Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

spectral nudging (fdda=2) in nested domains

This post was from a previous version of the WRF&MPAS-A Support Forum. New replies have been disabled and if you have follow up questions related to this post, then please start a new thread from the forum home page.

vinodkumar

New member
Hello,
I am trying to run WRF with spectral nudging only for the d01 domain in my nested domain set-up according to the namelist as attached in this thread. What I observe is that there are large differences in the output for d01 if I enable nudging (max_dom=2) or not (max_dom=1). In my opinion, d01 output should be identical of max_dom if I do 1-way nesting (feedback=0). As a side note, the nudging is much weaker if I set up max_dom=2. I have attached a plot of monthly mean wind speeds at 14 UTC for ERA5 (nudging data) and the two WRF outputs (max_dom = 1 &2) herewith.
Any leads will be much appreciated.

Best regards,
Vinod
 

Attachments

  • wrffdda_ws.png
    wrffdda_ws.png
    2 MB · Views: 927
  • namelist.input
    10.8 KB · Views: 72
Dear Ming Chen,
Sorry if the question was not clear. Here I try to summarize my problem and thoughts again.
I want to perform simulations in a one-way nested domain (feedback=0) such that spectral nudging is enabled in the outer domain (d01).
Spectral nudging is performed with respect to ERA5 reanalysis data (3-hour resolution). This is also plotted in the rightmost panel.
The WRF simulated wind speeds are shown in the middle panel.

I also performed another simulation without nesting (max_dom=1). My problem is that the WRF simulation results are different in this case as compared to the previous case.
In my opinion, the simulations should be identical in both cases because in the first case, feedback is turned off.

Hope I am clearer this time.

Best regards,
Vinod
 
Vinod,
Thanks for the clarification. Would you please let me know which version of WRF did you run?
Also, would you please run two test cases without spectral nudging, i.e.,
(1) run a single domain case (max_dom=1, grid_fdda =0)
(2) run a nesting case with feedback off (max_domain = 2, feedback =0, grid_fdda =0)
Then compare the results of D01 from the two tests. Please let me know whether there is large difference between them. Thanks.
 
Dear Ming,
I am using WRF-4.2.2 version.
Many thanks for suggesting these diagnostics. I can now see that the differences were there even before spectral nudging. Here are the respective plots. The plot on the left is for a single domain and that on the right with max_dom=2. As you recommended, grid_fdda =0 in both the runs.
However, I am still not sure, why this is happening.
Thanks a lot for your kind assistance.

Best regards,
Vinod
 

Attachments

  • wrf_ws.png
    wrf_ws.png
    2.2 MB · Views: 897
Hi Vinod,
I run two cases, one is a single domain case and the other is a nested case with feedback off. I can confirm your concern that the results of D01 are different between the two cases. Theoretically, we expect that results should be identical. However, because of the difference in physics/dynamics between the two domains, somehow the child domain still exert certain impacts on results of its parent domain. Turning feedback off cannot completely isolate the child domain from its parent domain.
 
Dear Ming,
Thanks a lot for checking this for me. Though this behaviour is not ideal, I understand your explanation.
I would like to add one further observation.
The differences are even stronger if I run WRF-Chem. For WRF only, it is not that huge.

Best regards,
Vinod
 
Vinod,
Probably you can post your question to wrf-chem forum. Chemical processes may be involved in the nested case and further increase the difference between parent and child domain.
 
Top