Problems with hail variables in HAILCAST module

[tmanzoni]

I ran into a problem while using the HAILCAST and LPI suites in the WRF model (version V4.2.2).
I have tried to plot with GrADS and ARWpost the following variables: HAILCAST_DHAIL1, HAILCAST_DHAIL2, HAILCAST_DHAIL3, HAILCAST_DHAIL4, HAILCAST_DHAIL5, HAILCAST_DIAM_MAX, HAILCAST_DIAM_MEAN, HAILCAST_DIAM_STD and LPI.
I added that vaiables modifing the Registry file and recompiling the WRF model.
The vairables HAILCAST_DIAM_MEAN and HAILCAST_DIAM_STD were plotted incorrectly (i used ARWpost and GrADS for post processing).
First i thought that somenthing went wrong during the simulation, so i checked the namelist.input file (in the attachement you can see the &physics used for the simulation) and ran again the simulation (both real.exe and wrf.exe) but nothing changed and I had the same error after plotting the maps again.
Finally I made a ncdump -v of the wrfout finding that the error was in the wrfout files. These two variables were completely out of scale; the variables should show hailstone's mean diameter and his standard deviation in millimeters.

Has anyone run into this problem?

In the immages below you can see an example of what GrADS plotted, the physics used while running the WRF model and an extract of the ncdump of the wrfout file

 

[Ming Chen]

Can you upload a single wrfout file for me to take a look? Also, please attach your full namelist.input. Thanks.

 

[tmanzoni]

Here the complete namelist.
I've followed the instruction on your homepage but seems there are problems in uploading the wrfout file on your Nextcloud (as you can see from the image).

 

[Ming Chen]

In your namelist.input, haildt = 0. However, the scheme is run only at every "haildt" seconds. I am not sure whether this is an issue.
Would you please specify haildt, then rerun the case? Please let me know whether this gives you reasonable results.

 

[tmanzoni]

I've run three different simulations using haildt = 10, haildt = 30, haildt=60; WRF sitll gives me unreasonable results about hail (negative values or hailstone 4 order of magnitude bigger than usual).

wrfout.tar.gz here you can download the wrfout (with haildt=0) file, i'm sorry that i sitll can't upload it on your Nextcloud account

 

[Ming Chen]

Sorry that I don't have an immediate answer to this issue ....

 

[tmanzoni]

Ok never mind.
Checking on WRF 's modules, I've seen that HAILCAST_DIAM_MEAN is the arithmetical mean of "HAILCAST_DHAIL" variables.
i'll calculate this this variable without the WRF's routine and overcame the problem.

Thank you

 

[tmanzoni]

Sorry that I don't have an immediate answer to this issue ....

Hello!
During this week I've ran other simulations, seems that the problem comes out when i'm turning on the lightning potential index (lightning_option = 3 in the name list), the hailcast module is ok.
Searching in the Forum i've found that other people had problems with the same options ( Large differences of simulated fields (U, V, W, T) with lightning_option = 3 (LPI) on and off ).

I'm using V4.2.2, do you know if someone has fixed this bug in V4.4? maybe i can resolve this problem uploanding new version

 

[Ming Chen]

Thanks for the update. I don't think there is bug fix done for the issue related to lightning_option.
Please avoid using this option at present. Also, not use the full spectral bin microphysics until it is updated. Use the fast version instead.

 

[n umakanth]

SIr, i am also facing the same issue when using lightning option. Please help me sir. Wihtout lightning option our WRF out file has no issue. WIth lighnting option why is the problem occuring.
&physics
physics_suite = 'CONUS'
mp_physics = 6, 6, -1,
do_radar_ref = 1,
cu_physics = 0, 1, 0,
ra_lw_physics = 1, 1, -1,
ra_sw_physics = 4, 4, -1,
bl_pbl_physics = 1, 1, -1,
sf_sfclay_physics = 91, 91, -1,
sf_surface_physics = 1, 1, -1,
radt = 12, 10, 30,
bldt = 12, 0, 0,
cudt = 5, 5, 5,
icloud = 1,
num_land_cat = 21,
sf_urban_physics = 0, 0, 0,
lightning_option = 3, 3, 3,