Box-aligned wave/stripe artifact inside nested refined domain with WRFInput + static AMR + terrain

[gioviciconte]

Hi,

I am testing a real-data nested ERF case initialized from WRF/WPS input files and I am seeing a persistent box-aligned wave/stripe feature inside the finest active refined patch. The artifact appears inside the level-3 / d04 refined patch. In ParaView, the plotted y_velocity field shows two strong, nearly vertical bands close to the d04 lateral boundaries, with a wave-like pattern inside the refined box. The structure appears aligned with the refinement patch rather than with the meteorological flow or terrain. Here a video showing its evolution. (In the gif I plotted x and y velocity in the smaller domain).

I am running ERF with WRF/WPS-generated wrfinput_d0* and wrfbdy_d01 files, using static nested AMR over complex terrain. The setup is intended to mimic a WRF-style nested mesoscale-to-LES cascade. This is the input file erf_inp.txt.

The horizontal refinement ladder is:

d01 / level 0: 11250 m
d02 / level 1: 3750 m
d03 / level 2: 750 m
d04 / level 3: 150 m

Any advice on the recommended ERF workflow for WRFInput + real-data static AMR + terrain would be very helpful.

Thanks.

[hgopalan]

Can you update to the latest version and check if the issues still persist?