Changes to the operational FVS database as part of SREF implementation:
- All Eta and RSM members dropped, replaced by additional WRF-NMM/ARW members and NEMS-NMMB members
- WRF-ARW: srarwctl, srarwn1, srarwn2, srarwn3, srarwp1, srarwp2, srarwp3
- WRF-NMM: srnmmctl, srnmmn1, srnmmn2, srnmmn3, srnmmp1, srnmmp2, srnmmp3
- NEMS-NMMB: srnmbctl, srnmbn1, srnmbn2, srnmbn3, srnmbp1, srnmbp2, srnmbp3
Changes to the operational FVS database:
- Rapid Update Cycle (RUC), replaced with Rapid Refresh (RAP), all "ruc" vsbd files
now named "rap"
- Add verification of Best (most unstable) CAPE
- Verifiction of CAPE for NAM/GFS now using field computped from virtual temperature
- Mesonet observations (MSONET) now separated from other surface data (ONLYSF)
- Obsolete ANYSFC verification type removed
- Big fixes:
- Restore verification of grids over Hawaii
- Fixed bug from 10/18/2011 package: on 10/18/11, code was changed to do the
model-to-ob location interpolation using nearest neighbor instead of lower left corner.
Changes to the operational FVS database in conjunction with the 18 October 2011 NAM implementation:
- Added the new NAM nested runs (0-60 h)
- 4 km CONUS
- 6 km Alaska
- 3 km Hawaii
- 3 km Puerto Rico
- Placeable fire weather nest (1.33 km if over CONUS; 1.5 km if over Alaska, 0-36 h)
- Added verifications of the NAM and RUC over the fire weather nest domain
- Changed most model FVS jobs to process 20 pressure levels (up to
50 mb) instead of 19. To compute 50 mb verification stats, the prepfits code was extrapolating pressure level data at 100 mb up to
50 mb, causing spurious 50 mb verification numbers.
- Fixes to code errors relating to visibility, CAPE, and PBL.
Code fix in parallel NAM verifications to allow for consistent ob counts for surface dewpoint verifications
between all models; previously, higher resolution grids had higher ob counts, leading
to inconsistent verification between the parallel 12 km NAM parent run and the nested runs.
The problem will be fixed in the operational NAM verification at a later date.
Changed all parallel NAM (including nests) jobs to process 20 pressure levels (up to
50 mb) instead of 19; prepfits code was extrapolating pressure level data at 100 mb up to
50 mb, causing spurious 50 mb verification stats. The problem will be fixed in the operational
NAM verification at a later date.
RUC verification extended to 18-h; this was accidently turned off in the 4/6/2010
||J. Zhu, Rogers, Shafran
Changes to the operational FVS database:
- A fix to the mismatch of the start hours for the NAMHW. This is the verification
for the NAM model on specific regions to be compared with the HIRESW model. Since in the
operational HIRESW runs, a certain domain or region only runs at certain cycle(s) of the
date (for example, the "east" domain runs at 00 and 12z only, "west" at 06z and "alaska"
at 18z), we will have to read in the input hours for the NAMHW verification to be same
as for the HIRESW.
- A change to verify the HIRESW on 5km grid (currently it's doing the 8km grid).
- Add the verification for GFSE. This is the verification for the GFS model from
90-192 hours for the comparison with the DGEX. Add verifications on the DGEX and GEFS for the
Alaska area and change to verify both models every 6 hours (current production is every 12 hours).
- Add the verification for the RUC First Guess.
- Add the verification for the NAM smartinit grids.
- Add 2 new variables to the verification codes: the Richardson number PBL and
the mixed-layer height. The original PBL is the TKE-based PBL.
- Rename the model name of the RTMA inside the vsdb files from "RTM" to "RTMA"
to be consistent with the model name in the file directory.
- Change the directory name for the RSM from "rsm" to "rsmh", which stands for "RSM Hawaii" region.
- Consolidate the gridtobs input parameter files among the NAM, NDAS, RUC, NAMHW, HIRESW,
SMARTINIT and SREF models.
- Consolidate the gridtobs-e input parameter files among SREFENS and SREFENSX
- A fix to the subroutine getbak.f (in PREPFITS) to output the Relative Humidity
(RH) verification on pressure levels for the GDAS and RUC models, and any other future model
grid that does not output specific humidity on pressure levels.
- Change to use NDAS prepbufr files for all the models (except for the AQM)
for data input consistency in the verification.
- Change the thresholds for verification on Visibility and Cloud-base Height.
- Changes are being made to all the non-aqm verification codes to ensure cleaner,
more graceful exits in times of job stoppages, for whatever reason, using the w3 lib routine
ERREXIT rather than STOP or ABORT.
- Verification of 15-h and 18-h RUC forecasts were inadvertently turned off when this package was
RUC forecast extended from 12 to 18-h; RUC verification extended to 18-h as of this date.
||J. Zhu, Rogers
Changes to the operational FVS database in conjunction with the 27 October 2009 SREF implementation:
Began copying Real-Time Mesoscale Analysis (RTMA) files to /mmb/wd20er/data/vsdb on tempest
(retroactive to 1 October 2009)
- Four WRF members added : NMMN2, NMMP2, ARWN2, ARWP2
- Four Eta members dropped : EDETN, EDETP, ESATN, ESATP
- Bias-corrected SREF mean (SRMEANX) restored
- Began copying SREFENS and SREFENSX to /mmb/wd20er/data/vsdb on tempest (retroactive to 1 October 2009)
||Rogers, J. Zhu, Shafran
Changes to the FVS database at /mmb/wd20er/data/vsdb on tempest:
New database (/mmb/wd20er/data/vsdb_aux) on tempest which contains FVS records of additional fields added by
P. Shafran over the years (not all of these are available for all models):
- E. Rogers' FVS jobs on the CCS turned off, start use of VSDB files produced in operations
- DGEXO renamed DGEX
- All bias-corrected SREF members, which were named ending in "BC" (NMMCTLBC, for example), are renamed
ending in "X" (NMMCTLX) to conform to naming convention in operational FVS jobs
- GFSEO (GFS verified from 90-192 h for comparison to DGEX) turned off
- NGM turned off
- Bias-corrected SREF ensemble mean forecast (SRMEANX) temporarily turned off pending script fix in SREF
- Added verification of specific humidity on pressure levels
- Added verification of GFS 1-deg lat/lon output
- Operational NAM VSDB file is split into "NAM40" (upper air verification using 40 km grids #212 and #216)
and "NAM" (upper air and surface verification using 12 km grids #218 and #242) to provide continuity
with E. Rogers' datebase and to allow FVS to run faster on tempest
- FVS records in NAM, GFS, and all SREF members VSDB files of new fields are extracted from
monolithic VSDB files and put into a new database (see below)
- PBL height
- Lifted Index
- Tropopause Pressure
- Precipitable Water
- Cloud Base Height
- Total Cloud Fraction
- Wind Chill Temperature
- Maximum Wind Gust
- Maximum Temperature
- Minimum Temperature
- All "FHO" statistics for surface variables
Begin verification of bias-corrected output from the NCEP Short-range Ensemble Forecast System (both individual members and SREF mean).
Filenames are the same as the current SREF FVS files, but with "bc" added (e.g., srmeanbc, etc.)
Upgrade to HIRESW runs:
Changes to HIRESW FVS files:
- Increase resolution from 5.2 km to 4.0 km
- Increase resolution from 5.8 km to 5.1 km
- Expanded large domains
- Western and Eastern U.S. domains roughly doubled in size
- 3 CONUS nests reduced to 2 overlapping domains (West-Central & East Central)
- New HIRESW run schedule:
- 0000Z : East/Central U.S., Hawaii
- 0600Z : West/Central U.S., Puerto Rico
- 1200Z : East/Central U.S., Hawaii
- 1800Z : Alaska, Puerto Rico
- CENTNMM, CENTEM, NAMCENTHW dropped
- AKEM renamed AKARW
- EASTEM renamed EASTARW
- HIEM renamed HIARW
- PREM renamed PRARW
- WESTEM renamed WESTARW
- Verifications for EASTNMM, EASTARW, NAMEASTHW, AKNMM,
AKARW, and NAMAKHW changed to reflect new East US and
Alaska HIRSEW model start times
- For upper air data, add verification over the entire
expanded 5 km output grid for the large HIRESW domains:
- East US (grid number 155)
- West US (grid number 156)
- Alaska (grid number 159)
Changed the grid used to define the data retention area for CONUS verifications of the NAM, GFS, RUC, NGM, parallel NAM,
parallel RUC, and all SREF members from grid #212 to grid #236 (set in "keeplist" file)
Began verifying 03z and 15z SREF runs
Turned on verification of 2-m dew point temperature for the following : NAM, RUC, HIRESW, parallel NAM, parallel RUC, all SREF members
Removed verification over DWRFE grid #163
Added the 6 new WRF members of the SREF:
- NMMCTL : WRF-NMM dynamic core, NCEP physics, control
- NMMP1 : WRF-NMM dynamic core, NCEP physics, positive perturbation
- NMMN1 : WRF-NMM dynamic core, NCEP physics, negative perturbation
- ARWCTL : WRF-ARW dynamic core, NCAR physics, control
- ARWP1 : WRF-ARW dynamic core, NCAR physics, positive perturbation
- ARWN1 : WRF-ARW dynamic core, NCAR physics, negative perturbation
Change to verification of SREF:
- Extented verification of ensemble mean and all members to 87-h
- Added verification of Alaska SREF output (grid #216) for ensemble mean and all
Change to output grid from GFS used in FVS:
Change is valid for both "GFS" records (0-84 h) and "GFSEO" records (90-192h)
- Old (up to 30 June 2005) : One-degree lat/lon grid interpolated to 80 km grid #211
- New (starting 1 July 2005): 1/2 degree lat/lon grid interpolated to 40 km grid #212
Changes/additions to FVS model names:
- "FNL" renamed "GDAS"
- "EDS" renamed "NDAS"
- "ETA" discontinued, added "NAM40", which is the operational NAM verified against all upper air data
using the 40 km grid #212 over the CONUS and 45 km grid #216 over Alaska. All previous on-line "ETA" FVS files
were renamed and copied to the "NAM40" directory.
- For the NAM archive (started 11/7/04, see 1/25/2005 entry below) copied all online "ETA" FVS files prior
to 11/7/04 and renamed them "NAM".
Added verification on the 20 km grid #252 for the RUC and RUCX
Changed names of NAM FVS files verified over HiResWindow Model domains:
- ETHA --> NAMAKHW (Alaska HIRESW domain, 00z cycle)
- ETHC --> NAMCENTHW (Central US HIRESW domain, 12z cycle)
- ETHE --> NAMEASTHW (East US HIRESW domain, 18z cycle)
- ETHH --> NAMHIHW (Hawaii HIRESW domain, 00z and 12z cycles)
- ETHP --> NAMPRHW (Puerto Rico HIRESW domain, 06z and 18z cycles)
- ETHW --> NAMWESTHW (Western US HIRESW domain, 06z cycle)
Operational Eta-12 officially renamed North American Meso Model (NAM). Started new
"NAM" verification files on 11/7/2004 which use grids #218 and #242 to verify upper air data in addition
to surface data.
Start of WRF HIRESW system, single NMM run replaced with 2 runs:
the WRF version of the Non-hydrostatic Meso Model
(WRF-NMM) and the WRF version of the NCAR Eulerian Mass Core (WRF-EM). Changes to FVS:
- Previous ops NMM names (EAST, CENT, WEST, AKHW, PRHW, HIHW) dropped
- CENTNMM --> Ops WRF-NMM run for Central US
- AKNMM --> Ops WRF-NMM run for Alaska
- EASTNMM --> Ops WRF-NMM run for Eastern US
- WESTNMM --> Ops WRF-NMM run for Western US
- HINMM --> Ops WRF-NMM run for Hawaii
- PRNMM --> Ops WRF-NMM run for Puerto Rico
- CENTEM --> Ops WRF-EM run for Central US
- AKEM --> Ops WRF-EM run for Alaska
- EASTEM --> Ops WRF-EM run for Eastern US
- WESTEM --> Ops WRF-EM run for Western US
- HIEM --> Ops WRF-EM run for Hawaii
- PREM --> Ops WRF-EM run for Puerto Rico
- Added CENTNMMX --> Parallel WRF-NMM run for Central US
- Added AKNMMX --> Parallel WRF-NMM run for Alaska
- Added EASTNMMX --> Parallel WRF-NMM run for Eastern US
- Added WESTNMMX --> Parallel WRF-NMM run for Western US
- Added HINMMX --> Parallel WRF-NMM run for Hawaii
- Added PRNMMX --> Parallel WRF-NMM run for Puerto Rico
The grid used to verify upper air data (raobs, etc) for model ETHA (Eta model over the Alaska HIRESW
domain) was changed from the 45 km grid #216 to the 11.25 km grid #242.
New 32-km SREF with physics diversity implemented; FVS database for SREF changed as follows:
- Old database names (ectl etn1 etn2 etp1 etp2 ekfc ekp1 ekp2 ekn1 ekn2 rctl rsn1 rsn2 rsp1 rsp2)
for 48-km ops SREF turned off
- New database names will be a continuation of those used for the 32-km SREF parallel:
- ebmjc : Eta, Betts-Miller-Janjic convection, ops Ferrier microphysics, control
- ekfcc : Eta, Kain-Fritsch convection, ops Ferrier microphysics, control
- ebmjn : Eta, Betts-Miller-Janjic convection, ops Ferrier microphysics, negative perturbation
- ebmjp : Eta, Betts-Miller-Janjic convection, ops Ferrier microphysics, positive perturbation
- ekfcn : Eta, Kain-Fritsch convection, ops Ferrier microphysics, negative perturbation
- ekfcp : Eta, Kain-Fritsch convection, ops Ferrier microphysics, positive perturbation
- esatn : Eta, Betts-Miller-Janjic convection w/saturated vapor pressure profiles, experimental
Ferrier microphysics, negative perturbation
- esatp : Eta, Betts-Miller-Janjic convection w/saturated vapor pressure profiles, experimental
Ferrier microphysics, positive perturbation
- edetn : Eta, Kain-Fritsch convection w/full cloud detrainment, experimental
Ferrier microphysics, negative perturbation
- edetp : Eta, Kain-Fritsch convection w/full cloud detrainment, experimental
Ferrier microphysics, positive perturbation
- rsasc : RSM, simplified Arakawa-Schubert convection, Zhou GFS microphysics, control
- rsasn : RSM, simplified Arakawa-Schubert convection, Zhou GFS microphysics, negative perturbation
- rsasp : RSM, simplified Arakawa-Schubert convection, Zhou GFS microphysics, positive perturbation
- rrasn : RSM, relaxed Arakawa-Schubert convection, Zhou GFS microphysics, negative perturbation
- rrasp : RSM, relaxed Arakawa-Schubert convection, Zhou GFS microphysics, positive perturbation
- "srm32" : Ensemble mean for 32-km SREF parallel turned off
Added parallel 32-km EDAS runs : EDASV, EDASL, EDASY, EDASZ, EDASW. Verified 00-h analysis at start
of EDAS and the last 3-h EDAS forecast, both valid at 00z or 12z.
Turned back on verification of surface ship data (PREPBUFR ob types 180/280, 182/282) in all models.
Turned off DGEX (parallel 00Z DGEX) and GFSE
Added DGEXO, operational DGEX from 06z/18z over CONUS, 00z/12z over Alaska. Created new
extended GFS archive called "GFSEO" for comparisons to DGEXO
Changed DGEX and GFSE verification jobs so that they used the same observation data set
Added 00Z DGEX (8-day eta extension parallel) for hours 84-192; added 00Z GFS for hours
84-192; 18z GFS for hours 192, 12Z GFS for hour 204; these GFS verifications designated "GFSE"
Fixed bug in SREF-mean prepfits code which caused the first forecast GRIB file read into
the code to not be used.
||Manikin, Benjamin, Rogers
Switched to using RUC2B 00z/12z analyses in verification scripts; this was originally done on
30 Dec 2002 but was undone at some point by persons unknown.
Added generation of regional verification statistics against rawinsondes over the CONUS using the 3
HIRESW CONUS domains
to split the CONUS into Eastern (G245), Central (G247) and Western (G246) regions.
- WRFC (NCAR WRF, central US) changed to CENTEM (EM=NCAR Eulerian Mass Core)
- WRFA (NCAR WRF, Alaska US) changed to AKEM
- WRFE (NCAR WRF, east US) changed to EASTEM
- WRFW (NCAR WRF, west US) changed to WESTEM
- Added CENTNMM (NMM WRF, central US)
- Added EASTNMM (NMM WRF, east US)
- Added WESTNMM (NMM WRF, west US)
- Added AKNMM (NMM WRF, Alaska)
|2003/09/29 (retroactive to 2003/09/01)
||Ek, Pyle, Rogers
New FVS surface subregions implemented:
Names changed/eliminated/combined with other regions:
- APL = Appalachians verification region
- NEC = Northern East Coast verification region
- SEC = Southern East Coast verification region
- GMC = Gulf of Mexico Coastal verification region
- MDW = Midwest verification region
- LMV = Lower Mississippi Valley verification region
- NPL = Northern Plains verification region
- SPL = Southern Plains verification region
- NMT = Northern Mountain verification region
- SMT = Southern Mountain verification region
- GRB = Great Basin verification region
- SWD = Southwest Desert verification region
- NWC = Northern West Coast verification region
- SWC = Southern West Coast verification region
- NFR = Northern Front Range verification region
- SFR = Southern Front Range verification region
- NMW = Northern Midwest verification region (renamed MDW, southern
boundary moved northward)
- SMW = Southern Midwest verification region (renamed LMV, excludes Gulf
Added verification of 51-84 h forecasts from 06z/18z runs
for the Eta and GFS.
SREF Mean statistics changed to include 5 Eta Kain-Fritsch members (now
a 15 member ensemble mean).
Added 5 SREF Eta Kain-Fritsch members to verification data base.
The official location of the Mesoscale Branch on-line FVS grid-obs
data base for all operational / parallel
runs (the default $VSDB_DATA path) on mmbsrv will
change from /export-6/emc2/data/vsdb (or /disk54/data/vsdb on mmbsrv)
to /export/mmbsrv/usr1/wd20er/data/vsdb (or /raid/usr1/wd20er/data/vsdb on mmbsrv).
Began verification of Alaska WRF nested runs (called WRFA)
Fixed error in WRF model parameter files for GRIDTOBS which caused
incorrect verification of surface fields.
Began verification of CONUS WRF nested runs WRFE (East US), WRFW (West US),
WRFC (Central US)
||Manikin, Benjamin, Rogers
Switched to using RUC2B 00z/12z analyses in verification scripts.
Added 5 SREF Eta Kain-Fritsch runs (designated EKFC, EKN1, EKN2, EKP1, EKP2)
to the FVS database (Turned off on 12/2/02, will be reactivated when KF runs
are turned back on in production)
Fixed script error so that verification files for all HiResWindow forecasts
valid at 21Z are computed correctly
AVN database in the EMC FVS system has been changed to GFS (for Global Forecast System),
the new official name for the NCEP global model forecast:
- Old holding directory : /export-6/emc2/data/vsdb/avn
- New holding directory : /export-6/emc2/data/vsdb/gfs
- Old FVS file name : avn_yyyymmdd.vsdb
- New FVS file name : gfs_yyyymmdd.vsdb
- "AVN" in all the above vsdb files changed to "GFS"
Removed use of ship observations (PREPBUFR ob types 180/280, 182/282) due to problems
with a Great Lakes buoy. This was done for all models with the Great Lakes in their domain.
Switched Eta Alaska surface verification to use the 11.25 km polar stereographic grid
#242 instead of grid #217
Switched HIRESW verification to use NMM runs; new model names :
Changed ops Eta verifications over HIRESW domains (model=ethe (East US), etac
(Central US), ethw (West US)) to use 12-km Lambert Conformal grid #218 instead
of 40 km grid #212)
- Alaska HIRESW : old=ak12 ; new = akhw
- Hawaii HIRESW : old=hi10 ; new = hihw
- Puerto Rico HIRESW : old=pr10 ; new = prhw
Corrected error in PREPFITS : Code set the default GRIB bitmap to 1, so that any model
whose verification grid had a GRIB bitmap=0 would attempt to verify data in the regions
outside of the model integration domain where the bitmap=0
Added verification of surface METAR data (PREPBUFR obs types 187/287) for all models.
Added verification of Eta surface forecasts (10-m wind, 2-m temp/RH) over the CONUS using
the 12-km Lambert Conformal grid #218, and over Alaska using the 22.5 km polar stereographic
Fixed temperature verification problem which observed temperatures were subjected to two
virtual to sensible temperature conversions.