Log for Pacific Northwest MM5 Numerical Forecast Verification

Verification Log Navigation

[ Model and Verification Problems | Observation Problems | Change Log | Realtime Verification Statistics ]


Model Problems


Problems with verify data due to model failures, lack of data, obs files not being created on time, the deletion of the mmout files before the errors were detected, etc.: 2003072200: MM5-ETA and MM5-GFS did not use ETA 221 grids to initialize surface parameters. 2003072112: MM5-ETA and MM5-GFS did not use ETA 221 grids to initialize surface parameters. 2003072012: MM5-ETA and MM5-GFS did not use ETA 221 grids to initialize surface parameters. 2003072000: MM5-ETA and MM5-GFS did not use ETA 221 grids to initialize surface parameters. 2003071912: MM5-GFS 36 and 12-km run ended at hour 63 in real-time due to lack of GFS grids. Run completed to 72 hours on July 28. 2003070200: MM5-ETA and MM5-GFS did not use ETA 221 grids to initialize surface parameters. 2003062712: MM5-GFS 4-km run ends at hour 36. wrf15 crashed. 2003061112: MM5-GFS and MM5-ETA did not use ETA 221 grids. 2003050912: MM5-GFS 36/12 km initially halted at hour 63 and 4-km delayed 2 hours. Runs completed late. 2003041612; MM5-GFS delayed 90 minutes due to problems with /home/disk/data. 2003032912: MM5-GFS did not run in real-time due to continuing /home/disk/data problems. 2003032812: MM5-ETA delayed due to /home/didk/data problem. 2003032800: MM5-GFS and MM5-ETA delayed due to /home/didk/data problem. 2003030500: MM5-GFS runs delayed due to problems stopping past cases. 2003030412: All runs delayed due to problems stopping past cases. 2003030400: All runs delayed due to problems with LDM corrupting incoming GRIB files. ETA 104 grids corrupted and MM5-ETA not run. 2003030312, 2003030300, 2003030212: All 3 MM5 runs did not complete in real-time due to memory SIMM failure on tahoma. 2003012712, 2003012700: MM5 runs delayed or did not run in real-time due to problems retrieving ETA 221 grids for initial conditions. 2003012512, 2003012500: MM5 runs did not run in real-time due to network outage at U.W. caused by hackers on 175 subnet. 2003011800: MM5-GFS delayed several hours due to MPI bug. 2003011612: MM5-GFS stopped at hour 69 in real-time due to late arrival of 72-hour GFS forecast. Run completed early the next day. 2003011512: MM5-GFS timed out waiting for hour 66 and needed manual preprocessing intervention after grid arrived late. 2002122600: MM5-GFS stopped at hour 21 due to a crash of wrf10. Run completed on Dec 30th. 2002121512: MM5-GFS ran only to hour 24 in real-time due to problems acquiring input grids. 2002112512: MM5-GFS delayed 5.5 hours due to wrf1 automount failure. 2002112500: MM5-GFS did not run in real-time due to wrf1 automount failure. 2002112400: MM5-GFS delayed due to wrf16 automount failure. 2002112212: MM5-GFS delayed 4.5 hours due to wrf1 automount failure. 2002111312: MM5 delayed at hour 40 due to crash of wrf1. 2002110812: MM5 delayed at hour 47 due to crash of wrf5. 2002110700: MM5 delayed at hour 42 due to crash of wrf5. 2002071200- 2002082712: No ETA 221 grids were used to initialize 3-D data in MM5 runs. The snow cover and soil temperatures, however, were used. 2002072512: NCEP backup test meant no ETA or AVN runs. MM5 was not run at all. 2002071612: Boundary condition problems caused delays in real-time. MM5 did not use ETA 221 grids in initialization. 2002071500: Same problems with our cloud-water initialization as the 2002071100 case. MM5 did not run in real-time. Archive version does not include cloud-water initialization, but does use ETA 221 grid initialization. 2002071100: Our cloud-water initialization scheme caused CFL errors in the Kain-Fritsch scheme and the model crashed. The run was delayed approximately 6 hours before a run initializaed without cloud-water and without ETA 221 grids could be started. 2002071000: MM5 did not use ETA 221 grids in initialization. 2002070712: MM5 did not use ETA 221 grids in initialization. 2002070112: MM5 did not use ETA 221 grids in initialization. 2002062812: MM5 did not use ETA 221 grids in initialization. 2002062500: MM5 did not run in real-time due to CFL errors over a convective line in New Mexico. The ETA 221 grids were the source of the problem. The initialization was switched to the ETA 104 grids and the forecast then ran to completion on June 27th. 2002061612: MM5 did not use ETA 221 grids in initialization. 2002061600: MM5 did not use ETA 221 grids in initialization. 2002052412: Real-time MM5 initially ran with AVN grids to hour 36. Subsequently rerun with ETA 104 grids. ETA-initialization was used for verification. 2002041712: Real-time MM5 did not use boundary conditions for hours 27 and 30 since they were posted extremely late at NCEP. 2002040612: Problems with receiving ETA 104 hour 51 file delayed real-time. ETA verification only goes to hour 48. 2002040112: MM5 ran 3 hours late due to problems at NWS ftp server in posting ETA 104 grids. 2002032700: MM5 4km not run in real-time due to problems at UW. 2002032612: MM5 did not use ETA 221 grids in initialization due to problems on NCEP ftp server. 2002022700: MM5 36 and 12 km runs stop at hour 39 due to lack of ETA 104 grids beyond that hour. 2002011512: All MM5 forecasts were delayed by 8 hours or more due to department computer (sunny) outage. 2002011500: MM5 4km domain graphics for hours 24-36 were delayed in posting to Web due to department computer outage. 2002010912: MM5 ran 3 hours late and did not use ETA 221 grids for initialization due to hanging of department LDM machine and main data disk. 2002010700: MM5 36/12 km ran only to hour 39 due to problems at NCEP. 2002010312: MM5 delayed 30 minutes due to tahoma crash/reboot due to parity error on CPU 8. 2001123112: MM5 delayed 3 hours 45 minutes due to rainier crash overnight and subsequent job backlogs on glacier. 2001123100: MM5 delayed half day for 36/12 km domains and 1.5 days for 4 km domain due to rainier crash overnight. 2001121812: Original run missed boundary conditions between hours 18 and 42. MMOUT files stored in 2001121812.rerun directory. Web graphics updated and verification rerun with new data. ~mm5rt/nobackup/2001121812.compare/ has some plots and verification files. 2001110312, 2001110300: MM5 ran late due a bug in our ETA 104 grid retrieval. No ETA verification or grids are available for these dates either since the grids were never placed in the proper directories. 2001102800: MM5 used AVN grids for boundary conditions, but ETA 221 grids as usual. OSO server problems meant only hour 0 was received from ETA. 2001102400: MM5 did not use ETA 221 grids in initialization, they arrived 2 minutes after the 15-minute wait expired. 2001102300: MM5 ran about 3 hours late due to previously undetected bug in pregrid.pl. 2001101712: MM5 initialized with AVN grids due to failure at NCEP. Output delayed approximately 3.5 hours. 2001101600: MM5 did not run in real-time due to anomalously small ETA 104 and 221 GRIB files for hour 0. 2001101012: NCEP backup test meant no ETA 104 or 221 grids received. MM5 initialized with 0Z run of AVN. 2001092600: MM5 4 km halted at forecast hour 34 due to tahoma memory cache error on CPU 11. Final 2 hours of forecast were run the next day since first restart file was corrupt. 2001070912: New jun01 domain. 2001073012: Web server was down from 12:03 pm to 7:51 pm. MM5 unaffected, but graphics from this run and prior runs were unavailable from Web until mid morning on the 31st. 2001072512: MM5 did not use ETA 221 grids in initialization because rainier was hung. 2001071900: MM5 runs delayed 3.5 hours due to rainier computer hanging. 2001071700: MM5 32 and 12 km domains ended at hour 48 due to a bug in transferring the boundary files. 2001071300: MM5 4 km verification missing. 2001071212: MM5 4 km delayed one hour due to tahoma CPU 1 board memory parity error. 2001061912: MM5 ran with NOGAPS initialization due to NCEP backup test, no verification statistics retained for this run. (Moved to verify/altinitmodel/nogaps). 2001052712, 2001052700, 2001052612: No MM5 forecasts due to tahoma crash and full disks. 2001052400: MM5 did not use ETA 221 grids in initialization. 2001052312: MM5 did not use ETA 221 grids in initialization. 2001052300: MM5 did not use ETA 221 grids in initialization. 2001051900: MM5 did not use ETA 221 grids in initialization. 2001040412: MM5 did not use ETA 221 grids in initialization. 2001031700: No verification statistics produced. 2001031500: No ETA verification statistics produced. 2001031200: MM5 did not use ETA 221 grids in initialization. 2001020612: MM5 did not use ETA 221 grids in initialization and model ran about 2 hours 45 minutes late due to NCEP backup operations test. 2001020112: MM5 did not use ETA 221 grids in initialization. 2001012300: MM5 ran 90 minutes late due to complications arising from switchover to new RAID disks and network problems. 2001012212: MM5 ran 4 hours late due to complications with switchover to new RAID disks. 2001012200: MM5 did not run in real-time due to switchover to new RAID disks, but finished on the 23rd. 2001011500: MM5 did not use ETA 221 grids in initialization. 2001011212: MM5 did not use ETA 221 grids in initialization. 2000122300: MM5 did not use ETA 221 grids in initialization. 2000121312: MM5 did not use ETA 221 grids in initialization. 2000120512: NCEP errors caused 2 1/2 hour delay. MM5 did not use ETA 221 grids in initialization. 2000120112: MM5 did not use ETA 221 grids in initialization. 2000113012: MM5 did not use ETA 221 grids in initialization. 2000112912: MM5 did not use ETA 221 grids in initialization. 2000112812: MM5 did not use ETA 221 grids in initialization. 2000111300 to 2000112800: MM5 forecasts all used SST data from 2000111112. 2000112512: MM5 did not use ETA 221 grids in initialization. 2000112312: MM5 did not use ETA 221 grids in initialization. 2000112012: MM5 did not use ETA 221 grids in initialization. 2000112000: MM5 d1 and d2 forecasts ended at hour 57 and 58 due to tahoma's disks filling up after a tahoma crash. MM5 4 km was 3 1/2 hours late. 2000111612: MM5 did not use ETA 221 grids in initialization. 2000111500: MM5 did not use ETA 221 grids in initialization. 2000110312: MM5 did not use ETA 221 grids in initialization. 2000110300: NCEP power outage caused large delays in MM5 startup. Real-time version ran with NOGAPS initialization and ended at hour 48 for 36/12 domains and at hour 29 for domain 3. Case was rerun later using ETA 104 grids for verification. MM5 did not use ETA 221 grids in initialization. 2000110112: MM5 did not use ETA 221 grids in initialization. 2000102712: MM5 did not use ETA 221 grids in initialization. 2000102612: MM5 did not use ETA 221 grids in initialization. 2000102012: MM5 did not use ETA 221 grids in initialization. 2000101812: MM5 did not use ETA 221 grids in initialization. 2000101612: MM5 did not use ETA 221 grids in initialization. 2000101212, 2000101200: MM5 4 km domain forecast hours 29 through 36 finished late. 2000100812: MM5 4 km domain had to be finished late. 2000100800: MM5 4 km domain had to be finished late and ended at hour 35. 2000100712: MM5 finished 36 and 12 km domains in real-time only up to hour 48. Hours 48 to 60 and the 4 km domain were finished later because of large problems acquiring the ETA 104 grids. MM5 did not use ETA 221 grids in initialization. MM5 verification for 36 and 12 km domains ends at hour 50, and, for the 4 km domain, ends at hour 29. 2000100600: MM5 did not use ETA 221 grids in initialization. 2000100512: MM5 did not use ETA 221 grids in initialization. 2000100412: MM5 did not use ETA 221 grids in initialization. 2000100312: MM5 did not use ETA 221 grids in initialization. 2000100212: MM5 did not use ETA 221 grids in initialization. 2000100200: MM5 did not use ETA 221 grids in initialization. 2000100112: MM5 did not use ETA 221 grids in initialization. 2000100100: MM5 did not use ETA 221 grids in initialization. 2000093012: MM5 did not use ETA 221 grids in initialization. 2000093000: MM5 did not use ETA 221 grids in initialization. 2000092912: MM5 did not use ETA 221 grids in initialization. 2000092900: MM5 did not use ETA 221 grids in initialization. 2000092900: MM5 did not use ETA 221 grids in initialization. 2000092912: MM5 did not use ETA 221 grids in initialization as slowness at ftp.ncep.noaa.gov continues. 2000092900: MM5 did not use ETA 221 grids in initialization. 2000092812: MM5 did not use ETA 221 grids in initialization. 2000092800: MM5 run one day late as problems with ETA 221 grid retrieval caused problems with deep soil initializations. 2000092712: MM5 did not use ETA 221 grids in initialization. 2000092700: MM5 did not use ETA 221 grids in initialization. 2000092600: MM5 did not use ETA 221 grids in initialization. 2000092400: MM5 did not use ETA 221 grids in initialization. 2000092212: MM5 did not finish due to lack of ETA 104 grids (serious problems receiving data from NCEP). MM5 did not use ETA 221 grids in initialization. 2000092200: MM5 did not use ETA 221 grids in initialization. 2000092112: MM5 initialized with AVN due to serious problems with ftp acquisition of ETA 104 grids from OSO server. 2000092012: MM5 did not use ETA 221 grids in initialization. 2000092000: MM5 did not use ETA 221 grids in initialization. 2000091912: MM5 did not use ETA 221 grids in initialization. 2000091900: MM5 did not use ETA 221 grids in initialization due to ftp.ncep.noaa.gov not being accessible via Internet2. 2000091812: MM5 had to be restarted at hour 36 since data was lost due to a full disk drive. Verification ends at hour 28 for domain 2. 2000091400: MM5 did not use ETA 221 grids in initialization. 2000091212: MM5 did not use ETA 221 grids in initialization. 2000090912: MM5 4-km simulation did not run in real-time due to failure to save forecast hour 16.5 from the 12-km run. 2000090900: MM5 ran late with AVN initialization and only ran to hour 18 for the 4 km domain due to unknown problem in generating initial conditions from ETA 104 grids. 2000082812: MM5 did not use ETA 221 grids in initialization. 2000082800: MM5 did not use ETA 221 grids in initialization. 2000082512: MM5 did not use ETA 221 grids in initialization. 2000082500: MM5 did not use ETA 221 grids in initialization. 2000082412: MM5 did not use ETA 221 grids in initialization. 2000082400: MM5 did not use ETA 221 grids in initialization. 2000082312: MM5 did not use ETA 221 grids in initialization. 2000082300: MM5 ran 3.5 hours late due to deparment computer changes. MM5 did not use ETA 221 grids in initialization. 2000081512: MM5 did not use ETA 221 grids in initialization. 2000081500: MM5 did not use ETA 221 grids in initialization. 2000081412: MM5 did not use ETA 221 grids in initialization. 2000080912: MM5 did not use ETA 221 grids in initialization. 2000080112: NCEP Internet outage caused us to start at 2 pm. No ETA 221 grids used in initialization. Web server down much of day. 2000073112: New jul00 domain. 2000072600: MM5 did not use ETA 221 grids in initialization. 2000072212, 2000072200: MM5 runs did not finish in real-time mode. Lightning caused a system failure on operational computer. No ETA 221 grids for 2000072200 run. 2000070512: MM5 did not use ETA 221 grids in initialization. 2000062812: MM5 computer rebooted at forecast hour 28 due to a memory parity error on CPU 17. 2000062712: MM5 did not use ETA 221 grids in initialization. 2000062612: MM5 did not use ETA 221 grids due to 3 hour 40 minute delay in ETA run at NCEP. 2000061912: MM5 finished very late due to full disks and need to rerun the complete forecast. 2000061512: MM5 did not use ETA 221 grids. 2000051512: MM5 did not use ETA 221 grids. 2000051212: ETA verification is missing. 2000050712: MM5 did not use ETA 221 grids. 2000050612: MM5 4 km run did not finish past hour 12 in real-time. 2000042912: MM5 did not use ETA 221 grids. MM5 4km run was 3 hours late due to damp's disk crash that caused rainier to hang. 2000041912: MM5 did not use ETA 221 grids. 2000041900: MM5 missing 4 km verification stats for hours 23 through 27. 2000041512, 2000041500: MM5 4 km domain run was significantly delayed due a bug in PERL. PERL 5.6.0 fixed this bug. 2000032812: MM5 did not use ETA 221 grids. 2000031812: Missing forecast hour 6 in MM5 d3 verification. 2000031712: MM5 did not use ETA 221 grids. 2000031700: MM5 did not use ETA 221 grids. 2000031000: MM5 did not use ETA 221 grids. 2000030800: MM5 did not use ETA 221 grids. 2000022912: MM5 did not use ETA 221 grids or cloud-water data in initialization. Verification of ETA missing hours 0-12. 2000022900: MM5 did not run in real-time due to Y2K leap year bugs in GEMPAK. MM5 did not use ETA 221 grids in initialization. Verification of ETA missing hours 0-24. 2000020900: MM5 did not use ETA 221 grids in initialization, ETA 104 was retrieved from very slowly from NCEP, but much faster from Joe Tenerelli. 2000020812: MM5 initialized late with AVN grids because NCEP comm.s problems meant ETA 104 did not arrive. 2000013100: MM5 forecasts end at hour 39 due to NCEP ETA problems. 2000012400: MM5 did not use ETA 221 grids in initialization. 2000011600: MM5 did not use ETA 221 grids in initialization. 2000010712: MM5 did not use ETA 221 grids in initialization. 2000010512: MM5 forecast run 4 late. 2000010500: MM5 forecast run 16 hours late. 2000010200: MM5 did not use ETA 221 grids in initialization. 2000010100: MM5 did not use ETA 221 grids in initialization. 1999122800: MM5 did not use ETA 221 grids in initialization. 1999122512: MM5 did not use ETA 221 grids in initialization. 1999122212: MM5 did not use ETA 221 grids in initialization. 1999121012: MM5 did not use ETA 221 grids in initialization. 1999121000: MM5 domain 2 verification is missing forecast hours: 15, 18, 21, 24, 27, 30, 33, 36, 39, 42, 45, 48. MM5 did not use ETA 221 grids in initialization. 1999120412: MM5 did not run until the 6th due to a tahoma crash. MM5 did not use ETA 221 grids in initialization. 1999120312: MM5 did not use ETA 221 grids in initialization. 1999120300: MM5 did not use ETA 221 grids in initialization. 1999120200: MM5 did not use ETA 221 grids in initialization. 1999120112: MM5 did not use ETA 221 grids in initialization. 1999120100: MM5 did not use ETA 221 grids in initialization. 1999113012: MM5 did not use ETA 221 grids in initialization. 1999113000: MM5 did not use ETA 221 grids in initialization. 1999112912: MM5 did not use ETA 221 grids in initialization. 1999112900: MM5 did not use ETA 221 grids in initialization. 1999112700: MM5 did not use ETA 221 grids in initialization. 1999112612: MM5 did not use ETA 221 grids in initialization. 1999112412: MM5 did not use ETA 221 grids in initialization. 1999112400: MM5 did not use ETA 221 grids in initialization. 1999112312: MM5 did not use ETA 221 grids in initialization. 1999112300: MM5 did not use ETA 221 grids in initialization. 1999112212: MM5 did not use ETA 221 grids in initialization. 1999112200: MM5 4 km run had to be run late due to problems running backend interp on blizzard. 1999112112: MM5 4 km run had to be run late due to problems running backend interp on blizzard. MM5 did not use ETA 221 grids in initialization. 1999112100: MM5 did not use ETA 221 grids in initialization. 1999112012: MM5 did not use ETA 221 grids in initialization. 1999111912: Forecast hour 18 was corrupted for domain 1, not included in verification statistics. MM5 did not use ETA 221 grids in initialization. 1999111812: MM5 did not use ETA 221 grids in initialization. 1999111700: Missing ETA verification due to missing fields in grids we received. MM5 did not use ETA 221 grids in initialization. 1999111500: Missing ETA verification due to missing fields in grids we received. 1999111012: MM5 did not use ETA 221 grids in initialization. 1999111000: MM5 was initialized with AVN since ETA 104 was not posted to OSO server. 1999110912: MM5 did not use ETA 221 grids in initialization. 1999110812: MM5 did not use ETA 221 grids in initialization. 1999110800: MM5 did not use ETA 221 grids in initialization. 1999110712: MM5 did not use ETA 221 grids in initialization. 1999110512: MM5 did not use ETA 221 grids in initialization. 1999110500: MM5 did not use ETA 221 grids in initialization. 1999102200: MM5 did not use ETA 221 grids in initialization. 1999101800: MM5 did not use ETA 221 grids in initialization. 1999101612: MM5 did not use ETA 221 grids in initialization. 1999101300: MM5 did not use ETA 221 grids in initialization. 1999100712: MM5 did not use ETA 221 grids in initialization. 1999100612: MM5 forecast in real-time ended at hour 18 for all domains due to computer errors at UW interfering with data retrieval. Model rerun later with mostly 6-hour boundary conditions out to hour 48. 1999100600: MM5 did not use ETA 221 grids in initialization. 1999100100: MM5 did not use ETA 221 grids in initialization. Missing ETA verification files. 1999093012: MM5 initialized with AVN not ETA since ETA did not arrive. 1999093000: MM5 did not use ETA 221 grids in initialization. Missing ETA verification files. 1999092912: MM5 initialized with AVN not ETA since ETA arrived 4+ hours late and incompletely. 1999092900: MM5 did not use ETA 221 grids in initialization. 1999092812: MM5 initialized with AVN not ETA due to fire at NCEP. 1999092800: MM5 initialized with AVN not ETA due to fire at NCEP. 1999092712: MM5 did not use ETA 221 grids in initialization. 1999092700: MM5 did not use ETA 221 grids in initialization. 1999092612: MM5 did not use ETA 221 grids in initialization. Eta verification ends at hour 42. 1999092600: MM5 did not use ETA 221 grids in initialization. 1999092512: MM5 did not use ETA 221 grids in initialization. 1999092500: Forecast used ETA 104 initialization to hour 36 only, and AVN thereafter. ETA verification ends at hour 36. MM5 did not use ETA 221 grids in initialization. 1999092400: MM5 did not use ETA 221 grids in initialization. 1999092312: MM5 did not use ETA 221 grids in initialization. 1999092112: MM5 did not use ETA 221 grids in initialization. NCEP problems caused 3+ hour delay. 1999092012: MM5 did not use ETA 221 grids in initialization. 1999091912: MM5 did not use ETA 221 grids in initialization. 1999091512: MM5 was initialized late with AVN due problems to receiving ETA in a timely manner. Also, tahoma got hung (tapes were being made) and was rebooted. 1999091500: MM5 did not use ETA 221 grids in initialization. 1999091412: MM5 did not use ETA 221 grids in initialization. Also forecast ends at hour 45. 1999091400: MM5 did not use ETA 221 grids in initialization. 1999091312: MM5 did not use ETA 221 grids in initialization. 1999091300: MM5 did not use ETA 221 grids in initialization. 1999091212: MM5 did not use ETA 221 grids in initialization. 1999090612: MM5 did not use ETA 221 grids in initialization. 1999090112: MM5 did not use ETA 221 grids in initialization. 1999083100: MM5 did not use ETA 221 grids in initialization. 1999083000: MM5 initialized late with AVN due to NCEP problems with ETA. 1999082912: MM5 initialized late with AVN due to NCEP problems with ETA. 1999082900: MM5 initialized late with AVN due to NCEP problems with ETA. 1999082812: MM5 did not use ETA 221 grids in initialization. 1999082800: MM5 did not use ETA 221 grids in initialization. 1999082612: MM5 did not use ETA 221 grids in initialization. 1999082400: MM5 did not use ETA 221 grids in initialization. 1999082000: MM5 did not use ETA 221 grids in initialization. 1999081912: MM5 did not use ETA 221 grids in initialization. 1999081412: Major system failure at NCEP meant no ETA forecast. MM5 run was late and used coarser AVN grids. 1999081700: MM5 did not use ETA 221 grids in initialization. 1999080100: Missing ETA verification. 1999072800: No MM5 run because NCEP did not run ETA (NCEP Cray 5 crashes). Model was run after 1999072812 using AVN 6-hour analyses. 1999072700: MM5 did not use ETA 221 grids in initialization. 1999072612: MM5 did not use ETA 221 grids in initialization. 1999072012: MM5 did not use ETA 221 grids in initialization due to 2:50 delay in receiving all NCEP grids. 1999071900: MM5 did not use ETA 221 grids in initialization. 1999070712: MM5 did not use ETA 221 grids in initialization. 1999070612: MM5 did not use ETA 221 grids in initialization. 1999070512: MM5 did not use ETA 221 grids in initialization. 1999070300: No MM5 forecast since never got complete ETA 00 hour fields. 1999070512: MM5 did not use ETA 221 grids in initialization. 1999070200, 1999070112, and 1999070100: No ETA 221 grids used in MM5 outputs used in verification. Due to a late rerun for verification, 221 grids were aged off. 1999063000: Verification files missing for all domains. 1999062812: MM5 did not use ETA 221 grids in initialization. Missing verification files for MM5 4 km domain. 1999062800: Missing MM5 verification files. 1999062712: Missing verification files for MM5 4 km domain. 1999062700: Missing MM5 verification files. 1999062612, 1999062600, 1999062512, 1999062500: Missing MM5 4 km verification files. 1999062400: Tahoma crashed (tapes were being made) model was late. 1999062312: Missing all MM5 verification files. 1999062112: MM5 did not use ETA 221 grids in initialization. 1999061700: MM5 did not use ETA 221 grids in initialization. 1999061600: MM5 did not use ETA 221 grids in initialization. 1999061512: MM5 did not use ETA 221 grids in initialization. 1999061212: Tahoma crashed, model was late and did not use ETA 221 grids in initialization. 1999061200: Cloud-water initialization corrected and turned on. 1999060912: Due to tahoma crash, did not use ETA 221 grids in initialization. 1999060900: Tahoma crashed, model run completed after 1999060912 run. 1999060812: Due to tahoma crash, MM5 domain 3 rerun at later date with cloud water initialization turned on for whole run.. 1999052812: Began using ETA 221 (36-km) gridded data for 0-hour initialization. 1999052612: Missing forecast hour 30 for domains 1 and 2 also missing hours 12, 27, 33, and 36 for d2. 1999052412: NCEP had serious computer problems. We received ETA104 grids only out to hour 18. MM5 eventually run with NOGAPS. 1999052200: Model crashed at hour 8 due to a CFL error. 1999052100: New 40-day running mean soil temperature initialization scheme turned on. 1999051812: Switched to new May99 model setup. Soil temperature initialization was not initially turned on, and because of the new incremental Interp, soil temperatures for the model run would be based on just the first 3 hours of ETA forecasts. 1999050312: Missing ETA verification files. 1999032312: MM5 forecasts end at hour 30 due to a very late start. 1999032100: No MM5 forecasts. 1999032012: No MM5 forecasts because ETA grids never arrived. 1999031712: MM5 forecasts ended at hour 42 due to incomplete boundary conditions from a very slowly arriving ETA. 1999031212: MM5 was initialized with NOGAPS 6-hourly data for forecasts hours 12 to 48. ETA was delayed at NCEP and hours 39 and 42 were not available. ETA verification is missing hour 39 since it never arrive: hour 42 arrived very late. 1999030412, 1999030500, 1999030512: Data decoding problems at the department meant that many observations were missing from 1999030606 through 1999030623, therefore, many obs missing from these verification files. 1999022812: MM5 did not run on time -- reason unknown -- but it was re-run later for verification purposes. Output to 48 hours with forecast hours 14 through 23 missing. Missing 4 km verification files. 1999022800: MM5 forecasts end at hour 44 due to bug in OS?. 1999022700: MM5 did not run on time -- reason unknown -- but, re-run later for verification purposes. 1999022612: MM5 forecasts stopped at hour 44 due to bug in OS? 1999022500: MM5 had to be initialized with 6-hourly ETA data for the 12-48 hour forecasts due to slow NCEP grid transmission. ETA verification has 3-hourly output to hour 18, 6-hourly thereafter. 1999022312: MM5 had to be initialized with 6-hourly ETA data for the 12-48 hour forecasts due to slow NCEP grid transmission. 1999022212: MM5 had to be initialized with 6-hourly ETA data due to NCEP computer problems. 1999022100: MM5 ends at hour 42 due to grid acquisition problems and late start. 1999020412: MM5 stopped at hour 36 due to incredibly slow transfer of ETA grids from NCEP. 1999020312: MM5 initialized with NOGAPS due to massive amounts of NCAR data missing and NCEP power outages that prevented the ETA from arriving on time. ETA verification ends at hour 33 (all that was available). Also, numerous obs missing between 99020308 and 99020315. This primarily affected the 99020112 verification, 99020200 and 99020212 were caught and re-run in time. 1999013100: MM5 crashed at hour 28 due to a power outage at the University. 1999012100: No MM5 forecast due to a tahoma crash caused by its inability to lock the processors. 1998123000: MM5 forecasts end at hour 39 (late start). 1998122200: MM5 forecasts end at hour 20 due to a late start. 1998121800: MM5 d2 missing hours 12, 24 and 36. 1998112500: MM5 forecasts end at forecast hour 42 due to a late start caused by late arriving NCEP grids. 1998102712 through 199811111: Every ETA verification file is missing hours that end at 00Z. 1999102700 through 1999080703: Missing ETA verification files. 1998100500: MM5 forecasts did not run due to campus power outage. 1998100400: MM5 forecasts did not run due to software problems. 1998092600: MM5 forecasts did not run due to software problems. 1998091800: MM5 forecasts end at hour 12 due to tahoma crash. 1998090800: MM5 forecasts end at hour 12 because the ETA grids from NCEP never fully arrived. 1998090700: MM5 forecasts not run due to tahoma problems. 1998090600: MM5 forecasts not run due to tahoma problems. 1998090512: MM5 forecasts end at hour 12. 1998090100: MM5 forecasts not run because ETA did not run at NCEP. 1998082412: MM5 forecasts end at hour 27 due to NCEP data server problems. 1998081712: MM5 forecasts end at hour 23 due to system error on tahoma. 1998080700: No MM5 forecast run, due to programming error. 1998080512: MM5 forecasts end at hour 35 due to local problems. 1998073000: MM5 forecasts end at hour 30 due to NCEP problems. 1998072200: MM5 forecasts end at hour 20 due to NCEP problems. 1998071700: MM5 forecasts end at hour 33 1998062200: MM5 did not run due to tahoma crash. 1998061900: MM5 did not run due to building construction work and NCEP problems. 1998061800: MM5 did not run due to computer crash and NCEP problems. 1998061700: MM5 forecasts end at hour 12 due to NCEP problems. 1998061100: MM5 did not run due to programming error. 1998061000: MM5 did not run due to programming error. 1998060800: MM5 verification and forecasts end at hour 12 (disk was full). 1998060712: MM5 verification and forecasts end at hour 13 (disk was full). 1998060512: MM5 verification and forecasts end at hour 38 (disk was full). 1998060500: MM5 verification unavailable (mmout files unavailable at time of verification). 1998060412: MM5 verification and forecasts end at hour 43 for 36-km domain, and hour 9 for 12-km domain. 1998060312: MM5 forecast verification ends at hour 24. 1998060300: ETA verification begins at hour 18. 1998060212: No verification statistics available. 1998060200: Verification missing hours: ETA 36, d2 27, d1 24. 1998060112: ETA verification ends at hour 9. 1998060100: MM5 forecasts end at hour 36, but verification ends at hour 21 for 36km domain. Some hours missing from other domains, as mmout files got corrupted after the run was complete. 1998053112: MM5 forecasts end at hour 30. 1998053100: MM5 forecasts end at hour 36. 1998053012: MM5 forecasts end at hour 30. 1998053000: MM5 forecasts end at hour 36. 1998052900: MM5 forecasts end at hour 12 due to memory chip failure. Verification starts at hour 3 and ends at hour 12. 1998052815: No Meso-Eta run at NCEP, due to tests on new ETA. 1998052800: ETA verification missing hour 24. 1998052712: ETA verification ends at hour 33. 1998052600: Verification ends at hour 24. 1998052312: MM5 forecasts end at hour 12. 1998052300: MM5 forecasts initialized with ETA, then boundary conditions for the 12 to 48 hour forecast were generated with NOGAPS. 1998052112: ETA verification ends at hour 3. 1998052100: MM5 forecasts end at hour 27. 1998052000: MM5 forecasts end at hour 12. 1998051712: no forecasts due to computer crash. 1998051700: no forecasts due to computer crash. 1998051612: no forecasts due to computer crash. 1998051600: MM5 computer crashed at hour 21 of forecast. 1998051012: MM5 forecasts end at hour 24. 1998050900: No MM5 forecasts. 1998050400: MM5 forecasts end at hour 12. 1998042800: No verification stats. 1998042712: No verification stats. 1998042700: Almost no verification stats. 1998042500: ETA verification ends at hour 9. 1998042312: ETA verification missing hours 6 through 18. 1998042300: ETA verification ends at hour 9. 1998042212: MM5 verification ran to forecast hour 31 only, ETA stopped at 30. 1998042000: No ETA verification due to NCEP Internet problems. 1998041912: No MM5 verification. 1998041312: ETA verification ends at hour 9. 1998041112: MM5 or NCEP failure (cause not known). 1998041100: MM5 or NCEP failure (cause not known). 1998041012: ETA verification missing hours 9 through 18. 1998041000: ETA verification missing hours 18 through 30. 1998040912: NCEP communications problems. No MM5 forecast. 1998040900: MM5 forecasts end at hour 43. 1998040712: Model failed to produce proper boundary conditions. No MM5. 1998040600: MM5 forecasts end at hour 36. 1998040300: MM5 or NCEP failure (cause not known). 1998033012: MM5 forecasts end at hour 12. 1998033000: MM5 forecasts end at hour 12. 1998032900: MM5 forecasts end at hour 12. 1998032600: NCEP problems, ETA never ran. No MM5 forecast or verification. 1998032512: ETA verification ends at hour 33. 1998032503: No ETA verification because no ETA forecast. 1998032500: NCEP problems, ETA never ran. No MM5 forecast or verification. 1998032200: No ETA grids received, hence no MM5 verification. 1998032115: No ETA grids received, hence no ETA verification. 1998032112: No ETA grids received, hence no MM5 verification. 1998032100: No MM5 forecast or verification. 1998031803: No Meso-Eta for verification either. 1998031800: NCEP problems meant no ETA was run. No MM5 either. 1998031700: No MM5 forecast. 1998031612: Undiagnosed problem with ETA grids. Coincidentally, ETA switched to 3-hourly output at this time. Model ran to 6 hours. 1998031512: Missing ETA fields, no verification. 1998031400: Missing ETA fields, no verification. 1998030803: No ETA forecast. 1998030800: No MM5 forecast because of NCEP problems, ETA not run. 1998030703: No ETA verification due to missing fields in gridded data. 1998030515: No ETA verification because of incomplete gridded data. 1998030412: Verification stats end at hour 35 for domains 1 and 2, and hour 34 for domain 3. 1998030200: Tahoma crashed, no MM5 forecasts. 1998030115: ETA verification starts at hour 6. 1998030112: Verification for domains 1 and 2 starts at hour 8. 1998030103: ETA verification starts at hour 18. 1998030100: Verification for domains 1, 2 and 3 starts at hour 20. 1998022815: ETA verification starts at hour 30. 1998022812: Verification for domains 1, 2 and 3 starts at hour 32. 1998022803: No ETA verification files. 1998022800: No verification files. 1998022715: ETA verification ends at hour 18. 1998022712: Verification ends at hour 22. 1998022703: ETA verification ends at hour 30. 1998022700: Verification for domain 3 ends at forecast hour 34. 1998022600: Tahoma crashed, no MM5 forecasts. 1998022515: ETA starts at hour 9. 1998022500: ETA missing hours 10 through 23. 1998022412: ETA missing hours 22 through 35. 1998022400: Domain 2 files deleted, only 12 to 24 hour forecasts available. 1998022315: Meso-Eta missing crucial fields for verification. 1998021912: Verification stops at hour 24 for domain 2 only. 1998021900: MM5 crashed at hour 16. 1998021815: Did not receive Meso-ETA grids for verification. 1998021800: Tahoma crashed overnight. No MM5 forecast. 1998021712: ETA 32-km failures, no MM5 forecast run. 1998021700: ETA 32-km failures, no MM5 forecast run. 1998021612: MM5 ran out to hour 12 only. 1998021603: High-resolution Meso-ETA was missing crucial fields. 1998021600: Model ran to hour 36, verification stops at hour 34. 1998021515: No high-resolution Meso-ETA received. 1998021503: No high-resolution Meso-ETA received. 1998021415: No high-resolution Meso-ETA received. 1998021412: MM5 ran out to hour 30 only. 1998021400: MM5 never ran. 1998021403: High-resolution Meso-ETA was missing crucial fields. 1998021312: ETA forecast files were never complete (missing f12) MM5 never ran. 1998021300: ETA missing hour 30. 1998021000: ETA forecast files were never complete (2nd run of new 32km ETA had problems), MM5 never ran. 1998020312: no Meso-Eta forecast received 1998020312: model crashed after f12, re-run to hour 25 1998020112: model never ran 1998012812: model crashed after f35 1998012700: ends f21 due to model problem. 1998012600: No ETA verification. 1998012512: ends f10 for d1, f24 for d2, and f36 for d3 1998012500: ends f22 for d1, f24 for d2 and d3 1998012412: ends f34 1998012312: ends f10 for d1, d2 and d3 f12 to f24 1998012212: ETA ends at hour 15. 1998012200: starts f11 ends f34. MM5 d2 verification ends at hour 20. 1998012112: d3 only, starts at hour 23 1998012100: MM5 4 km domain verification missing hours 23 through 34. 1998012012: Forecast ends at hour 34 1998011612: MM5 4 km domain verification stops at hour 25. 1998011312: MM5 4 km domain verification starts at hour 23. 1998011300: MM5 4 km domain has only forecast hours 35 and 36 in verification history files. 1997123100: MM5 4 km domain has only forecast hour 36 in verification.
[ Model and Verification Problems | Observation Problems | Change Log | Realtime Verification Statistics ]

Observation Problems


990306 -- Hours 06 through 21 have many missing observations due to local computer problems (disk going bad on lorrie). 990121 -- UW,UW wind sensor reporting calm winds during 20 to 40 knot winds.
[ Model and Verification Problems | Observation Problems | Change Log | Realtime Verification Statistics ]

Changes to MM5 Verification


12/13/2001 Beginning with 2001121000, observations must be at grid point 1.5 or greater to be included withing a domain, since mygrintp.f truncates to the lower integer for the interpolation scheme, we had some 0 indices. Also, made small modification to code to use GN version so that no additional gplt process is spawned. This does not affect results. 5/16/2001 Beginning with 2001051600, make_obs changes. All observation files in ~mm5rt/data/obs/ will now have the following: . station reference numbers for surface and upper-air stations in characters 30:35 of the station name (char*40). . the bogus flag has been changed from True to False . the geopotential height field for surface obs has changed from missing, -888888., to the station elevation Note: as of the this time, all obs prior to 2001051600 do not have these modifications. This will be rectified soon. 2/12/2001 Formula change for wsp2, temp, and temp2 values starting with 2001021000. The 10-meter wind speed and the surface and 2-meter temperatures have been re-formulated to account for the varying height of the lowest half sigma level. That height is calculated assuming that the virtual temperature follows a standard 6.5K/km lapse rate in the altimeter equation (Wallace and Hobbs p. Old formulas z_ref = 40. elevdiff = StnElevation - terht(srf_idx) [Model's StnElevation] grid_temp = t(srf_idx) - .0065*(elevdiff - z_ref) logzozref = LOG(10./z_ref) logzozndev = LOG(z_ref/z_naughtdef) grid_wsp_reduced = grid_wsp * (1 + ( logzozref /(LOG(z_ref/z_naught(srf_idx))))) New formulas ghtsfc = terht(i,j) - tv(kmax)/.0065 * ((Psurf/prs(kmax))**(Rgas*-.0065/9.81) - 1.) z_ref = ghtsfc(srf_idx) - terht(srf_idx) elevdiff = StnElevation - terht(srf_idx) [Model's StnElevation] grid_temp = t(srf_idx) - .0065*(elevdiff - z_ref) logzozref = LOG(10./z_ref) logzozndev = LOG(z_ref/z_naughtdef) grid_wsp_reduced = grid_wsp * (1 + ( logzozref /(LOG(z_ref/bcz_naught(srf_idx))))) Because the new z_ref is less than the old static one (40 m), temperatures will be sligtly cooler (roughly 0.03 Kelvin) and wind speeds a little higher (roughly 0.04 m/s) 11/15/2000 Began using GEMPAK 5.6 and ICAO lat/lon/elevations for upperair sites within the bounding 36 km domain with 2000111212 verification data. 10/17 and 10/20/2000 Added BC RWIS, Puget Sound Energy, USGS, and Seattle City Light observations to the verification data base. Removed Sequim ob from Ecology type. Added ceiling, visibility, cloud cover, 6-hr temperature extremes, 3-hr pressure tendencies and trends, and wind gusts to the observational data set (beginning on October 20). These fields are not yet output by the model. 9/27/2000 ETA switched from 32 km / 45 level resolution to 22 km / 50 level resolution with the 2000092612 run. EDAS now includes cloud radiances and domain has been expanded westward. The MM5 is still initialized with the highest resolution grids we receive, namely, the 80 km ETA 104 grid. 6/15/2000 After noticing extremely low wind speeds over largely sloping terrain, at Mt. Rainier locations, we have removed the Ustar wind-speed drag modifications from the MM5 starting with the 2000061600 run. 3/28/2000 to 4/11/2000 Switched to binary format for history files and saved approximately 48%, 14 GB, in storage space. Format is unformatted, direct access with each forecast hour indicating the line number of the last entry for that hour. 12/21/1999 Bug fixes for latest, dec99, modeling system: * Several fields in the mifc character array of the mmout files were seriously garbled. Terrain variance file reading moved back to location in code prior to CALL ADDALL fixed this. Bad headers for 1999121300 through 1999122012. * Soil temperature data initialization using 40-day running mean was inoperative for forecasts 1999121300 through 1999122112. Turned back on with 1999122200. * Landuse type and roughness length incongruencies were used with verification data for 1999121300 through 1999121900. Result: 10-meter wind speeds generally seemed too high by up to 1/2 m/s. 11/5/1999 Bug in make_obs.pl meant that no upper air data was stored in the obs files and verification files for the period 1999102100 through 1999110212. Fixed bug in verify that resulted in losing the model predictions for some Snotel and all unknown obs types. All Schoolnet and other non-ecology sites included in the ~ldm/data/ecology/ files will be deleted from verification statistics prior to 1999110400, since they have no useful information. 10/7/1999 Discovered that Hydromet stations (type HM) all had incorrect elevations (feet instead of meters) for all observations prior to 1999100700. Thus, they should be excluded from any temperature statistics (where elevations are used) prior to this date. 9/23 and 9/28/1999 Ran verify/util/fixasos.pl to correct the mislabelling of several uncommissioned ASOS reports in ~mm5rt/data/obs and in history files. This primarily affected duplicate KNUW and KNRA obs. Any other obs in surf_hist files that were affected by the same dates, were also cleansed of erroneous duplicate obs. Ran verify/util/rewrite_ecology_and_doqc_990910.pl to change verification surf_hist files for Ecology names and ids. Gross-error qc similar to qc.pl was part of this script and failed sites were indicated in surf_hist*qcfail files. Ran verify/util/fix_more_duplicates.pl to correct accidental duplication of any obs in surf_hist files. (Less than 12 files affected). 9/17/1999 OTIS GEMPAK SST grids changed again (with NAWIPS 5.4.12 change). Starting with 1999091512, SST is found with GFUNC = WTMPK@0%NONE. Note: previous change from GFUNC = TMPK@1%DPTH to GFUNC = WTMPK@0%PRES occurred on 1999090912. USAF snow GEMPAK files changed with 1999091600. Snow has changed from GFUNC = SNDM@0%PRES to GFUNC = SNDM@0%NONE. 9/17/1999 Starting with 1999091800 run, MMRESTEMP for backend/one-way 4km run has been changed from FALSE to TRUE. Coarse domain substrate temperature will be used for the 4 km domain 3 run. 9/16/1999 Starting with 1999091612 run at forecast hour 21, the LATITCRS and LONGICRS arrays in the MM5 domain 3 output are correct. Before then, duplication occurred over the nest points. Verification is unaffected because it does not use those header arrays for anything. 9/9/1999 OTIS SST grids used in initialization (gridder) program changed name and location information for the SST field starting with 1999090912. Gempak parameters changed from gfunc=tmpk@1%dpth to gfunc=wtmpk@0%pres. Appropriate entries were changed in preprocess.pl for MM5 run 1999091000, the first run to use the new SST field designators. 8/31/1999 Changed locations and some names of cstcn stations in the surf_hist files. 8/19/1999 Fixed upper-air reports that were missing from 1999080212 through 1999081900 ~mm5rt/data/obs files due to Y2K update. Added primarily duplicate Canadian buoy data to cstcn decoded files and enhanced the lat/lons of BC buoys beginning with 1999082000. Expect to see improved wind verification statistics for type CC obs starting with that date. 7/22/1999 Added the following obs types to those used by make_obs and verification: RWIS/DOT Snotel -- will need good QC Also, Ecology data recently came back online after being gone for several months. 6/15 to 6/16/1999 Fixed the following corrupted history files: 1998060100: fixed d1 surface and upper-air history file duplicates 1998060712: fixed surface and upper-air history file duplicates 1998060812: fixed surface and upper-air history file duplicates 1998061712: fixed surface and upper-air history file duplicates 1998080912: fixed d2 surface and upper-air history file duplicates 1998081612: fixed d1 surface and upper-air history file duplicates Also, reran spatial and timeseries files to incorporate these changes. 5/1/1999 Fixed precip output files that were fouled up for 1999030900 through 1999043012. Switch to dynamic memory allocation caused only one element to be written out, rather than complete array. 3/11 to 3/13/1999 Modified MM5 verification to use true bilinear interpolation to grid points. Modified ETA verification to use bilinear interpolation instead of bicubic interpolation. Tests showed smaller errors with bilinear -- too much smoothing with bicubic. Fix corrupted history file surf_hist_mm5_d3_1999013000 that had erroneous observations added to forecast hour 12. Also corrected the corresponding values in the spatial files and the timeseries files. Fixed MM5 and ETA verifcation for the recording of 24-hour precip values in the surf_hist files. MM5 verifcation was always missing before; ETA was reporting missing when the observation was missing. Now, a model value is reported for the 6-hour and 24-hour precip values, regardless of the observed value. Precip fix to MM5 also fixed the short-term problem with precip values that were erroneously large in the forecast files. 2/17 - 2/19/1999 Noted that all surf_hist files through 1998081000 have hour H3Z data in each hour (H3 - 1)Z slot, i.e., 02, 05, 08, 11, 14, 17, 20, and 23Z data is incorrect since it is actually 03, 06, 09, 12, 15, 18, 21, and 00Z data. Noted that as of 1998080617, the RAWS sites changed their names, see http://www.atmos.washington.edu/mm5rt/raws_name_changes.html. All spatial files were re-written to reflect the new name, also rewrote them for the Ecology sensor name changes that were initiated in our department. Fixed some corrupted lines in surf_hist_eta_d1_1998052603, surf_hist_mm5_d1_1998052812, surf_hist_mm5_d3_1998052700, surf_hist_mm5_d1_1998102000, surf_hist_mm5_d1_1998102012, surf_hist_mm5_d1_1998102100, surf_hist_mm5_d1_1998102112, surf_hist_mm5_d2_1998102000, surf_hist_mm5_d2_1998102012, surf_hist_mm5_d2_1998102100, surf_hist_mm5_d2_1998102112, and surf_hist_eta_d1_1998053003. 2/10/1999 Added initialization time to spatial history files to speed up calculations. 12/23/1998 Partially fixed make_obs to handle surface pressures above 1050 mb. Only partial because it works only for those stations that also report an altimeter setting. The period of bad pressures in the verification files extends from around 1998121700 through 1998122000. Fixed a bug in make_obs.pl that was not waiting long enough to acquire Agrimet sites. Remade obs files for period 1998121600 through 1998122206. Agrimet sites begin regularly appearing 1998122307. Sites appear in verification first in 1998122200, sporadically in 1998122212 and 1998122300, then regularly beginning with 1998122312. Three sites are included within the 4km domain: AGOLW -- Goldendale, WA AHRHW -- Harrah, WA AMASW -- Manson, WA 12/14 and 12/18/1998 Fixed surf_hist mm5_d2 and mm5_d3 files for the period 1997092912 through 1998012000 and 1998031712 and 1998012712 to correct forecast date for forecast hours 21, 24, 27, 30, 33, 36, and 39. On 18th, fixed the corresponding spatial files that had already been created. Fixed a corrupted line in surf_hist_mm5_d1_1997121912 near entry for JMQY,UN,1997121912,1997122006,18. 12/10/1998 Hydromet data turned on. Added winds, one-hour precip, and several comments to the decoded obs. 12/07 and 12/08/1998 Fixed duplication of data from forecast hour 0 to 46 in 1998031400 files. Fixed surf_hist and upper_hist files for 1998031612 through 1998031712 that had coarse domain locations for the next MM5 coarse domain. Restored surf_hist files for 1997092912 through 1998121312 for all MM5 domains. No ETA comparisons for this period. Fixed post-processing statistics program to use knots, as opposed to meters per second. 11/20/1998 Ecology sensor names were changed at 19Z: "1F" => "LACEY", "3F" => "DUWAM", "6G" => "ZION", "1M" => "SOMTN", "3K" => "MEADO", "6H" => "WOODL", "2C" => "INDIN", "3L" => "POULS", "6J" => "BPAV", "2D" => "FS12", "3O" => "BCNH2", "7B" => "YAKCC", "2E" => "54AVE", "4A" => "MARYS", "8E" => "MONRO", "2F" => "ALEXR", "4D" => "EVRET", "8F" => "CRWNZ", "2G" => "KENT", "4E" => "LKFOR", "8N" => "ROCKW" "2H" => "PUYAL", "4J" => "LYNNW", "2L" => "ENUM", "4K" => "NOAA", "2O" => "SEATN", "5A" => "STEXA", No adjustment was initially made to rename stations in past history files -- this finally occurred on 2/17 - 2/19/1999 as history and spatial files were rewritten with the new names. 11/10 and 11/11/1998 Started verifying the ETA 104 grids. Went back to 98102712 forecast. Fixed wind direction in ETA verification of upper air observations. Initial bug whereby ETA forecast hours > 0 that ended at 00Z were not included was not fixed until forecasts from 98111200. Hence, 1998102712 through 1998111112 inclusive are missing 00Z data beyond forecast hour 0 (ETA only). 10/1 and 10/2/1998 Fixed corrupted temperatures and near-surface temperatures in all history files for the CMAN stations from 25 March 1998 to 28 April 1998. Ran qc_surf.pl to eliminate horribly wrong fields in surf_hist files for December 1997 through September 1998. 9/23/1998 Rewrote December 1997 through August 1998 history files in updated (comma-separated) format, converted to 4-digit years, and converted 3-letter ids to 4-letter ids. 9/1/1998 Added PAWS network data to verification statistics. Effective time is forecast 98083012. 8/11 to 8/12/1998 Moved all verifcation files into subdirectories based on the year and month (yyyymm). Renamed all verification files to use 4-digit years instead of 2-digit years. History output files from 1998080912 forecast on contain model initial and forecast times in yyyymmddhh format, all dates prior still have yymmddhh format for forecast and initial times in the file, though the file name uses yyyymmddhh format. 7/23/1998 Fixed a bug in the make_obs script that was erroneously duplicating upper air observations into off hours. Observation and history files containing observations prior to 98072311 are affected by this. A remedy is to select verification statistics at 12-hour increments only (which is what one should select anyways). 7/2/1998 Noted that Meso-Eta has come in sporadically since supposedly being turned off at NCEP. Hence, some ETA verification has been successful over the past month. 6/10/1998 Corrected some history files written between 4/26/98 and 5/2/98 that had erroneous fields added in and final fields deleted. Some moving stations end up with an MM5 coarse-domain y-coordinate of -999.99 for this period. 6/3/1998 Meso-Eta was permanently cancelled by NWS. Changes to verify not yet implemented. 4/29/98 to 5/4/98 Fixed ref_temp in verify code. Now read in from model header. Added obs type to surf_hist and upper_hist files. Changed surf_hist and upper_hist output format to use commas instead of spaces to separate items. All surf_hist and upper_hist files re-written to this format on May 6, 1998. Programs rewrite_surfhistory.pl and rewrite_upperhistory.pl can change between the formats. 4/23/98 More obs implemented in verification. Additions of types AV, HM, AP, and RW. 3/25/98 Changed to 4-letter station ids on the 98032512 verification files. All previous files rewritten 9/24/1998 to conform to 4-letter station ids. Verification software began using all of the extra observations. Verification software began using 24-hour observed precipitation. The following is a list of MM5 runs that had low resolution terrain data: 98031712 98031812 98031900 98031912 98032000 98032012 98032212 98032300 98032312 98032400 98032412 98032512 3/18/98 Increase in number of observations making it into verification finally realized. Increased from 150 at 00Z to 1125 at 00Z for the whole 36-km domain. MM5 domains shifted. Switched back to mm5v2.7 with ustar mods. Began using 1/4 degree SST data. Began using ETA 3-hourly output (old used 6-hourly outputs). 2/11/98 Added all the extra obs to the observation files, though they were not used by verification until 98032512. 2/4/98 U-star mods were missing from model runs 98012900 through 98020400. 1/26/98 Fixed ref_temp in mm5_verify code. Noted that main.f need to correct dates (off-hour mmout files get the wrong obs files). 1/20/98 Switched to version 2.7 of MM5 and read_model.f. Initial U-star mods added. Pressure calculations in upper_hist files out of whack due to ref_temp set to 275 while verify code maintained the old 290. This corrected on 98012400 verify files. 1/12/98 Added 1-hour and 24-hour (at 36 and 48 hours) precipitation fields, but it did not become fully operational until March 25, 1998 12Z. 12/18 - 12/19/97 Added hourly output to 12 km and 4 km domains after forecast hour 12. 9/29/97 4 km domain added.

[ Model and Verification Problems | Observation Problems | Change Log | Realtime Verification Statistics ]

Realtime Verification Statistics
Back to MM5 Alternate Format Page
Back to MM5 Standard Form Page