WRF Northern Hemisphere Polar Projection

[wangb]

When I use WRF to simulate the northern hemisphere, the grid distance is 108km. The simulated 2m temperature is up to 10 degrees different from the NCDC observation. Is my simulation wrong? Or is it a resolution issue?
I sincerely hope that someone who has simulated similar cases can share a reasonable method for model testing.

 

[kwerner]

Are you saying that you are setting dx/dy = 108000?

 

[wangb]

Are you saying that you are setting dx/dy = 108000?

yes!

 

[kwerner]

Do you know the resolution of the input (meteorological) data you're using? 108 km is probably entirely too coarse. Your d01 should have a ratio of about 3:1 or 5:1 that of the input data. So, for e.g., if you're using GFS 0.25 degree data, DX/DY or d01 could be set to 9000 (9 km) and this would be reasonable.