[Met_help] [rt.rap.ucar.edu #96175] History for Can CMORPH products of NETCDF format be applied in MET?
John Halley Gotway via RT
met_help at ucar.edu
Wed Aug 26 12:24:40 MDT 2020
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
To whom that could be concerned,
These days I have been trying FSS examination on MET-8.1.2 platform by employing CMORPH data downloaded from https://rda.ucar.edu/datasets/ds502.1/#!description, but little advances left behind me. The issues are as follows:
1. Can this kind of netcdf format data be used directly in MET? Is it necessary to re-edit the variable name? Such as, cmorph_precip in NETCDF structure should be renamed APCP or other else? I can not find any clues from the observation datasets description, https://dtcenter.org/community-code/model-evaluation-tools-met/input-data. It seems to me that the MET can not read CMORPH precipitation products directly from the website, but I do not know how to covert it to meet the requirements. Please let me know it.
2. Should the data be re-organized by a single time? The raw datasets I download are organized by 8 observation frequency in a single file.
3. When I execute the grid_stat command using the NETCDF format file as observation data directly, it always pops up the following text:
DEBUG 1: Default Config File: /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
DEBUG 1: User Config File: /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
ERROR :
ERROR : parse_vx_grid() -> The forecast and observation grids do not match:
ERROR : Projection: Lambert Conformal Nx: 711 Ny: 531 Lat_LL: 17.122 Lon_LL: -59.428 Lon_orient: -87.850 Alpha: 1269.303 Cone: 0.716 Bx:r 355.0641 By: 957.9183 !=
ERROR : Projection: Lat/Lon Nx: 1440 Ny: 480 lat_ll: -59.875 lon_ll: -0.125 delta_lat: 0.250 delta_lon: 0.250
ERROR : Specify regridding logic in the config file "regrid" section.
ERROR :
The Lambert Conformal data is from the UPP processing of WRFOUT, and the Lat/Lon projection data is CMORPH precipitation product. From the reminder, I do not know how to fix the regrid section in the configuration file whatever setting to_grid to "NONE", "OBS" or "FCST". How to remap the projection of two kind of source data? Should I have to cut the coverage as regional model domain first? I feel the projection is the key issue from the above text, but I do not know how to make them consistent. Please let me know if my guess is right.
Thank you.
I am looking forward to your news.
With my best regards,
Yu Xin
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Can CMORPH products of NETCDF format be applied in MET?
From: John Halley Gotway
Time: Tue Aug 11 16:05:55 2020
Yu Xin,
I see you have a question about using CMORPH data in MET. I pulled
this
sample file:
https://rda.ucar.edu/data/ds502.1/cmorph_v0.x/0.25deg_3hly/netcdf/2020/CMORPH_V0.x_0.25deg-
3HLY_20200809.nc
Running ncdump, I see that this file follows the NetCDF Climate
Forecast
(CF) convention. And MET should support NetCDF files following that
convention.
*ncdump -h CMORPH_V0.x_0.25deg-3HLY_20200807.nc
<http://CMORPH_V0.x_0.25deg-3HLY_20200807.nc> | grep Conventions*
* :Conventions = "CF-1.5" ;*
So you're running Grid-Stat to compare Lambert Conformal UPP output to
global Lat/Lon CMORPH data. This is actually very easy to fix. I'd
recommend modifying the Grid-Stat configuration file like this:
*regrid = {*
* to_grid = FCST;*
* method = BUDGET;*
* width = 2;*
* vld_thresh = 1.0;*
* shape = SQUARE;*
*}*
This will automatically regrid the CMORPH data to your model domain
using
budget interpolation, which is appropriate for precip data.
Please give that a shot and let me know how it goes.
Thanks,
John Halley Gotway
On Mon, Aug 10, 2020 at 1:36 AM Yu Xin via RT <met_help at ucar.edu>
wrote:
>
> Mon Aug 10 01:36:15 2020: Request 96175 was acted upon.
> Transaction: Ticket created by learnerxy at 163.com
> Queue: met_help
> Subject: Can CMORPH products of NETCDF format be applied in
MET?
> Owner: Nobody
> Requestors: learnerxy at 163.com
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=96175 >
>
>
> To whom that could be concerned,
> These days I have been trying FSS examination on MET-8.1.2 platform
by
> employing CMORPH data downloaded from
> https://rda.ucar.edu/datasets/ds502.1/#!description, but little
advances
> left behind me. The issues are as follows:
> 1. Can this kind of netcdf format data be used directly in MET? Is
it
> necessary to re-edit the variable name? Such as, cmorph_precip in
NETCDF
> structure should be renamed APCP or other else? I can not find any
clues
> from the observation datasets description,
> https://dtcenter.org/community-code/model-evaluation-tools-
met/input-data.
> It seems to me that the MET can not read CMORPH precipitation
products
> directly from the website, but I do not know how to covert it to
meet the
> requirements. Please let me know it.
> 2. Should the data be re-organized by a single time? The raw
datasets I
> download are organized by 8 observation frequency in a single file.
> 3. When I execute the grid_stat command using the NETCDF format file
as
> observation data directly, it always pops up the following text:
> DEBUG 1: Default Config File:
> /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
> DEBUG 1: User Config File:
> /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
> ERROR :
> ERROR : parse_vx_grid() -> The forecast and observation grids do
not
> match:
> ERROR : Projection: Lambert Conformal Nx: 711 Ny: 531 Lat_LL:
17.122
> Lon_LL: -59.428 Lon_orient: -87.850 Alpha: 1269.303 Cone: 0.716 Bx:r
> 355.0641 By: 957.9183 !=
> ERROR : Projection: Lat/Lon Nx: 1440 Ny: 480 lat_ll: -59.875
lon_ll:
> -0.125 delta_lat: 0.250 delta_lon: 0.250
> ERROR : Specify regridding logic in the config file "regrid"
section.
> ERROR :
> The Lambert Conformal data is from the UPP processing of WRFOUT, and
the
> Lat/Lon projection data is CMORPH precipitation product. From the
reminder,
> I do not know how to fix the regrid section in the configuration
file
> whatever setting to_grid to "NONE", "OBS" or "FCST". How to remap
the
> projection of two kind of source data? Should I have to cut the
coverage as
> regional model domain first? I feel the projection is the key issue
from
> the above text, but I do not know how to make them consistent.
Please let
> me know if my guess is right.
> Thank you.
> I am looking forward to your news.
>
>
> With my best regards,
>
>
> Yu Xin
>
------------------------------------------------
Subject: Can CMORPH products of NETCDF format be applied in MET?
From: Yu Xin
Time: Wed Aug 26 12:04:32 2020
Hello John,
Thank you for your answer. I've got it under your patient help.
The ticket can be closed now.
See you later.
Warm regards,
Yu
At 2020-08-12 06:05:56, "John Halley Gotway via RT"
<met_help at ucar.edu> wrote:
>Yu Xin,
>
>I see you have a question about using CMORPH data in MET. I pulled
this
>sample file:
>
>https://rda.ucar.edu/data/ds502.1/cmorph_v0.x/0.25deg_3hly/netcdf/2020/CMORPH_V0.x_0.25deg-
3HLY_20200809.nc
>
>Running ncdump, I see that this file follows the NetCDF Climate
Forecast
>(CF) convention. And MET should support NetCDF files following that
>convention.
>
>*ncdump -h CMORPH_V0.x_0.25deg-3HLY_20200807.nc
><http://CMORPH_V0.x_0.25deg-3HLY_20200807.nc> | grep Conventions*
>
>* :Conventions = "CF-1.5" ;*
>
>
>So you're running Grid-Stat to compare Lambert Conformal UPP output
to
>global Lat/Lon CMORPH data. This is actually very easy to fix. I'd
>recommend modifying the Grid-Stat configuration file like this:
>
>
>*regrid = {*
>
>* to_grid = FCST;*
>
>* method = BUDGET;*
>
>* width = 2;*
>
>* vld_thresh = 1.0;*
>
>* shape = SQUARE;*
>
>*}*
>
>
>This will automatically regrid the CMORPH data to your model domain
using
>budget interpolation, which is appropriate for precip data.
>
>
>Please give that a shot and let me know how it goes.
>
>
>Thanks,
>John Halley Gotway
>
>On Mon, Aug 10, 2020 at 1:36 AM Yu Xin via RT <met_help at ucar.edu>
wrote:
>
>>
>> Mon Aug 10 01:36:15 2020: Request 96175 was acted upon.
>> Transaction: Ticket created by learnerxy at 163.com
>> Queue: met_help
>> Subject: Can CMORPH products of NETCDF format be applied in
MET?
>> Owner: Nobody
>> Requestors: learnerxy at 163.com
>> Status: new
>> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=96175 >
>>
>>
>> To whom that could be concerned,
>> These days I have been trying FSS examination on MET-8.1.2 platform
by
>> employing CMORPH data downloaded from
>> https://rda.ucar.edu/datasets/ds502.1/#!description, but little
advances
>> left behind me. The issues are as follows:
>> 1. Can this kind of netcdf format data be used directly in MET? Is
it
>> necessary to re-edit the variable name? Such as, cmorph_precip in
NETCDF
>> structure should be renamed APCP or other else? I can not find any
clues
>> from the observation datasets description,
>> https://dtcenter.org/community-code/model-evaluation-tools-
met/input-data.
>> It seems to me that the MET can not read CMORPH precipitation
products
>> directly from the website, but I do not know how to covert it to
meet the
>> requirements. Please let me know it.
>> 2. Should the data be re-organized by a single time? The raw
datasets I
>> download are organized by 8 observation frequency in a single file.
>> 3. When I execute the grid_stat command using the NETCDF format
file as
>> observation data directly, it always pops up the following text:
>> DEBUG 1: Default Config File:
>> /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
>> DEBUG 1: User Config File:
>> /public/soft/met-8.1.1/share/met/config/GridStatConfig_default
>> ERROR :
>> ERROR : parse_vx_grid() -> The forecast and observation grids do
not
>> match:
>> ERROR : Projection: Lambert Conformal Nx: 711 Ny: 531 Lat_LL:
17.122
>> Lon_LL: -59.428 Lon_orient: -87.850 Alpha: 1269.303 Cone: 0.716
Bx:r
>> 355.0641 By: 957.9183 !=
>> ERROR : Projection: Lat/Lon Nx: 1440 Ny: 480 lat_ll: -59.875
lon_ll:
>> -0.125 delta_lat: 0.250 delta_lon: 0.250
>> ERROR : Specify regridding logic in the config file "regrid"
section.
>> ERROR :
>> The Lambert Conformal data is from the UPP processing of WRFOUT,
and the
>> Lat/Lon projection data is CMORPH precipitation product. From the
reminder,
>> I do not know how to fix the regrid section in the configuration
file
>> whatever setting to_grid to "NONE", "OBS" or "FCST". How to remap
the
>> projection of two kind of source data? Should I have to cut the
coverage as
>> regional model domain first? I feel the projection is the key issue
from
>> the above text, but I do not know how to make them consistent.
Please let
>> me know if my guess is right.
>> Thank you.
>> I am looking forward to your news.
>>
>>
>> With my best regards,
>>
>>
>> Yu Xin
>>
------------------------------------------------
More information about the Met_help
mailing list