Entrainment Rate and Coefficient in CAM6?

[bdav010]

Hi All,

I want to play around with the entrainment rate and the entrainment coefficient within the current SCAM model. In the previous iteration of the model (CAM5), this seems to be explicitly calculated such that the values in the equation can be modified. I'm wondering if this is still the case for CAM6? I know that CLUBB replaced a few things but I can't definitively determine if this is now handled implicitly in CLUBB or if it has been moved somewhere else. Any information someone can provide on this would be greatly appreciated. Thanks!

Bri

 

[aherring]

Can you clarify what you did in CAM5 that you're wanting to do in CAM6? Specify the entrainment rate in the CAM5 shallow (UW) and/or deep (ZM) schemes?

 

[bdav010]

I haven't ever done anything in CAM5. Based on documentation I found (image below), I'm looking to change the entrainment rate w_e and entrainment coefficient, A. However I can't seem to find the comparable equation for CAM6 in any documentation that would lead me to where I could do this in the source code. I have CAM6 cloned to my code directories along with the necessary forcing data and SCAM code I want to use. But I am trying to figure out whether this process is now in CLUBB and I need to find an alternative way (like using CAM5) to experiment with the entrainment coefficient and rate within the model.

 

[aherring]

That particular entrainment rates is for UW moist turbulence scheme, which does PBL mixing. There are two other convection schemes in CAM5, UW shallow and ZM deep, which have their own entrainment formulations. See 4.54->4.56 in that CAM5 doc. I'll note the buoyancy sorting makes it difficult to really control the entrainment flux in the UW shallow scheme.

in CAM6, CLUBB handles PBL mixing and shallow convection. Deep convection is still handled by ZM. You cannot directly control the entrainment rate in CLUBB due to the formulation of the problem as prognostic moments with a PDF based closure. You can control dissipation coefficients for each of the moments and you can also play around with the mixing length, which along with prognostic tke controls the dissipation time-scale used for the prognostic equations.

 

[bdav010]

Thank you for the response! I thought that was the case but I wasn't sure if CLUBB took over that aspect. I was looking to alter the scheme for PBL mixing but instead I'll try changing the moment coefficients and mixing length. Thank you again.