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

Feedback=1 option

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.


New member
Dear WRF Support,

Although I have yet to test in v4+ generation of WRF-ARW, I did note this potential bug in v3911 so suspect it is still present (no updates or fixes in v4 have mentioned it). The issue occurs with the SWDOWN & SWNORM fields along the lateral boundaries of inner (child) nests, but only when (i) feedback=1 and (2) topo_shading=1 & slope_rad=1 are turned on. I have seem with use of the Goddard SW scheme, but also with Dudhia SW (maybe not quite as pronounced). I have not looked at using other SW schemes, but suspect it occurs sing any of them. The hunch is that the issue is with some interaction of feedback=1 interpolations at the boundary when shading or slope is used with a SW scheme. Feedback=0 eliminates the issue. Contouring SWNORM, SWDOWN, and even T2 or TSK fields with very fine contour intervals will show it up near the lateral boundaries when it occurs. In something a bit unrelated, I have also noted that when I place very nice nests (say at sub-km grid spacing) over complex terrain & contour fields such as T2 with very fine contour intervals and after using feedback=1 (smooth=0), I can see reflections of the inner nest boundary within the parent nest I am contouring.
I kind of expect the last issue to exist as a byproduct of the 2-way interpolations necessary in feedback=1 (thus, try to avoid if possible placing nest lateral boundaries over regions of very complex terrain), but I would not expect the issues with SWNORM/SWDOWN fields on the lateral boundaries by using feedback=1 and topo_shading;slope_rad both set to 1. I will attach a few plots. I just want to know if anyone has noted this kind of behavior before, since feedback=1 typically isn't used except for very high res simulations. if necessary, I will collect some of my namelist info used and attach later. But...everything else unchanged, the SWDOWN/SWNORM lateral boundary issue goes away just by changing feedback=1 to feedback=0.

Robert Dumais
US Army Research Lab


  • 932test.png
    115.9 KB · Views: 3,277
  • 932test.png
    115.9 KB · Views: 3,276
  • 935test.png
    485.5 KB · Views: 3,276
Hi Robert,
I apologize for the delay in response. We have all been out of the office for the past couple of weeks and are getting caught-up now. Thank you for reporting this. Take a look at this pull request we have on GitHub:
This *should* fix the problem you're seeing. Can you test this out and let us know? Once you modify the code, you'll need to recompile, and since it's a registry change, you'll first need to issue a './clean -a' and then reconfigure before recompiling.