METplus Use Case: PointStat - Standard Verification of CONUS Surface
This use case utilizes the MET Point-Stat tool.
Optional: Refer to the MET Users Guide for a description of the MET tools used in this use case.
Optional: Refer to the METplus Config Glossary section of the METplus Users Guide for a reference to METplus variables used in this use case.
-
View Configuration File
Many of the options are similar, except the fields typically at the surface (BOTH_VARn_LEVELS = Z2 and Z10 or integrated BOTH_VARn_LEVELS = L0). Also, the Point-Stat MESSAGE_TYPE is set to a specific keyword ONLYSF for only surface fields. Finally, the PB2NC_INPUT_TEMPLATE includes a da_init specification in the input template to help determine the valid time of the data.
...
PB2NC_INPUT_TEMPLATE = nam.{da_init?fmt=%Y%m%d}/nam.t{da_init?fmt=%2H}z.prepbufr.tm{offset?fmt=%2H}
Using this configuration file, you should be able to run the use case using the sample input data set without any other changes.
-
Run the PointStat use case
${METPLUS_BUILD_BASE}/parm/use_cases/model_applications/medium_range/PointStat_fcstGFS_obsNAM_Sfc_MultiField_PrepBufr.conf \
${METPLUS_TUTORIAL_DIR}/tutorial.conf \
config.OUTPUT_BASE=${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc
This may take a few minutes to run.
-
Review the output files
Inspection of the file shows that statistics for TMP, RH, UGRD, VGRD, UGRD_VGRD are available. Also, based on the number listed after the line type (SL1L2 and VL1L2), there are between 8263 - 9300 points included in the computation of the statistics. The big question is why are there no statistics for TCDC and PRMSL? Let's look at the log files.
Open the log file and search on TCDC, you will see that there is an error message stating "no fields matching TCDC/L0 found" in the GFS file. That is why TCDC does not appear in the output.
Look for PRMSL/Z0 in the log file. We can see the following:
DEBUG 2: Number of matched pairs = 0
DEBUG 2: Observations processed = 441178
DEBUG 2: Rejected: station id = 0
DEBUG 2: Rejected: obs var name = 441178
You will note, the number of observations processed is the same as the number rejected due to a mismatch with the obs var name. That suggests we need to look at how the OBS variable for PRMSL is defined.
-
Inspect configuration file and plot fields
Note that PB2NC_OBS_BUFR_VAR_LIST = PMO, TOB, TDO, UOB, VOB, PWO, TOCC, D_RH, where PMO is the identifier for MEAN SEA-LEVEL PRESSURE OBSERVATION according to https://www.nco.ncep.noaa.gov/sib/decoders/BUFRLIB/toc/prepbufr/prepbufr_bftab/. Let's use Plot-Data-Plane to confirm this identifier will provide valid observations for Point-Stat to use.
${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc/nam/conus_sfc/20170601/nam.2017060100.nc \
${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc/nam/conus_sfc/20170601/nam.2017060100.ps \
-obs_var PMO
${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc/nam/conus_sfc/20170601/nam.2017060100.ps \
${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc/nam/conus_sfc/20170601/nam.2017060100.png
display ${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc/nam/conus_sfc/20170601/nam.2017060100.png
-
Update configuration file and re-run
FCST_VAR7_LEVELS = Z0
OBS_VAR7_NAME = PMO
OBS_VAR7_LEVELS = Z0
${METPLUS_TUTORIAL_DIR}/user_config/PointStat_Sfc2.conf \
${METPLUS_TUTORIAL_DIR}/tutorial.conf \
config.OUTPUT_BASE=${METPLUS_TUTORIAL_DIR}/output/PointStat_Sfc2
There is now a line with PRMSL listed and statistics reported.