Phoenix case study runs with sf_urban_physics = 0, but with BEP-BEM either runs into errors or keeps running without outputs except d01's 1st timestep

M

maryamfazeli

New member
I'm running a Phoenix case study for a heatwave period in 2019 with a parent and a nested domain. Gridded urban parameters are ingested using the WRFUP tool for d02 only. Running wrf.exe with BEP-BEM, it either runs into errors (see the attached wrf.exe log and rsl files + namelist.input), or keeps running without outputs except for d01's first time step. This is while I have used the same case setup for other cities without any problems. I have listed my unsuccessful attempts to resolve the issue below. Note that the case runs with sf_urban_physics = 0. I appreciate any suggestions on how to solve the problem. Thanks.

Unsuccessful attempts:
  • running for a different heatwave period in 2022
  • reducing the time step from 25 to 20, 15, 10, 5
  • running with w_damping = 1 and epssm = 0.2, 0.7, and 0.9 as suggested for high topography cases here and here
  • reducing d02 domain size (to lower the terrain height in the top right corner to <1000m from 2000-2500m)
  • using w2w for urban parameters
  • using the default MODIS dataset without any urban parameters ingested (num_land_cat = 21 and use_wudapt_lcz = 0)
  • running without excluding written wrfout variables
  • running for d01 only
  • running with an enlarged d01 (to test a different d01 domain)
  • using more CPUs
Notes:
  • I'm using WRF-Comfort installed on WRF v4.5.2.
  • Please view the geo_em files, here.
 

Attachments

Hi, Just letting you know we've reached out to someone who may know a bit more about this. They will respond as soon as they get a chance.
 
Hi,

Thanks for your reply. I could just fix the errors. The problem was that the following urban fields were filled by default for both d01 and d02 after running geogrid.exe - which is not the case for non-American cities. I zeroed them out and handled them only through W2W and WRFUP. This aligns with my observation that the domain could be run with sf_urban_physics = 0 (which uses no urban parameter) and not BEP-BEM. The default values filled by WPS were causing problems.

The fields: URB_PARAM, FRC_URB2D, CANFRA (2011 NLCD Canopy Percent), IMPERV (2011 NLCD Imperviousness Percent)

Best,
Maryam
 
You are right. By default, only CONUS region has the valid values for those urban maps. You will need to prepare the data for your specific case. Another note that you may want to keep in mind is that even for those urban parameters within CONUS, they may not be fully consistent across each other over some urban regions. For example, one grid with higher urban fraction may not have valid URB_PARAM values, or vice versa. Just a caution.
 
cenlinhe said:
You are right. By default, only CONUS region has the valid values for those urban maps. You will need to prepare the data for your specific case. Another note that you may want to keep in mind is that even for those urban parameters within CONUS, they may not be fully consistent across each other over some urban regions. For example, one grid with higher urban fraction may not have valid URB_PARAM values, or vice versa. Just a caution.
Click to expand...
Noted. Thank you so much for your reply.
 
You must log in or register to reply here.
Back
Top