EUMETNET has communicated to its members the intention to switch off GTS transmission of E-ASAP data with the linked letter.

Here follows a list of GTS headers: 

TEMP A: USVX01 EDZW
TEMP B: UKVX01 EDZW
TEMP C: ULVX01 EDZW
TEMP D: UEVX01 EDZW
Conversion of TEMP to BUFR from surface to 100 hPa: IUKX40 EDZW

HiRes Bufr from surface to burst height: IUSX40 EDZW


Here is a catalogue of ASAP vessels .


On 14 October 2014, DWD announced by METNO C that the TAC bulletins (U[SKLE]VX01 EDZW) will be terminated on 10 November 2014.  Source: ftp://ftp.wmo.int/wmo-ddbs/OperationalInfo/VolumeC1/From_WMO/METNO/2014/MC4214.txt


This page will be updated when more information will be available.


Investigation of BUFR SHIP TEMP data for 1-5 July 2014

I did an in depth investigation of the BUFR SHIP TEMP data available in the ECMWF system for these five days and found a number of problems or features.  These have been discussed with the E-ASAP Operational Service Manager and his responses have been incorporated in the notes below.

  1. Reports from the ship reporting as ASDK3 have a missing identifier in the BUFR data (as processed by ECMWF).  It turns out that the identifier is right justified (rather than left justified as it should be) by the GRAW software (the other E-ASAPs use Vaisala or Modem systems).  This will be corrected soon.
  2. BUFR reports from ASFR04 contain many erroneous dew point temperatures (some higher than air temperature).  This ties in with a comment by Pat Pauley at  https://confluence.ecmwf.int/display/TCBUF/Errors+in+BUFR+radiosonde+data "Negative dew point depressions (temperature minus dew point temperature) on some significant wind levels in BUFR data from France."  There are also noisy winds in some of the Modem BUFR reports (both ASAP, ASFR*, and land).  Reported to MeteoFrance.
  3. A few BUFR reports (both ASAP and land) have very noisy dew point temperatures in the upper stratosphere (the humidity is not generally used at these levels).
  4. Metadata.  In some cases the callsign has '0' inserted, eg ASEU3 becomes ASEU03 in BUFR.  In SHIP TEMP code the positions are reported to 0.1 degree lat/long (mostly truncated rather than rounded as I would have expected) in BUFR they are reported to four decimal places.  The TAC report usually has the nominal ascent time eg 1100 UTC the BUFR report has the actual report time eg 1033 UTC (sometimes seconds as well) - in a few cases the BUFR time is over an hour earlier.
  5. There are two BUFR reports due to the requirement to deliver data from surface to 100 hPa and from surface to burst height. The low resolution for the first part (surface to 100 hPa) is to save satellite traffic costs instead of transmitting redundant high resolution data.  The full report is at high resolution.
  6. There were two non E-ASAP vessels reporting in TAC code during this period.  DBLK wasn't reporting in BUFR.  "The configuration on board the RV Polarstern (ASAP station DBLK) shall be modified in October 2014 (during dockyard stay of the ship in Bremerhaven) to transmit BUFR."  Data from the Japanese ship JGQH was available in BUFR, apparently reformatted from TEMP (some issues are being investigated). 
  7. There were some TEMP DROP reports, these did not appear in BUFR (as processed at ECMWF).

Bruce Ingleby  24 July 2014