Failed run of ensemble_stat in METplus 5.0.0

[BinbinZhou-NOAA]

After modified the ensemble_stat conf file following the instruction and use case example of METplus/5.0.0, the run is still not
successful. The output_base directory shows that the stage sub-directory is not created, and ensemble_stat.log file is not created either, so I can not see the error message.
The Ensemble_stat conf file is like below (all external variables are defined and exported in another script file)

###############################################
[config]

OUTPUT_BASE = {ENV[output_base]}
LOG_LINE_FORMAT =

LOOP_BY = VALID
VALID_TIME_FMT = %Y%m%d%H
VALID_BEG = {ENV[vday]}{ENV[vbeg]}
VALID_END = {ENV[vday]}{ENV[vend]}

METPLUS_PATH = {ENV[METPLUS_PATH]}
VALID_INCREMENT = {ENV[valid_increment]}

LOOP_ORDER = times

PROCESS_LIST = EnsembleStat

ENSEMBLE_STAT_ENS_VLD_THRESH = 0.5
ENSEMBLE_STAT_ENS_THRESH = 0.5

FCST_VAR1_NAME = HGT
FCST_VAR1_LEVELS = P500
BOTH_VAR1_NAME = {FCST_VAR1_NAME}
BOTH_VAR1_LEVELS = {FCST_VAR1_LEVELS}

lead = {ENV[lead]}
LEAD_SEQ = {lead}
MODEL = {ENV[MODEL]}
model = {ENV[model]}

ENSEMBLE_STAT_REGRID_TO_GRID = {ENV[regrid]}

OBTYPE = {ENV[OBTYPE]}
MASKPATH = {ENV[maskpath]}
TMP_DIR = {OUTPUT_BASE}/tmp

ENSEMBLE_STAT_DESC = NA
ENSEMBLE_STAT_REGRID_METHOD = NEAREST
ENSEMBLE_STAT_REGRID_WIDTH = 1
ENSEMBLE_STAT_REGRID_VLD_THRESH = 0.5
ENSEMBLE_STAT_REGRID_SHAPE = SQUARE

METPLUS_ENS_SSVAR_BIN_SIZE = 1000000
METPLUS_ENS_PHIST_BIN_SIZE = 0.05

ENSEMBLE_STAT_MASK_GRID = G003
ENSEMBLE_STAT_MASK_POLY = {MASKPATH}/G003_NHEM.nc, {MASKPATH}/G003_SHEM.nc, {MASKPATH}/G003_TROPICS.nc, {MASKPATH}/Bukovsky_G212_CONUS.nc, {MASKPATH}/Bukovsky_G212_CONUS_East.nc, {MASKPATH}/Bukovsky_G212_CONUS_West.nc
METPLUS_CI_ALPHA = 0.05

ENSEMBLE_STAT_INTERP_FIELD = BOTH
ENSEMBLE_STAT_INTERP_VLD_THRESH = 1.0
ENSEMBLE_STAT_INTERP_SHAPE = SQUARE
ENSEMBLE_STAT_INTERP_TYPE_METHOD = NEAREST
ENSEMBLE_STAT_INTERP_TYPE_WIDTH = 1

ENSEMBLE_STAT_OUTPUT_FLAG_ECNT = BOTH
ENSEMBLE_STAT_OUTPUT_FLAG_RPS = NONE
ENSEMBLE_STAT_OUTPUT_FLAG_RHIST = NONE
ENSEMBLE_STAT_OUTPUT_FLAG_PHIST = NONE
ENSEMBLE_STAT_OUTPUT_FLAG_ORANK = NONE
ENSEMBLE_STAT_OUTPUT_FLAG_SSVAR = NONE
ENSEMBLE_STAT_OUTPUT_FLAG_RELP = NONE

ENSEMBLE_STAT_NC_ORANK_FLAG_LATLON = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_MEAN = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_RAW = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_RANK = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_PIT = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_VLD_COUNT = TRUE
ENSEMBLE_STAT_NC_ORANK_FLAG_WEIGHT = FALSE

GRID_WEIGHT_FLAG = COS_LAT

modelhead = {ENV[modelhead]}
modelpath = {ENV[modelpath]}
modelgrid = {ENV[modelgrid]}
modeltail = {ENV[modeltail]}

gdaspath = {ENV[gdaspath]}
gdashead = {ENV[gdashead]}
gdasgrid = {ENV[gdasgrid]}

climhead = {ENV[climhead]}
climpath = {ENV[climpath]}
climgrid = {ENV[climgrid]}
climtail = {ENV[climtail]}

members = {ENV[members]}

location of ensemble_stat and grid_stat MET config files

ENSEMBLE_STAT_CONFIG_FILE = {METPLUS_BASE}/parm/met_config/EnsembleStatConfig_wrapped

#Defines the name of final metplus conf
METPLUS_CONF = {OUTPUT_BASE}/final_ens_{model}_{ENV[vbeg]}_upper.conf

FCST_ENSEMBLE_STAT_INPUT_DATATYPE = GRIB
OBS_ENSEMBLE_STAT_INPUT_GRID_DATATYPE = GRIB

ENSEMBLE_STAT_N_MEMBERS = {members}

LOG_LEVEL = DEBUG

LOG_ENSEMBLE_STAT_VERBOSITY = 2
ENSEMBLE_STAT_OUTPUT_PREFIX = {MODEL}_g2g_BIN1_FHR{lead?fmt=%3H}

[dir]
FCST_ENSEMBLE_STAT_INPUT_DIR = {modelpath}
OBS_ENSEMBLE_STAT_GRID_INPUT_DIR = {gdaspath}
ENSEMBLE_STAT_OUTPUT_DIR = {OUTPUT_BASE}/stat/{model}

ENSEMBLE_STAT_CLIMO_MEAN_FILE_NAME = {climpath}/mean_{valid?fmt=%m%d}
ENSEMBLE_STAT_CLIMO_MEAN_USE_FCST = True
ENSEMBLE_STAT_CLIMO_STDEV_FILE_NAME = {climpath}/stdev_{valid?fmt=%m%d}
ENSEMBLE_STAT_CLIMO_STDEV_USE_FCST = True

[filename_templates]

FCST_ENSEMBLE_STAT_INPUT_TEMPLATE = atmos.{init?fmt=%Y%m%d}/{modelhead}/{modelhead}.ens??.t{init?fmt=%H}z.{modelgrid}{lead?fmt=%3H}{modeltail}

OBS_ENSEMBLE_STAT_GRID_INPUT_TEMPLATE = atmos.{valid?fmt=%Y%m%d}/{modelhead}/{gdashead}.t{valid?fmt=%H}z.{gdasgrid}.f000.grib2

[BinbinZhou-NOAA]

One Bug in my conf file: BOTH_VAR should be OBS_VAR. After fixing this bug, the run is successful!