Removing topography

Any problems/questions that might not fall into the other topics in this category.

Removing topography

Postby erik » Wed Aug 20, 2008 8:54 am

Hi. I'm trying to remove topographic features, such as Greenland and Spitsbergen, in real data cases. So far I have done this "manually" on the geo_em NETCDF files directly, simply by setting the HGT_M values to zero for each grid point inside a region. I then proceed with running metgrid.exe and the model itself. The obvious question is: Is this too simple? If so, what should I do?

Thanks in advance for your help!
Erik Kolstad
Bjerknes Centre for Climate Research
Bergen, Norway
erik
 
Posts: 3
Joined: Wed Aug 20, 2008 8:46 am
Location: Bergen, Norway

Re: Removing topography

Postby PBLer » Wed Aug 20, 2008 12:05 pm

That *might* be all you need, but I usually modify the slope fields as well as the HGT_U and HGT_V fields.

HGT_M:description = "Topography height" ;
SLPX:description = "df/dx" ;
SLPY:description = "df/dy" ;
HGT_U:description = "Topography height"
HGT_V:description = "Topography height" ;
PBLer
 
Posts: 38
Joined: Wed May 07, 2008 5:11 pm

Re: Removing topography

Postby erik » Sat Aug 23, 2008 10:18 am

Thanks a lot, PBLer. I managed to find out about HGT_U and HGT_V myself and ran some tests. Of course, it does make a difference that the staggered grid topography is also modified. I'm also setting SLPX and SLPY to zero now; more tests will reveal the importance of this.

PBLer wrote:That *might* be all you need, but I usually modify the slope fields as well as the HGT_U and HGT_V fields.

HGT_M:description = "Topography height" ;
SLPX:description = "df/dx" ;
SLPY:description = "df/dy" ;
HGT_U:description = "Topography height"
HGT_V:description = "Topography height" ;
Erik Kolstad
Bjerknes Centre for Climate Research
Bergen, Norway
erik
 
Posts: 3
Joined: Wed Aug 20, 2008 8:46 am
Location: Bergen, Norway

Re: Removing topography

Postby gustafson » Tue Sep 02, 2008 1:36 pm

Erik,

Were you able to finally get this to work? I've been trying to setup an aquaplanet simulation with a global domain and have run into a similar learning curve that you are facing. I am trying to start with a global GFS analysis product and then flatten out the world, prescribe an SST, and then set the surface to water along the equator to mid-latitudes and ice polewards of 60 deg. Ignoring the water/ice surface issues, I have it working for the atmosphere part. My problem was stability issues when bringing the 3-D initial fields down to the surface. Along the Andes the model kept blowing up. I finally applied two passes of a 1-2-1 smoother to the dynamic fields and this seems to have fixed the stability problem as the model starts up--I was at least able to calculate out for a week or so and the fields look smooth.

My problem now is getting the prescribed water/ice correct. The model keeps resetting the XLAND and LANDMASK variables back to land. If anybody has ideas on how to force the water flag to water so that it stays, I'd love to hear some tips. My current iteration is to reset the wrfinput_d01 values for LU_INDEX, LANDMASK, SNOALB, LANDUSEF, SOILCBOT, SOILCTOP, TSLB, SMOIS, SH2O, SEAICE, IVGTYP, VEGFRA, SST, HGT, TSK, TMN, and XLAND to the values that the ocean typically has in a normal "terrain enabled" world. But, XLAND and LANDMASK get reset to 1 leading to issues with surface fluxes, etc.

-Bill
William I. Gustafson Jr.
Atmospheric Science and Global Change Division
Pacific Northwest National Laboratory
http://www.pnnl.gov/atmospheric/staff/staff_info.asp?staff_num=5716
gustafson
 
Posts: 14
Joined: Wed Apr 16, 2008 11:45 am

Re: Removing topography

Postby andyb » Tue Nov 25, 2008 5:43 pm

Hi,

I am new to wrf and having trouble with a few things. I have successfully run WRF, now i'm trying to run it without topography in some areas. I was wondering how to set the HGT_M, etc, to zero for a particular gridpoint. Whenever i open geo_em to modify it, i get several thousand lines of unrecognisable code.

Thank you for any help
Andy
andyb
 
Posts: 1
Joined: Mon Nov 24, 2008 12:51 pm

Re: Removing topography

Postby wrfhelp » Tue Nov 25, 2008 10:58 pm

You could start with the utility program read_wrf_nc.f. Please see this site for information:

http://www.mmm.ucar.edu/wrf/users/utilities/util.htm

wrfhelp
wrfhelp
 
Posts: 68
Joined: Tue Apr 15, 2008 1:33 pm

Re: Removing topography

Postby mjm » Wed Apr 21, 2010 7:03 pm

Those of you thinking about making changes to the input files fed into WRF should check out the NCO utility (http://nco.sourceforge.net/).
I have found this software to be extremely useful in manipulating the geo_em & met_em netcdf files produced by WPS.

mj
mjm
 
Posts: 20
Joined: Sat Oct 11, 2008 12:04 am

Re: Removing topography

Postby srs_atmo » Sat Mar 26, 2011 1:58 am

I was wondering if someone could tell me how to calculate SLPX and SLPY? I am reducing topography in certain areas, but not removing it all the way and I want to make sure I am able to calculate the slope fields correctly.
srs_atmo
 
Posts: 1
Joined: Sat Mar 26, 2011 1:56 am

Re: Removing topography

Postby tomk » Mon Jun 06, 2011 2:27 am

I've been playing around with this and had the same problem as Bill. I was able to convert land points to ocean points by setting the following fields in my wrfinput files:

XLAND=2
LANDMASK=0
LU_INDEX=16
IVGTYP=16
ISLTYP=14
HGT=0
VEGFRA=0
ALBBCK=0.08
SHDMAX=0
SHDMIN=0
SNOALB=0.08
TMN=SST

If I omit the LU_INDEX,IVGTYP,ISLTYP changes, wrf resets XLAND and LANDMASK back to their land values. This may be due to dyn_em/module_initialize_real.F, where it looks like wrf checks to make sure the land/sea points are consistent with the vegetation and soil points.

-Tom
tomk
 
Posts: 7
Joined: Mon Jun 06, 2011 2:17 am

Re: Removing topography

Postby whylyb » Sat Aug 27, 2011 6:19 pm

Hi Tom,

If you use land surface modules, you have to change LU_INDEX and LANDUSEF also, otherwise WRF will reset LANDMASK. I met this problem before and solved it by changing LU_INDEX AND LANDUSEF.

But I have a problem with MYJ boundary scheme, when I change LANDMASK, and run with MYJ scheme, then I got the wind/pressure/etc. all become "infinity". But if I run with MYNN scheme, there will be no problem. Anybody knows why? Thanks.

Yubin

tomk wrote:I've been playing around with this and had the same problem as Bill. I was able to convert land points to ocean points by setting the following fields in my wrfinput files:

XLAND=2
LANDMASK=0
LU_INDEX=16
IVGTYP=16
ISLTYP=14
HGT=0
VEGFRA=0
ALBBCK=0.08
SHDMAX=0
SHDMIN=0
SNOALB=0.08
TMN=SST

If I omit the LU_INDEX,IVGTYP,ISLTYP changes, wrf resets XLAND and LANDMASK back to their land values. This may be due to dyn_em/module_initialize_real.F, where it looks like wrf checks to make sure the land/sea points are consistent with the vegetation and soil points.

-Tom
whylyb
 
Posts: 2
Joined: Tue May 24, 2011 8:14 pm

Next

Return to Miscellaneous

Who is online

Users browsing this forum: No registered users and 4 guests

cron