about SKEB in WRFV3.9

Topics specifically related to any of the WRF runtime options, such as SST update, DFI, nudging, restarts, global runs, adaptive time step, ndown.exe, tc.exe, moving nest, etc.
Post Reply
summer2021
Posts: 4
Joined: Tue Mar 02, 2021 7:01 am

about SKEB in WRFV3.9

Post by summer2021 » Wed Mar 03, 2021 8:25 am

Hi,
I have two questions about SKEB in WRF.
How to download and test a version of the module with the latest changes dyn em/module stoch.F ? Is this web address http://www.cgd.ucar.edu/berner/module stoch.F valid mentioned in SKEB_in_WRF.pdf ?
The other question is that in WRFUsersGuid.pdf , Tuning strategies are discussed in Romine et al. 2014(http://journals.ametsoc.org/doi/citedby ... 14-00100.1) . IF I test namelist like the following, the SKEB parameter in namelist , gridpoint variance (SPPT) and length scale (SPPT) and decorrelation time (SPPT) were set at the same time.The gridpoint variance (SPPT) and length scale (SPPT) and decorrelation time (SPPT) couldn't change the field RT_TENDF_STOCH , RU_TENDF_STOCH, and RV_TENDF_STOCH.My question is that how to change the stochastic perturbations of SKEB by setting the gridpoint variance (SPPT) and length scale (SPPT) in namelist similar as the artical (Romine et al. 2014).
Thank you! Looking forward to your reply!

Ming

I used WRFV3.9. Here is a part of my namelist.input.
&stoch
skebs = 1
stoch_force_opt = 0,0,
skebs_vertstruc = 0,
stoch_vertstruc_opt = 0,1,
tot_backscat_psi = 1.0E-05,1.0E-05,
tot_backscat_t = 1.0E-06,1.0E-06,
ztau_psi = 10800.0,
ztau_t = 10800.0,
rexponent_psi = -1.83,
rexponent_t = -1.83,
zsigma2_eps = 0.0833,
zsigma2_eta = 0.0833,
kminforc = 1,
lminforc = 1,
kminforct = 1,
lminforct = 1,
kmaxforc = 1000000,
lmaxforc = 1000000,
kmaxforct = 1000000,
lmaxforct = 1000000,
nens = 100
iseed_skebs = 811,
sppt = 0
iseed_sppt = 25
gridpt_stddev_sppt = 0.5
stddev_cutoff_sppt = 2.0
lengthscale_sppt = 90000.0
timescale_sppt = 21600.0
sppt_vertstruc = 0
/

kwerner
Posts: 2287
Joined: Wed Feb 14, 2018 9:21 pm

Re: about SKEB in WRFV3.9

Post by kwerner » Mon Mar 08, 2021 5:57 pm

Hi Ming,
The .pdf document's web links are outdated. Take a look at the SKEBS section in our newest version of the Users Guide for updated links. As for your other question, I'm going to try to contact a colleague for assistance and will get back to you.
NCAR/MMM

romine
Posts: 1
Joined: Mon Mar 08, 2021 6:01 pm

Re: about SKEB in WRFV3.9

Post by romine » Wed Mar 10, 2021 3:16 am

Hi Ming,

I deferred your question to Kevin Lupo, and not surprisingly he came up with a better answer to your question than I would have. Here is what he suggested:

Thanks for sending this my way. While I'm more familiar with SPPT than SKEBS, there are a couple of things I can address in the user's second question. First, the RT_, RU_, and RV_TENDF_STOCH fields are for SKEBS, and shouldn't directly respond to the SPPT tuning parameters. I think the SKEBS tuning parameters would be the ztau_, rexponent_, tot_backscat_, and zsigma2_ parameters, though I have less familiarity with these (but the user does appear to have them configured as in the Romine et al. 2014 article).

Second, the user would want to output the unitless rstoch field to examine the effect of changing the gridpt_stddev_sppt, lengthscale_sppt, and timescale_sppt parameters. Importantly, they also need to set sppt = 1 to turn on SPPT. But, as in above, these have no control on SKEBS fields.

Hope this is helpful,
Kevin Lupo
klupo@albany.edu

summer2021
Posts: 4
Joined: Tue Mar 02, 2021 7:01 am

Re: about SKEB in WRFV3.9

Post by summer2021 » Thu Mar 11, 2021 3:23 am

kwerner wrote:
Mon Mar 08, 2021 5:57 pm
Hi Ming,
The .pdf document's web links are outdated. Take a look at the
Skebs Section in our newest version of the Users Guide for updated links. As for your other question, I'm going to try to contact a colleague for assistance and will get back to you.
Thanks very much. But the updated links in the newest version of the Users Guide still can't be opened .

summer2021
Posts: 4
Joined: Tue Mar 02, 2021 7:01 am

Re: about SKEB in WRFV3.9

Post by summer2021 » Thu Mar 11, 2021 3:29 am

romine wrote:
Wed Mar 10, 2021 3:16 am
Hi Ming,

I deferred your question to Kevin Lupo, and not surprisingly he came up with a better answer to your question than I would have. Here is what he suggested:

Thanks for sending this my way. While I'm more familiar with SPPT than SKEBS, there are a couple of things I can address in the user's second question. First, the RT_, RU_, and RV_TENDF_STOCH fields are for SKEBS, and shouldn't directly respond to the SPPT tuning parameters. I think the SKEBS tuning parameters would be the ztau_, rexponent_, tot_backscat_, and zsigma2_ parameters, though I have less familiarity with these (but the user does appear to have them configured as in the Romine et al. 2014 article).

Second, the user would want to output the unitless rstoch field to examine the effect of changing the gridpt_stddev_sppt, lengthscale_sppt, and timescale_sppt parameters. Importantly, they also need to set sppt = 1 to turn on SPPT. But, as in above, these have no control on SKEBS fields.

Hope this is helpful,
Kevin Lupo
klupo@albany.edu
Thank you very much ,@romine and Kevin Lupo.

kwerner
Posts: 2287
Joined: Wed Feb 14, 2018 9:21 pm

Re: about SKEB in WRFV3.9

Post by kwerner » Thu Mar 11, 2021 6:48 pm

Hi Ming,
I apologize about the link. I've corrected it above. Try this:
https://www2.mmm.ucar.edu/wrf/users/docs/user_guide_v4/v4.2/users_guide_chap5.html#stochastic
NCAR/MMM

Post Reply

Return to “Special Running Options”