Did nudging support go away?

Forum dedicated to older versions of EMS package (WRFEMS v3.2, v3.1 or older). Support is user-to-user based, so please help others if you can.
Post Reply
pattim
Posts: 187
Joined: Sun Jun 24, 2012 8:42 pm
Location: Los Angeles, CA, USA

Did nudging support go away?

Post by pattim » Sun Jul 08, 2012 4:24 pm

I'm running the 3.2 beta and noticed that the run_nudging.conf file doesn't exist on my system, and that ems_prep and ems_run won't accept the "--nudge" flag. Did support for nudging go away for the beta? Is there a way around this? Thank You.

meteoadriatic
Posts: 1600
Joined: Wed Aug 19, 2009 10:05 am

Re: Did nudging support go away?

Post by meteoadriatic » Sun Jul 08, 2012 4:29 pm

I'm not aware that this has ever been in WRFEMS?

pattim
Posts: 187
Joined: Sun Jun 24, 2012 8:42 pm
Location: Los Angeles, CA, USA

Re: Did nudging support go away?

Post by pattim » Sun Jul 08, 2012 5:31 pm

meteoadriatic wrote:I'm not aware that this has ever been in WRFEMS?
Hi Meteo:
It's in the users manual chapter 8 (8.6.12) and Appendix F: "3D Analysis Nudging in the WRF EMS"
Patti

j1d1w1
Posts: 34
Joined: Wed Mar 13, 2013 2:02 am

Re: Did nudging support go away?

Post by j1d1w1 » Sun Mar 17, 2013 10:09 pm

Is there anything new on this subject? I, too, had the impression from Ch7 of http://strc.comet.ucar.edu/software/newrems/ that ems_prep takes an option --nudge, but find that in practice it does not (and that there is no run_nudging.conf file in ../conf/ems_run). Effectively, what I'm interested in doing is downscaling from an NCEP reanalysis to obtain something like an analysis on a much finer scale. The minimum simulation length is 6 hrs, and it seemed to me that in this case nudging towards the next analysis (on the outer domain, only) is a good idea if it can be done. Thanks, JDW

Antonix
Posts: 260
Joined: Fri Oct 16, 2009 8:53 am

Re: Did nudging support go away?

Post by Antonix » Mon Mar 18, 2013 6:31 pm

for nudging (grid nudging) add (and edit) this line in /wrfems/data/tables/wrf/namelist.arw

&fdda
grid_fdda = 1, 1, 1,
gfdda_inname = "wrffdda_d<domain>",
gfdda_end_h = 24, 24, 24,
gfdda_interval_m = 360, 360, 360,
fgdt = 0, 0, 0,
if_no_pbl_nudging_uv = 0, 0, 0,
if_no_pbl_nudging_t = 1, 1, 1,
if_no_pbl_nudging_q = 1, 1, 1,
if_zfac_uv = 1, 1, 1,
k_zfac_uv = 10, 10, 10,
if_zfac_t = 0, 0, 0,
k_zfac_t = 10, 10, 10,
if_zfac_q = 0, 0, 0,
k_zfac_q = 10, 10, 10,
guv = 0.0003, 0.0003, 0.0003,
gt = 0.0003, 0.0003, 0.0003,
gq = 0.0003, 0.0003, 0.0003,
if_ramping = 1,
dtramp_min = 60.0,
io_form_gfdda = 2,
/

j1d1w1
Posts: 34
Joined: Wed Mar 13, 2013 2:02 am

Re: Did nudging support go away?

Post by j1d1w1 » Tue Mar 19, 2013 2:12 pm

Thanks Antonix. Just a clarification: the user guide http://strc.comet.ucar.edu/software/newrems/, particularly Ch8 and Appendix F, uses the terminology "Analysis nudging''. Does your term "grid nudging" mean the same thing? I'm waiting for a run without nudging to end, then I'll be ready to give this a go. John

Antonix
Posts: 260
Joined: Fri Oct 16, 2009 8:53 am

Re: Did nudging support go away?

Post by Antonix » Tue Mar 19, 2013 6:08 pm

yes is the same thing.
the technique FDDA comprises three techniques.
grid nudging.
spectral nudging.
Nuding observation.

with the code that you wrote you can use the grid nudging technique.
but also the spectral technique is easy. just use the right code.
is very similar to this. search on the forum ("official") and find it!

meteoadriatic
Posts: 1600
Joined: Wed Aug 19, 2009 10:05 am

Re: Did nudging support go away?

Post by meteoadriatic » Tue Mar 19, 2013 7:03 pm

Can anybody explain in simple terms, what exactly can we get by this nudging thing? Thx :)

j1d1w1
Posts: 34
Joined: Wed Mar 13, 2013 2:02 am

Re: Did nudging support go away?

Post by j1d1w1 » Wed Mar 20, 2013 2:38 pm

In very simple terms, one is reducing the freedom of the forecast by constraining it, to some degree, to reproduce a reference state. In principle that reference state would be an analysis (or reanalysis) or even a sequence of analyses. I guess any or all of the model's dependent variables could be "nudged" to the reference state. The crudest way to do this is "relaxation," whereby every time a variable U is advanced from its old value U(t-1) at time t-1 to its new value U(t) at time t, one performs

U_nudged(t)=alpha*U(t)+(1-alpha)*Ureference(t)

where 0 < alpha <= 1. I don't know whether this is done on every time step, or at the end of the (e.g. 6 hour) interval between the fields one has grabbed using ems_prep to "drive" wrf (i.e. provide initial conditoins and the evolving boundary conditions)... this is probably in the documentation somewhere.

I would guess in the case of "observation nudging" one has reference values for only a few variables at a few spots (which might not coincide with grid points?), so the adjustment is more restricted.

Nudging should in principle be useful in case one is attempting to use a dynamical model to downscale met observations to attempt to represent the state at finer scales. It isn't something one would use in the context of a pure forecast.

Post Reply