Welcome, Guest
Username: Password: Remember me
General topics

TOPIC: Swiss obs missing for Canari

Swiss obs missing for Canari 4 months 1 week ago #2551

  • Bert van Ulft
  • Bert van Ulft's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 109
  • Thank you received: 22
Dear all,

in a reanalysis simulation for the NETHERLANDS we saw a drop in the number of observations used by Canari (purple line):

WINS50_43h21_fERA5_CTL_ptA_Nobs_all_hours.png


Looking with obsmon at a map is seems the Swiss observations are no longer included since 2020111015 (up to at least 2020123121):
2020111101220201111015
obsmon_can_rh2m_2020111012.png
obsmon_can_rh2m_2020111015.png
Since 2020111015 the BUFR codes seem to have changed and there are many "inconnu warnings:"
* WARNING - BATOR : template inconnu pour fichier N.   5722
      301150      307080        5001        6001        7001
The version used by the experiment is tag harmonie-43h2.1, but the two maps above are with a develop version of last week.
Does anybody know how this can be fixed?

cheers,

Bert
Last Edit: 4 months 1 week ago by Bert van Ulft.

Swiss obs missing for Canari 4 months 6 days ago #2552

  • Eoin Whelan
  • Eoin Whelan's Avatar
  • OFFLINE
  • Gold Boarder
  • Posts: 221
  • Thank you received: 45
Hi Bert,

Nice time-series plots!

Quick answer: If you can find a "307080 5001 6001 7001" definition in the param files, copy it, change the "4" to a "5" and add the 301150 to give something the following:
BUFR synop_1
5 2 2 28
codage     1  301150
codage     2  307080
codage     3  005001
codage     4  006001
codage     5  007001
control    1      42 second delayed replication factor's position
control    2       2 iteration's number for RR
offset     1       4 nb of replicated B descriptors
offset     2       5 nb of replicated B descriptors
values    -1  007063 Depth below sea/water surface
values    -1  022043 Sea/water temperature
values    -1  001101 State identifier
values     1  001001 WMO block number
values     2  001002 WMO station number
values     3  001015 Station or site name
values     4  002001 Type of station
values     5  004001 Year
values    10  005001 Latitude
values    11  006001 Longitude
values    12  007030 Height of station ground above mean sea level
values    13  007031 Height of barometer above mean sea level
values    14  010004 Pressure
values    15  010051 Pressure reduced to mean sea level
values    19  007004 Pressure (standard level)
values    20  010009 Geopotentiel height of the standard level
values    22  012101 Temperature/dry-bulb temperature
values    23  012103 Dew-point temperature
values    24  013003 Relative humidity
values    28  013023 Total precipitation past 24 hours
values    30  020010 Cloud cover (total)
values    37  031001 Delayed replication factor (the first in file)
values    61  013013 Total snow depth
values    72  004024 Time period in hours
values    73  013011 Total precipitation/total water equivalent of snow
values    86  004025 Time period (-10, or number of minutes after change)
values    87  011001 Wind direction
values    88  011002 Wind speed
/BUFR synop_1

If you can provide sample I can take a look.

Eoin
Last Edit: 4 months 6 days ago by Eoin Whelan.
The following user(s) said Thank You: Bert van Ulft

Swiss obs missing for Canari 4 months 6 days ago #2553

  • Bert van Ulft
  • Bert van Ulft's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 109
  • Thank you received: 22
Thanks Eoin! I have given it a try, but Bator failed and didn't produce an ecma file, subsequently causing Canari to fail. I will do a bit more testing and get back to you. It is a default simulation with develop on ECMWF.

Bert
Last Edit: 4 months 6 days ago by Bert van Ulft.

Swiss obs missing for Canari 4 months 6 days ago #2557

  • Bert van Ulft
  • Bert van Ulft's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 109
  • Thank you received: 22
Adding the block below to const/bator_param/param_bator.cfg.conv.mars bring back the Swiss. It adds 4 WIGOS parameters, all others are shifted 4 positions.
As mentioned this morning it has no impact on the upper air analysis. Looking at Swiss obs for a 15UTC cycle:
  • the old-style obs (307080 5001 6001 7001) are present for 14h (typicalTime), with masterTablesVersion=16
  • the new style obs (301150 307080 5001 6001 7001) are present for 15h & 16h, with masterTablesVersion=29
But if and how this can explain the different behaviour for surface and upper air analysis I don't know.
I will create a pull request for this change after tests have completed.
#NNN
BUFR synop_1
5 2 2 32
codage     1  301150
codage     2  307080
codage     3  005001
codage     4  006001
codage     5  007001
control    1      46 second delayed replication factor's position
control    2       2 iteration's number for RR
offset     1       4 nb of replicated B descriptors
offset     2       5 nb of replicated B descriptors
values    -1  007063 Depth below sea/water surface
values    -1  022043 Sea/water temperature
values    -1  001101 State identifier
values     1  001125 WIGOS identifier series
values     2  001126 WIGOS issuer of identifier
values     3  001127 WIGOS issue number
values     4  001128 WIGOS local identifier (character)
values     5  001001 WMO block number
values     6  001002 WMO station number
values     7  001015 Station or site name
values     8  002001 Type of station
values     9  004001 Year
values    14  005001 Latitude
values    15  006001 Longitude
values    16  007030 Height of station ground above mean sea level
values    17  007031 Height of barometer above mean sea level
values    18  010004 Pressure
values    19  010051 Pressure reduced to mean sea level
values    23  007004 Pressure (standard level)
values    24  010009 Geopotentiel height of the standard level
values    26  012101 Temperature/dry-bulb temperature
values    27  012103 Dew-point temperature
values    28  013003 Relative humidity
values    32  013023 Total precipitation past 24 hours
values    34  020010 Cloud cover (total)
values    41  031001 Delayed replication factor (the first in file)
values    65  013013 Total snow depth
values    76  004024 Time period in hours
values    77  013011 Total precipitation/total water equivalent of snow
values    90  004025 Time period (-10, or number of minutes after change)
values    91  011001 Wind direction
values    92  011002 Wind speed
/BUFR synop_1
Time to create page: 0.076 seconds