Ferrier Microphysics in ARW?

Topics related to current and future physics in the WRF as well as any problems you may have.

Ferrier Microphysics in ARW?

Postby tletcher12 » Wed Apr 25, 2018 1:41 pm

Hi all,

I'm using WRF Version (3.8.1) ARW core to run a physics-ensemble and I have been including a run with the Ferrier Microphysics (option 5) since I'm trying to stick with physics options that are well validated and use, and from what I can gather, this is the microphysics scheme used operationally in the NAM.

I'm seeing some pretty funky results from the Ferrier scheme, both quantitive and qualitative differences. These differences are especially pronounced when looking at snow and snow mixing ratio. Furthermore, the Ferrier scheme does not seem to put any data into the accumulated snow (SNOWNC) variable, despite the fact that snow is included as a prognostic species. I'm running simulations at both convective resolution and convection parameterized scales, the issues appear in both.

So I'm curious if anyone has any idea as to why Ferrier seems to be a bit off, and is missing some of the output variables. My only speculative thought is that since Ferrier is operational in the NAM, is it best to use with the NMM core? or should it work equally as well in ARW? Is anyone aware if I need to turn on another physics switch (e.g., boundary layer or cumulus parameterization) to run Ferrier correctly? Should I use something other than Ferrier to fill out the ensemble? Morrison perhaps?

I don't have a ton of experience using Ferrier, but from what I can gather, Ferrier looks like a solid scheme, so I don't know why it appears to be "behaving badly"

Any insight is appreciated.
tletcher12
 
Posts: 9
Joined: Sat Dec 18, 2010 9:38 pm

Re: Ferrier Microphysics in ARW?

Postby sudheer717 » Mon May 28, 2018 11:56 am

Hi,

I have worked briefly with both Ferrier and Morrison-2 Moment Microphysics (M2M, option 10) schemes in ARW. For dry cases, Ferrier gave very good values when compared with observations, but recently for a weakly convective case I've been working on, M2M gave better results than Ferrier (it underestimated the values by 50-70%).

As far as the missing output variables, look into the Registry files and try adding the variables you want to an aux-history stream.

HTH
--
Sudheer
--
Sudheer
Ph.D. Candidate
Univ of Texas
San Antonio, USA
sudheer717
 
Posts: 8
Joined: Mon Jan 18, 2016 8:46 pm


Return to Physics

Who is online

Users browsing this forum: No registered users and 2 guests

cron