ncatted" command to do that.

If you'd like me to take a closer look, please send me some sample
data and
I can tell you what exact commands are required.

And I'll make a note that we should change that error to a warning in
the
next release.

Thanks,
John Halley Gotway

On Wed, Nov 23, 2016 at 12:46 PM, Daniel Halperin - NOAA Affiliate via
RT <
met_help at ucar.edu> wrote:

>
> Wed Nov 23 12:46:57 2016: Request 78757 was acted upon.
> Transaction: Ticket created by daniel.halperin at noaa.gov
>        Queue: met_help
>      Subject: pcp_combine -subtract question
>        Owner: Nobody
>   Requestors: daniel.halperin at noaa.gov
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78757 >
>
>
> Hi met_help,
>
> I'd like to verify different thresholds of forecast temperature
anomalies.
> I've used regrid_data_plane to put my forecast, observation, and
> climatology grids onto one common grid.
>
> I am able to successfully create the observed temperature anomaly
grid by
> using pcp_combine to subtract the climatology grid from the
observation
> grid.  However, I receive an error when trying to subtract the
climatology
> grid from the forecast grid because the initialization times are not
the
> same:
>
> ERROR  :
> ERROR  : do_sub_command() -> init_time1 (20160204_000000) must be
equal to
> init_time2 (20160214_000000) for subtraction.
> ERROR  :
>
> I see in the User's Guide that there is a way to ignore the same
initial
> time requirement when using the sum command.  Is there a way to
ignore the
> requirement when using the subtract command?  I'm currently using
V5.1.
>
> Thanks,
> Dan
>
>

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