2014-09-09/22 Per Undén



UERRA MST meeting 3, 8 October 2014



      1. Present:

Per Undén, Dale Barker (Skype), Manola Brunet, Andrea Kaiser-Weiss, Albert Klein-Tank, Eric Bazile, Frank Kaspar (observer).

      1. Agenda

Action points from MST 2.

  1. How to add observations for use in WP2

  2. How and where to store / archive UERRA and access them for WP3 (also EURO4M data)

  3. Update of work in the WPs

  4. Deliverables
    In particular the Dissemination plan (WP7), Coordination plan (WP6), Web portal (WP9)

  5. Periodic Report 31 December with Scientific part, time schedule and for the GA

  6. Meetings and communications

    EMS/ECAC meeting in Prague 6-10 October

    GA 2015
    Dates and times set and communicated in May (and on the Web site)
    Programme and planning etc. and for the ESAB

  7. Common web portal WP9 ?

  8. Other items

  9. Next meeting



  1. Action points from MST2

  1. The Coordinator contacts the Project Officer Stijn Vermoote with an update on the status of the early Deliverables including the delayed ones.

This has been done and since it is early in the project the delays should be quite manageable.

  1. Per will discuss with Jan Keller and Dale about any changes to UBO's reanalysis work.

Christian Ohlwein replied explaining the change of focus for HErz Centre and some foreseen delays of the work in UERRA. Per sent a long reply (Dale cc:ed) analysing the situation and the delays should be managable and also the planned 5 years of reanalysis. (A further reply is expected soon when Sarah Jones is back). Dale advised to inform Stijn about changes of the work. Per will take it up anyway in a coming update of our status. Action.

  1. Andrea will circulate the latest draft version of the D3.2 Evaluation procedures report with a deadline of 30 September for update. The updates should be in the on-line documents, as instructed by Andrea.

Done and D3.2 has been delivered.

  1. Manola will contact us to see what is available nationally in scanned format. Manola has offered the countries to digitize manual records. Also discussions about which data to digitize from Romania have not been completed yet.

We got the reply from Romania and it was negative. No extra data can be transferred due to national data policies. Contacts with MF and DWD were positive and will be followed up more. Slovenia can supply some but Libya is difficult due to unrest in the country. NCDC and Peter Thorne have WW2 data which will be checked for inventory.

  1. Manola will contact Roxana Bojariu again soon and will make a CC to Per to have him in the loop.

Done.

  1. How to add observations from WP1 for use in WP2

Rescued and processed observations from WP1 are processed by Harry (CRU) to be put in ODB format and such a batch will be done soon. ECMWF will only update MARS annually with such data, according to what Ian Harries and Manola were told in their May meeting along with Dick Dee, Manolo Fuentes and Hans Hersbach at the ECMWF headquarters. It may not be a big problem since the Met Office production runs will not start before 2016. SMHI need to start earlier but may benefit later on. If there are substantial extra data that can be used for RRA in the meantime, we would like to receive information of the content from WP1. EURO4M data are still sitting outside of ECMWF (at the ZENODO data bank: https://zenodo.org/record/7531#.VD0yfhadiXQ, ECA&D, ISTI databank: http://www.surfacetemperatures.org/databank and at the ISPD: https://reanalyses.org/observations/international-surface-pressure-databank the air pressure data ). Currently only air pressure hourly observations digitized and quality controlled under EURO4M are being converted into the ODB format, but not the remaining daily maximum and minimum temperatures and rainfall amounts. If these daily observations are also of interest for any component of the UERRA work-load, it should be defined both the data format and the delivery mechanism that suits better with the relevant UERRA WPs. They need to be transferred as well. (Any more thoughts how to do this , through CRU?, Manola and Albert?). Action.

The ECA&D has got much more data now and for the 2D- downscaling RA this data set should be checked for extra observations. (For MF to check with KNMI).

The national French data (especially the daily precipitation) used in EURO4M MESCAN and MESAN may not be possible to store at ECWWF due to MF data policy . This may not be a problem for WP3 anyway since mainly TEMP data in the ODBs will be compared.

  1. How and where to store the RA data

It is a critical issue how to store the RA data for use in WP3, in the first instance. Now, in the early stages EURO4M data will be used. The period done by the MO is in MARS but the SMHI data still needs to be extracted from their MARS and stored for DWD at ECMWF.

ECMWF would prefer to store in GRIB2 and Per explained that HARMONIE is not ready yet for GRIB2 and it may take half a year or so. Work is needed to define parameters with respect to the GRIB2 tables and possibly contact WMO if new parameters need to be added. There may be some short cut using other formats for other projects at ECMWF and store the parameters listed by WP3 at least.

The MO did tweak their output to be able to archive in MARS but SMHI has worked with ECMWF to have an updated version which can handle other models like HIRLAM and HARMONIE/ALADIN. We need to discuss this with ECMWF (again) and Per will take the contacts (Action Per) .

Depending on the outcome and the availability for work by ECMWF it will be some time (months?) before WP3 can use the new data.

The Deliverable of MARS archives at M48 is to late for the project, so we will at least need archives of the subset of common parameters.

The producers need to send data to ECMWF (to be agreed with them) to inspect and check how to archive. (Action for Andrea to provide Dale with parameter list collecting what WP3 requires short-term).

  1. Update of the work in the WPs

WP1.

We may soon run out of post-1961 data to process due to the problems mentioned above. In that case it is justified to do more early data than planned. Even if they are not used in WP2 they constitute a contribution to the meteorological community and the global RAs.


WP2.

The setting up of the RA systems is going on at the MO and at SMHI. The choice of ensemble variational method has not been made yet but both Ens-4DVar and 4D Ens-Var are there but availability of boundary ensemble forcing influences the decision as well as efficiency. The rain data from E-OBS will be used.

SMHI and MF have tested model versions and there are some issues to be sorted out on biases and small scale precipitation (as well as the archiving). The interactive LAI does not seem to work in DA and it needs to be thought about too. (It is probably only a small refinement but mentioned in the DoW as is the EKF for soil (Per added these comments).

MF has developed and implemented methods for creating pseudo obs from Tmin and max so they can use such daily observations in case of missing hourly data.

UniBonn activities were discussed above.


WP3.

The Deliverables are there and work continues now on data sets that are available. We need to collect all the data on which the preliminary evaluation routines should run in one place, with comparable parameter names. The MARS archive would be the best option. If not possible, we need netcdf fields at ecfs. A request will be formulated to WP2 (see Action Andrea above).

WP4.

The issues now are with MARS as mentioned above.

WP6.

The Scientific Coordination plan has been written and it seems acceptable and there were no comments on that (before or at the meeting).

WP7.

The Dissemination Plan (D 6.1, very late) has been written some months ago and more comments have been received from Albert. It can soon be finalised. It will be sent to the other MST members as well (Action).

WP8.

The D8.1 Initial review of user requirements was discussed. It will build on the work in EUOR4M and other projects. Andrea collected user requirements at the WP3 workshop and wishes and needs to be involved in the process when Albert works on this (for M12). (DWD has 1 pm on this too),

WP9.

A telephone conference has been held with 4 of the 5 Coordinators of the Copernicus Projects. Information was exchanged and e.g. information about each each other's meetings is valuable in order to attend some of those.

The D9.1 Common web portal was discussed and now, lately via e-mail. CLIPC is supposed to provide a web portal for tools and data access for all the data produced of our projects. Under that there should be a common web page (pages) and information about our projects. (It is not primarily an UERRA, action and we need to inform our PO about the Deliverable. Discussion has continued between CLIPC and us via mail after the MST meeting, N.B.).

 

  1. Deliverables

The Dissemination plan and the Coordination Plan are basically ready and will soon be delivered.

The Common Web portal/ Common web page D9.1 needs to done when the 5 CCCS projects have agreed and CLIPC implemented pages and information filled from e.g. UERRA in our case. The PO will be informed in connection with a status update (Action, Per)

  1. Periodic Report and Scientific report

The reporting is due for M1-M12 and our GA takes place 27-28 January. The ESAB should come and they will need to have the Scientific reports in time before the meeting. It means that the Scientific reports from each partner needs to be received mid December. The Coordinator will set up a template for these reports and send out to the MST (Action) and then to the partners for the actual reporting. (Before 1 December)

  1. Meetings and communication

The arrangements for the GA in Tortosa were discussed. There are plenty of meeting rooms and a block booking for 30 will be made at the Parador (it is recommended to stay there).

Manola will send details to Per and he will send it all out to partners before the end of October (Action).

  1. The Web Portal/common web page

See above under Actions from MST2.

  1. Other items

None.

  1. Next meeting

A telephone meeting will be arranged with a Doodle and for arranging the GA and reporting etc. , so early December.

  1. Action points

  1. Per will update the status of the work and Deliverables and particularly about UniBonn to the PO.

  2. Manola will investigate and discuss how to transfer EURO4M SLP observations to ECMWF.

  3. Per will contact ECMWF (Manuel Fuentes) on how to be able to archive MO, SMHI and COSMO RRA data in MARS.

  4. Andrea will ask WP2 and Uni Bonn to put their available products (also from EURO-4M) covering at least 2008-2009, at least the parameters required for preliminary evaluation, to ECMWF for MARS archiving.

  5. Per will disseminate the Dissemination Plan to the MST.

  6. Per will compile a template for the Scientific reports and send it to the MST.

  7. Manola will compile practical information about the GA in Tortosa and Per will send it out to partners at the end of October.