Configuration for new MLCAPE in the new beta

[PerryShafran-NOAA]

Hi, all,

I have been asked to test the new capability in dtcenter/MET#1824 to test for Mixed-Layer CAPE (MLCAPE). I am unsure of the settings to use in the configuration for PB2NC and PointStat to be able to test, since this is a new capability I haven't used before. Can someone assist with the settings here?

Thanks!

Perry

[TaraJensen]

@hsoh-u - This is time sensitive. Can you help Perry out?

[JohnHalleyGotway]

@PerryShafran-NOAA, I saw your email to @jprestop and I, but figure I'd followup here.

Turning on the output for MLCAPE in the PB2NC config file should be pretty straight-forward. Here's an example where @hsoh-u added it to a config file run in MET's unit tests:

obs_bufr_var = ["D_CAPE", "D_PBL", "D_MLCAPE"];

The "D_MLCAPE" tells PB2NC to derive it. Also note, that this line has been added to the default PB2NC config file:

obs_prepbufr_map = [
...
   { key = "D_MLCAPE"; val = "MLCAPE"; }
];

So you request it as "D_MLCAPE" in "obs_bufr_var" and you should see observations of "MLCAPE" show up in the output. Increasing the verbosity level should print more details about the derivation. I'm not sure what verbosity level to recommend though.

[PerryShafran-NOAA]

Thanks, John! Can you also assist with the settings in PointStat? Did you use an NCEP model file such as NAM or RAP?

Right now, the setting for surface-based CAPE is:

FCST_VAR11_NAME = CAPE
FCST_VAR11_LEVELS = L0
FCST_VAR11_OPTIONS = cnt_thresh = [ >0 ];
FCST_VAR11_THRESH = >500, >1000, >1500, >2000, >3000, >4000
OBS_VAR11_NAME = CAPE
OBS_VAR11_LEVELS = L0-100000
OBS_VAR11_OPTIONS = cnt_thresh = [ >0 ]; cnt_logic = UNION;
OBS_VAR11_THRESH = >500, >1000, >1500, >2000, >3000, >4000

What changes for MLCAPE?

Thanks!

Perry

[JohnHalleyGotway]

@PerryShafran-NOAA I have not actually run Point-Stat to verify MLCAPE. If your model output file contains Mixed Layer CAPE data, and its named MLCAPE in the GRIB1/2 output file, it should be as easy as adding a new entry to your METplus config file. I assumed you'd want to use the same setup as you're using for CAPE?

OBS_VAR12_NAME = MLCAPE
OBS_VAR12_LEVELS = L0-100000
OBS_VAR12_OPTIONS = cnt_thresh = [ >0 ]; cnt_logic = UNION;
OBS_VAR12_THRESH = >500, >1000, >1500, >2000, >3000, >4000

[PerryShafran-NOAA]

Interestingly, there is no grib value output of MLCAPE. I have the following types of CAPE (this is from RAP):

230:11382185:vt=2022030301:surface:2 hour fcst:CAPE Convective Available Potential Energy [J/kg] - this is surface-based CAPE
289:15257835:vt=2022030301:180-0 mb above ground:2 hour fcst:CAPE Convective Available Potential Energy [J/kg
292:15507077:vt=2022030301:90-0 mb above ground:2 hour fcst:CAPE Convective Available Potential Energy [J/kg]
294:15541945:vt=2022030301:255-0 mb above ground:2 hour fcst:CAPE Convective Available Potential Energy [J/kg]
299:15810877:vt=2022030301:0-3000 m above ground:2 hour fcst:CAPE Convective Available Potential Energy [J/kg]

As you see, I have many different layers to choose from. Which is the one I should be using, and how would I post that in PointStat?

Thanks!

Perry

[PerryShafran-NOAA]

But the more important thing is what I should be looking for in the PB2NC output to know that this worked. Because I'm not finding anything specific to MLCAPE derivation in the PB2NC output.

[hsoh-u]

@hertneky said the 0-90 mb layer for MLCAPE (UPP)

[PerryShafran-NOAA]

@hsoh-u @JohnHalleyGotway Yep, got that info from colleagues here.

Now how would I read that particular record in PointStat? I'm not sure how to write the FCST_VAR_LEVELS in PointStat to be sure I am reading the correct record.

[hsoh-u]

[log messages]
pb2nc does not log messages for mlcape. The number of CAPE and MLCAPE data is logged with the debug level 3.

Case 1: D_CAPE only
Case 2: D_MLCAPE only
Case 3: Both D_CAPE and D_MLCAPE

log_cape_only:DEBUG 3: Derived CAPE = 184      Zero = 62

log_mlcape_only:DEBUG 3: Derived CAPE = 184       Zero = 84

log_both_capes:DEBUG 3: Derived CAPE = 368   Zero = 146

[R script to see the MLCAPE]

Rscript  met/share/met/Rscripts/pntnc2ascii.R prepda.2018091100.add_cape.nc | grep MLCAPE
ADPUPA 78988 20180911_000000 12.200  -68.970    54 MLCAPE NA     0.00000 2   1398.900000
ADPUPA 80001 20180911_000000 12.580  -81.700     2 MLCAPE NA     0.00000 2   1510.900000
ADPUPA 78954 20180911_000000 13.070  -59.500    47 MLCAPE NA     0.00000 2   1343.400000
ADPUPA 78384 20180911_000000 19.300  -81.370     3 MLCAPE NA     0.00000 2   2281.200000
...

[PerryShafran-NOAA]

So it should look like this?

FCST_VAR17_NAME = CAPE
FCST_VAR17_LEVELS = P90-0
FCST_VAR17_OPTIONS = cnt_thresh = [ >0 ];
FCST_VAR17_THRESH = >500, >1000, >1500, >2000, >3000, >4000
OBS_VAR17_NAME = MLCAPE
OBS_VAR17_LEVELS = L0-100000
OBS_VAR17_OPTIONS = cnt_thresh = [ >0 ]; cnt_logic = UNION;
OBS_VAR17_THRESH = >500, >1000, >1500, >2000, >3000, >4000

The FCST_VAR_NAME says CAPE and not MLCAPE because it is labelled as CAPE in the grib2 file - is that correct?

Thanks!

[PerryShafran-NOAA]

OK, this seemed to work. I think that we are good but I wanted to make sure with folks like Geoff and Logan that the MLCAPE values are OK before proceeding. We can talk about this at the Monday METplus user meeting.

Perry

[georgemccabe]

Hi @PerryShafran-NOAA, I am just following up on this discussion. Did you confirm with Geoff and Logan that the MLCAPE values are OK?

[PerryShafran-NOAA]

Yes, we are good on that end! I apologize for not reporting sooner, but our team is satisfied with the development of MLCAPE in the new version of METplus.

[georgemccabe]

Great to hear, thanks!