Push notifications in your browser are not yet configured.

You are not logged in, you may not see all content and functionalities. If you have an account, please login .

Push notifications in your browser are not yet configured.

## Hail variables in output

Dear colleagues, I have maybe a foolish question, associated with many hail variables in model output. I have run model with itype_gscp=4, and included in output list TQH , HAIL _GSP variables. But, the output lffd… are not containing these variables. I have seen many presentations including ones from COSMO -seminars, and it states about comprising hail parameterization in COSMO model code. As I have understand, it was included in itype_gscp=4 (‘graupel scheme’), or I’m not right?

Could you suggest, which namelist parameters values I should to set to have TQH and HAIL _GSP in output?

Thanks a lot!

Dear Vladimir,

To produce HAIL output, you need to use the second moment scheme. This is not a namelist option. It requires the model to be compiled with this scheme (including the second moment scheme code and the flag - DTWOMOM _SB).

However, if you want to produce realistic amount of hail, you also need to use a relevant resolution (same is true for the “graupel scheme”).

At the resolution you use, the convective plumes are not resolved enough to produce realistic vertical wind speeds. This results in an severe underestimation of your graupel estimation.

To summarise, if you want to produce graupel output, increase the resolution of your model to the convection-permitting scale (e.g. 0.0275°) (this will be very very expensive for your domain) and include the second moment scheme (about 40% increase in computational cost).

If you want more information on this or have some more questions, do not hesitate to contact me directly.

Kind regards,

Erwan

Sorry small mistake:

Dear Vladimir,

To produce HAIL output, you need to use the second moment scheme. This is not a namelist option. It requires the model to be compiled with this scheme (including the second moment scheme code and the flag - DTWOMOM _SB).

However, if you want to produce realistic amount of hail, you also need to use a relevant resolution (same is true for the “graupel scheme”).

At the resolution you use, the convective plumes are not resolved enough to produce realistic vertical wind speeds. This results in an severe underestimation of your graupel estimation.

To summarise, if you want to produce

hailoutput, increase the resolution of your model to the convection-permitting scale (e.g. 0.0275°) (this will be very very expensive for your domain) and include the second moment scheme (about 40% increase in computational cost).If you want more information on this or have some more questions, do not hesitate to contact me directly.

Kind regards,

Erwan

Dear Erwan, thanks a lot for your explanation and useful recommendations! Of course, I have planned to increase resolution of my model run (via downscaling, i.e. the domain area would be smaller), therefore, I will try to take into account your suggestions and to compile a new binary.