Page 1 of 1

WRF changing XLAND

PostPosted: Thu Jan 08, 2009 8:53 pm
by jzambon
Hey all,

I'm currently having a problem running WRF v2.1...

I setup an idealized simulation of a tropical cyclone (TC) making landfall. I created the wrfinput/bdy files using a hacked version of WRFSI. The hack works fine, it levels the land and then replaces it with water to eliminate problematic missing pressure/temperature values.

I then set the LANDMASK values to land (1) and XLAND values to land (1) using the read_wrf fortran program. Once the simulation starts, the LANDMASK remains the same through the entire run. However, the XLAND reverts back to entirely water at F000.

Anyone have any suggestions on getting the XLAND factor to remain the same in the simulation as in wrfinput?


Re: WRF changing XLAND

PostPosted: Thu May 12, 2016 2:40 pm
by jzambon
Its been over 7 years since I posted this. But I finally figured out the solution. In addition to setting LANDMASK, XLAND. You need to set other variables (e.g. HGT, LANDUSEF, etc) to water values. Otherwise something will overwrite it.

(just kidding, I figured this out a while ago and someone linked me this unresolved post today)

Re: WRF changing XLAND

PostPosted: Wed Jan 30, 2019 3:47 pm
by JamesH
Hi jzambon,

I'm running into the same problem (with WRFv3.9.1.1). I'm also trying to setup an idealized simulation of a TC making landfall. In my case, I'm using a matlab script to write in a meridional land boundary to wrfinput_d01 by changing LU_INDEX and XLAND. The roughness length (z0) values update just fine after the initial time, but it appears that the model still thinks that the land area is ocean even though XLAND=1, as the surface moisture fluxes into the atmosphere are way too high over land.

Do you remember which other variables need to be changed? I tried LANDUSEF but it didn't work, although I'm not sure if I'm using it correctly.


Re: WRF changing XLAND

PostPosted: Thu Feb 14, 2019 10:24 pm
by jzambon
There are a bunch that I had to change in addition to XLAND and LANDMASK. Also look into HGT and LANDUSEF.

The easiest thing to do might be to start from geogrid.exe, modify the geo_em output there and then work forward. This way everything that's written by metgrid and real will be congruent.

Good luck!

Re: WRF changing XLAND

PostPosted: Thu May 23, 2019 3:07 pm
by theanh0508

Could you please instruct me more detailed on how to set up landmass in the WRF idealized experiment?