394 
FXCN01 CWAO 290330

OPERATIONAL RUN STATUS
GEM-RGNL IS COMPLETE AND WAS BEHND SCHELDE FOR SOME PRODUCTS BY
5-10MINUTES. ALL OPS RUN ARE ON SCHED THIS EVENING.

GEM MODEL INITIALIZATION FROM 290000 UTC
TRIAL FIELDS AND RUN R1 BOGUS
THE 500MB HGTS WERE TO LO BY 1DAM OVR BC/PRAIRIES. TRIAL IS TOO DEEP WITH NU
UPR LO WITH ITS CENTRE AND FOR ITS ERN PORTION AMD ALSO TOO DEEP BY 1DAM WITH
UPR TROF XTNDG OVR NWT. AMPLITUDE OVR NRN QC UPR RDG WERE TOO HI BY 1DAM.
NO BOGUS WAS PERFORMED.

DATA COVERAGE
RESOLUTE (71924) AND INUVIK (71957) RAOBS ARE MISSING.
OTWZ, NR NORMAL DATA COVERAGE THIS EVENING WITH NR/ABV WKLY AVG NUMBER
OF OBS FOR ALL DATA TYPES.

REJECTED DATA, RUN R1 OA
70200 : PAOM: NOME, AK: 250 AND 500MB WND CORRECTLY RJCTD AS WRONG DIR.
71603 :CYQI:YARMOUTH, NS: 1000MB WND RJCTD AS WRONG DIR AND TOO STRONG. OBS
LUKS OK.
ALL OTHER REJECTS NOT SIGNT.

R1 OA HAS CORRECTED A BIT WKNESS NOTED WITH TRIAL FIELDS..BY LESS THAN 1DAM.
OA HAS ALSO WEAKEN ATLC UPR LO BY 1DAM. OTWZ, ANAL INCREMENTS ARE QUITE SMALL
THRUT NOAM.



GEM PERFORMANCE FOR 24 HR PD ENDING 290000Z

SHORT RANGE VRFCTN SHOWS THAT MAIN FCST ERROR WAS OVR WRN AB WHERE MODEL WAS
TOO INTENSE WITH VORT MAX, LOWERING HEIGHTS TOO MUCH; AS A RESULT, ERROR UP TO
4 DAMS NOTED JUST E OF BC/AB BORDER. WITH HIR AMPLITUDE OF UPPER RDG
THRU BC/AB, DOWNSTREAM IMPULSE MOVD FURTHER S THRU SRN PRAIRIES GVG ERRORS OF
3-4DAMS IN THIS AREA.
OVR ARCTIC RGNS, MODEL WAS QUITE GUD WITH VERTICAL FEAT AND MINOR WKNSS NOTED
IN THIS AREA.
OVR ERN RGNS, UPPER TROF THRU NERN ONT IS NOT QUITE AS DEEP AS FCST BY ABT 3
DAMS WHILE SRN PART OF THIR UPPER TROF THRU LOWER GREAT LKS WAS NOT FCST DEEP
EOUGH BY ABT 1 DAM.
CLOSED LO S OF MARITIMES WAS FCST 2 DAMS TOO DEEP. ASSOCIATED SFC LO WAS ALSO
TOO DEEP BY ABT 4 MBS.

PCPNWISE, MODEL WAS QUITE GUD WITH AREAS OF PCPN THRUT CDN RGNS; GEMREG QPFS
WAS ON THE HI SIDE THO WITH MAX AMOUNTS.
OVR WRN RGNS, PCPN PATTERN WAS VERY WELL FCST BY GEMREG LAST 24HRS; MODEL WAS
ESPECIALLY GUD WITH AXIS OF MAX PCPN ON WRN SIDE OF VRISLD AND W COAST. MODEL
WAS ALSO FAIRLY GUD WITH MAX AMOUNTS XCPT OVR CSTL SXNS OF CNTRL COAST RGN
WHERE MAX AMOUNTS LUK ON THE HI SIDE BASED ON AVLBL SFC REPORTS.
OVR ONT RGNS, MODEL DID A GUD JOB ALSO WITN PCPN PATTERN; HWVR, MODEL LUKS
ON THE HI SIDE WITH MAX AMOUNTS THRU AXIS OF HIR PCPN N OF LK SUP.
OVR ARCTIC RGNS, VRFCTN OF PCPN ENVELOPE IS GUD BUT MAX AMOUNTS NOT SUPPORTED
BY SFC REPORTS.
OVR MARITIMES, GEMREG IS A BIT TOO FAR N WITH LEADING EDGE OF PCPN.


GEM 12 AND 24 HR PROG

MAIN ADJUSTMENT NOTED COMPARED TO PREVIOUS INTEGRATION IS OVR CNTRL RGNS WHERE
MODEL IS NOW DEEPER WITH LEADING IMPULSE MOVG FROM SWRN SK TO LK SUP BY
T+24HRS; UPSTREAM IMPULSE OVR WRN AB, ON ERN SIDE OF UPPER RDG IS NOW NOT QUITE
AS DEEP AS PREVIOUSLY FCST AS IT MOVS THRU CNTRL RGNS,
THESE ADJUSTMENTS ARE JUSTIFIED AND ACCEPTED AS MODEL WAS TOO DEEP WITH IMPULSE
THRU SWRN AB AND NOT DEEP ENOUGH WITH LEADING ONE. INITIALIZATION OF THESE
IMPULSES LUKS FINE TONIGHT.
AS A RESULT, ASSOCIATED FNTL TROF MOVG THRU SRN PRAIRIES IS A BIT DEEPER THAN
SUGGESTED BY PREVIOUS INTEGRATIONS AND THIS IS ALSO ACCEPTED.
QPFS PATTERN ASSOCIATED WITH THIS SYSTEM DIFFERS ALSO A BIT FROM PREVIOUS RUNS
WITH MORE COMPACT PCPN PATTERN THRU SRN SK/MB NEXT 12HRS; THIS PATTERN FITS
BETTER WITH CLOUD PATTERN ON IMAGERY AS WELL AS RADAR ECHOES THRU THIS AREA.
HWVR, MODEL QPFS MAY BE A TAD HI WITH MAX AMOUNTS ASSOCIATED WITH THIS SYSTEM
ERLY IN PD IN VIEW OF INTENSITIES OF RADAR ECHOES. FOR 12-24HRS PD, GEMREG
QPFS LUK MORE RSNBL WITH LOWER AMOUNTS THRU NWRN ONT COMPARED TO PREVIOUS RUN
AND IS ACCEPTED.
GEMREG NOT INTENSE ENOUGH WITH CVCTV ACTIVITY OVR MIDDLE MISS VALLEY ERLY IN
PD; HWVR, THIS ACTIVITY WILL REMAIN S OF OUR RGNS AS IT DRFT EWD S OF LOWER
GREAT LKS.

OVR WRN RGNS, MODEL FAIRLY CONSISTENT WITH PREVIOUS INTEGRATION HANDLING
EVOLUTION OF FNTL TROF OFF W COAST MOVG RAPIDLY INLAND THRU SRN BC NEXT 24HRS.
QPFS PATTERN ALSO FAIRLY SIM; IN VIEW OF CONSISTENCY AND GUD VFRCTN LATELY IN
THIS AREA, GEMREG QPFS GNLY ACCEPTED FOR THIS SYSTEM XCPT IN WAKE OF IT WHERE
PCPN AMOUNTS PROBABLY ON THE HI SIDE AGAIN OVR SWRN BC IN ONSHR/UPSLP CONDS.

OVR ARCTIC RGNS, MODEL ALSO FAIRLY CONSISTENT FROM RUN TO RUN HANDLING
EVOLUTION OF CLOSED LO CNTR DRFTG SLOLY NNWWD.
TIGHT CYCLONIC FLO ASSOCIATED WITH THIS SFC FEAT WILL CONTINUE TO GV MDT-STRNG
WNDS TO MOST OF CNTRL/ERN ARCTIC RGNS. CMC-OBJ SVR WX CHARTS STILL RSNBL WITH
OVERVIEW OF PARTS OF ERN ARCTIC RGNS WITH STRONG WNDS.
FOR PCPN, BASED ON RECENT VFRCTNS, MAX AMOUNTS PROBABLY ON THE HI SIDE AGAIN,
BUT GNRL SOLN LUKS RSNBL.

OVR ATLC PROVINCES, MODEL NOT QUITE AS DEEP WITH VERTICAL LO LIFTG NWD NEXT
24HRS. THIS SLIGHTLY WKR SOLN IS CORRECT AS MODEL VRFD TOO DEEP WITH THIS
FEAT.
ASSOCIATED PCPN SPREADING THRU ATLC PROVINCES NEXT 24HRS; GEMREG QPFS PATTERN
LUKS MORE RSNBL WITH LOWER AMOUNTS IN VIEW OF CLOUD PATTERN ON IMAGERY BUT
CONFIDENCE STILL LO ON DETAILS OF PCPN PATTERN AFECTING MARITIMES NEXT 24HRS.
END/LAROCQUEM/ROY