Problems interpolating SST --sfc file

Questions and other topics related to UEMS 15.
Post Reply
norulz
Posts: 92
Joined: Thu Mar 12, 2015 1:43 pm

Problems interpolating SST --sfc file

Post by norulz » Sat Sep 10, 2016 8:17 am

Hello users,

I am trying to use a high resolution source for SST in the Mediterranean (1km). The Grib file includes only one variable for the temperature and no mask of its own. When using this file I am getting temperature artifacts on coastlines only. I suspect that the problem is a masking issue or an interpolation one.

I found that this is not an unknown issue - see following links - but no solution suggested has helped so far
http://forum.wrfforum.com/viewtopic.php?f=6&t=4512
https://meteo.unican.es/trac/wiki/WRFKn ... astallines
http://mailman.ucar.edu/pipermail/wrf-u ... 01471.html

My METGRID.TBL for SST is currently as follows:

Code: Select all

========================================
name=SST
        interp_option=sixteen_pt+four_pt+wt_average_16pt+search
        missing_value=-1.E30
        interp_mask=LANDSEA(1)
        masked=land
        fill_missing=0.
        flag_in_output=FLAG_SST
========================================
I have tried changing LANDSEA to LANDMASK and also tried other cocktails of interpolations with no affect at all

The only hint of a masking problem in prep_metgrid2.log is the following line:

Code: Select all

INFORM: Field LANDSEA.mask does not have a valid mask and will not be checked for missing values


The gfsp25 initialization files include a land/sea mask named "LAND" which I believe is used by metgrid as "LANDMASK"

Code: Select all

2016-09-10 07:35:55.478 --- Read in static field LANDMASK.
Can anyone please help on this?

David

windyweek
Posts: 28
Joined: Thu Aug 21, 2014 8:46 am

Re: Problems interpolating SST --sfc file

Post by windyweek » Sun Oct 09, 2016 5:12 pm

Hi David,

I was in the same boat and was able to find a solution.

The problem is a masking issue. I'll try to explain a bit down below.

When we try to input SST data, LANDMASK or LANDSEA won't ever be 100% accurate masks simply because they come from different sources. LANDMASK comes from the land use dataset (geogrid program) and LANDSEA comes out of the GFS dataset (if used). For the latter you can look at Vtable.GFS and you'll see it there.

What I did was to get the "original" mask (the one that comes with the SST dataset) and describe it in new Vtable all along with the actual data:

Code: Select all

GRIB1| Level| From |  To  | metgrid  | metgrid | metgrid                                 |GRIB2|GRIB2|GRIB2|GRIB2|
Param| Type |Level1|Level2| Name     | Units   | Description                             |Discp|Catgy|Param|Level|
-----+------+------+------+----------+---------+-----------------------------------------+-----------------------+
  11 |   1  |   0  |      | SST      | K       | Sea Surface Temperature		 		|  10 |  3  |  0  |   1 |
  81 |   1  |   0  |      | SST_MASK | proprtn | Land/Sea flag (1=land, 2=sea)           |  2  |  0  |  0  |   1 |
-----+------+------+------+----------+---------+-----------------------------------------+-----------------------+
Then in the METGRIB.TBL.ARW, you can use the new mask:

Code: Select all

========================================
name=SST
	missing_value=-1.E30
	flag_in_output=FLAG_SST
	interp_option=sixteen_pt+four_pt+wt_average_4pt+wt_average_16pt+search
	fill_missing=0.
	masked=land
	interp_mask=SST_MASK(2)
========================================
I am yet about to run that in production but in my testing environment SST looks good now in met_em* files.

Cheers,
Ivan

norulz
Posts: 92
Joined: Thu Mar 12, 2015 1:43 pm

Re: Problems interpolating SST --sfc file

Post by norulz » Sun Oct 09, 2016 6:32 pm

Hi Ivan,

Thanks for the input. I"ll give it a try in the coming days.

David

Post Reply