WRF gives incorrect position of vortex

[Liliputian_123]

Dear WRF-experts,

I am running WRF version 4.0.3.
I am using the 1 degree x 1 degree FNL grib2 files (https://rda.ucar.edu/datasets/ds083.2/)

Attached are the outputs of my WRF run. Winds are at 925hPa level.

METEM plot of wind vectors




WRFOUT file
(rain and winds)


I am also attaching my namelists here.

PROBLEM:

[1] The location of the simulated vortex (along 8 degrees North) is always incorrect. I have tried different cumulus and microphysics combinations. The vortex is always located to the north and doesnt match the timing of the input files (sometimes a bit delayed).

[2] Another weird thing is that there are no wind vectors between 6-8N; 122-127E at 850hPa, considering the topography here is only 1200 meters (highest peak).

Any suggestions on how to run this correctly? Apologies, this is also the first time I encountered this type of mismatch.

 

[kwerner]

Hi,
I'm not able to say at this time what the exact cause of this is, but I do have a couple of suggestions that could potentially help in making your simulation closer to reality. First, the input data you are using is very coarse. I would suggest trying to use the 0.25 degree GFS data (https://rda.ucar.edu/datasets/ds084.1/). If you do that, your outer domain could have a much higher resolution, with an even higher inner domain resolution. You could use a 3:1 ratio, and use something like 9km and 3km. I also would suggest using much larger domains (e_we and e_sn larger than 100 x 100), but if you increase the resolution, you will need to increase the domain sizes to get the coverage you're interested in. Using anything <= 100 will not allow enough space for large systems to propagate through, often providing unrealistic results. Take a look at this document, which provides best practices for setting up your domain: https://www2.mmm.ucar.edu/wrf/users/namelist_best_prac_wps.html
See if doing that helps at all.

 

[Liliputian_123]

Dear kwerner,

Thank you for your suggestions.
It's just weird that I am not getting wind values over the southern Philippines.

In fact, I have to set the "extrapolation" to true in the namelist.ARWpost to get them:
These lines:

&interp
 interp_method = 1,
 interp_levels = 1000.,950.,925.,900.,850.,800.,750.,700.,650.,600.,550.,500.,450.,400.,350.,300.,250.,200.,150.,100.,
 extrapolate = .true.

 

[kwerner]

I'm not sure which storm this is, or how strong it was, but is it possible that the pressure in the center of the vortex is low enough that at 1200 m, 850 hPa is actually below the ground?

 

[Liliputian_123]

@kwerner

Maybe you are right. This is just a Low pressure area on 06 UTC January 16, 2017 not even a tropical cyclone. Below is the weather chart.



A westward propagating Low pressure area (cyclonic vortex). I plotted the input boundary condition and it got the correct location of the vortex before it hits land. But in most of my simulations, the timing is not correct.

 

[kwerner]

Hi,
The vortex-following capability was written to work well over water, but it has trouble when crossing land because of the need to update the domain's terrain/landuse. If your area of low pressure is crossing land, you may be seeing problems due to that. There is a work-around that you can read about in Chapter 5 of our Users' Guide (see section f - specifically the TERRAIN_AND_LANDUSE options):
https://www2.mmm.ucar.edu/wrf/users/docs/user_guide_v4/v4.1/users_guide_chap5.html

And see this page to obtain the data:
https://www2.mmm.ucar.edu/wrf/users/hurricanes/moving_nest.html

As for the timing being off, unfortunately, that's just the nature of modeling - sometimes it gets it right, and sometimes it doesn't. You could try different types of input data, but there isn't really a great solution for this.