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

Lateral Boundary conditions and Temporal resolution for WRF - Best Practice?

J

jorge_p

New member
Hi all,

I was wondering if there is any source/publication/documentation(or even a quick tip) available discussing the advantages and/or disadvantages of using high temporal resolution lateral boundary conditions (e.g. Hourly reanalysis ERA5 for both the surface and the atmosphere).
Here are some points that I've been brainstorming on and might help the discussion regarding LBCs and "more frequent is better":
  • Forecast Accuracy (better or worse?)
  • Computational performance (slow down?)
  • Increased storage needs (large GRIB files, even if cropped over area of interest)
  • Parent domain size and resolution (can we reduce extent and increase spatial resolution? - always taking into account reanalysis available resolution)
  • Applications that would benefit more than others (e.g., high-res urban simulations).
Hope my question and notes make some sense and will be happy to hear your advice/thoughts/links on this.

Best,

Giorgos
 
I will be watching this thread too, as I wondered that too. I always assumed that the relationship between the boundary update timestep and the size of the domain is similar to the CFL condition - you'd want the update time to be smaller than the time needed for the synoptic disturbances to propagate across the domain.

But I am not sure why exactly, that's just a hunch for me.
 
@William.Hatheway thanks for the link(s)!

I try to visit them every time they get updated (and more often than that). Nevertheless I haven't found something that informs, in detail (at least shown on the available slides), on the topic of temporal resolution of LBCs, apart from "the more frequent the better" (hope I haven't missed something).

Thanks again,

Giorgos
 
jorge_p said:
@William.Hatheway thanks for the link(s)!

I try to visit them every time they get updated (and more often than that). Nevertheless I haven't found something that informs, in detail (at least shown on the available slides), on the topic of temporal resolution of LBCs, apart from "the more frequent the better" (hope I haven't missed something).

Thanks again,

Giorgos
Click to expand...
Are you referring to how often you should have input files, ie. 1-hour, 3-hour, etc? @jorge_p
 
We recommend to update lateral boundary as frequently as possible mainly because the large-scale circulation can evolve rapidly, and thus the frequent update can assimilate large-scale information in a more accurate and timely manner.
 
jorge_p said:
Hi all,

I was wondering if there is any source/publication/documentation(or even a quick tip) available discussing the advantages and/or disadvantages of using high temporal resolution lateral boundary conditions (e.g. Hourly reanalysis ERA5 for both the surface and the atmosphere).
Here are some points that I've been brainstorming on and might help the discussion regarding LBCs and "more frequent is better":
  • Forecast Accuracy (better or worse?)
  • Computational performance (slow down?)
  • Increased storage needs (large GRIB files, even if cropped over area of interest)
  • Parent domain size and resolution (can we reduce extent and increase spatial resolution? - always taking into account reanalysis available resolution)
  • Applications that would benefit more than others (e.g., high-res urban simulations).
Hope my question and notes make some sense and will be happy to hear your advice/thoughts/links on this.

Best,

Giorgos
Click to expand...
Update/Comment on my own questions, regarding:
  • Computational performance (slow down?): No significant slow down (~ 10% slower) with 1hr LBCs compared to 6 hr LBCs (3 nests 25km down to 1km - with feedback), size :
e_we = 210, 196, 206,
e_sn = 125, 176, 166,
 
You must log in or register to reply here.
Top