Hi, I am getting high dry bias for non local pbl schemes in WRF but temperature output is very good statistically compared to local schemes. What are the options for reducing this bias without changing the pbl scheme (maybe in dynamics options?). I would very much like your input as I have little experience with WRF yet.
Regards
Neeldip
You can try to change other physics options than PBL. I suggest you to test the WDM5 Microphysics scheme (mp_physics = 14). I obtained good results in specific humidity with it.
Other suggestion is to run WRF with Nudging.
I have noticed this behavior using the YSU PBL scheme, which is a nonlocal PBL that causes an overmixed PBL, especially over areas with a well-mixed atmosphere. If you are finding dry bias over monitoring stations in urban areas, I recommend the use of the urban canopy model (sf_urban_physics), even the simpler one ( sf_urban_physics =1). Additionally, I would also activate the mosaic LU categories in NOAH LSM ( sf_surface_mosaic =1). These two options will incorporate sub-grid processes that help in heat flux, temperature, and humidity.
if you are interested, you can read this paper (DOI: 10.1175/WAF-D-14-00105.1) which makes a review of the PBL schemes in WRF.
Be careful with microphysics options, because I already tested some if them over Brazil and I found huge differences in modeled rain.
Regards