Attendees

• Tiziana Pacagnella (TP)
• Richard Swinbank (RS)
• Ervin  Zsoter ( EZ)
• Richard Mladek (RM)

Meeting Minutes

1) TIGGE-LAM archive

• the current status quo was shortly presented by RM
• RM mentioned  technical issues related to multi-model LAM-EPS systems (e.g. GLAMEPS or PEPS) which were not present in TIGGE(global) processing chain in the past. To those peculiarities can be counted e.g. varying EPS size, different available parameters for different EPS members, varying forecast length and integration domain within the members etc. It was suggested to try to adapt the current TIGGE-LAM archive scripts to be able to digest such multi-model data in the best possible way (datasets usability must be of the highest importance e.g. for PEPS all forecasts will be archived only on the common interpolated grid). The step-by-step approach for archiving of some currently incomplete datasets was suggested what means to archive at the moment what's available and ask providers to try to provide as much of the defined TIGGE-LAM parameters in near future as possible.
• TP mentioned possibility to include some case study ensemble forecasts in the archive in the future e.g. HYMEX or FROST-2014 ; Brian Golding is keen that these data be available to support the HIW project. It  would  depend also on agreement with ECMWF and archive  status  after the end of GEOWOW
• RS shortly mentioned a situation with MOGREPS-UK ensemble system. The first data samples are expected soon. Some parameters will  not be included (CAPE, CIN) as they are not produced as prognostic fields by the model. Large scale precipitation is not included either due to the model characteristics (convection-permitting with 2.2km resolution )
• TP mentioned theoretical possibility of archiving model outputs hourly instead of 3-hourly in the future
  

2) TIGGE/TIGGE-LAM time-series archive

• RM mentioned ongoing preparatory technical work towards the time-series archive (NetCDF format is being tested etc.)
• RS mentioned a potential strong interest of many users for on-the-fly GRIB2 to NetCDF conversion. It would be the best if such feature is available directly in the TIGGE data portals
• what to store from model forecasts for each point location? The proposal is to store 4 surrounding closest points with some additional basic information (distance to the point location,  land sea mask, orography height(?), the best recommended interpolated or nearest land point value for given location(?)) 
• parameter list
  TIGGE
  
  • create a proposal for  GIFS-TIGGE   working group members . The document about time-series datasets prepared by EZ can be updated and used for that (1st & 2nd order parameters, interpolation methods etc.)
  • requirements of HIW (High Impact Weather) project should be taken into account
  • include any  upper level data ?   (or just surface parameters , as listed in current document ) ?
  • the parameter list should depend on agreed priorities and involved partners' requests (is energy sector a core member for HIW?)
  TIGGE-LAM
  • possibility to add some new parameters in the future if required by users or project partners (e.g. fluxes or some vertical level data)
• point location list
  TIGGE
  
  • use GTS synop locations ( at minimum the list of GCOS climate monitoring stations, but could also  include  up-to-date station list)
  TIGGE-LAM
  
  • some existing high density stations' location might be used (e.g. precipitation stations network; contact Anna Gheli)
  • providers will be asked for their suggestions  for stations within their forecast model domain ; priorities will be set up (contact Lawrence Wilson)
  • TP mentioned an example of 1500 special points of interest in Italy related to civil protection network
  • how many point location forecasts can be afforded to archive? Is there any upper limit? Some reasonable limits must be found and agreed by all partners.    [Is 10,000 locations in any one model domain reasonable?] 

3) TIGGE-LAM presentation at SRNWP meeting (Antalya 30.9.-3.10.)

• RM will verify a possibility to participate at the SRNWP meeting (funding needed from GEOWOW). It would be an excellent opportunity to meet most of the data providers and discuss topics mentioned above.
• RM + TP jointpresentation (or poster?) would have to be prepared including:
  • GEOWOW introduction
  • TIGGE-LAM and prepared time-series archives introduction
  • possible on-line demonstration
  • mention the opened topics (station & parameters lists etc.)
  • mention the potential for archiving of some case study forecasts in future (HYMEX, FROST..)

Follow-up actions

• The document about TIGGE time-series archive requirements and proposals prepared by EZ in the past will be updated (action RM & EZ) and sent to the partners for discussion
• The partners (data providers, subject matter experts etc.) will be asked in advance by e-mail (action RM) for their suggestions about point location selections especially in relation to TIGGE-LAM over European area
• some obsolete information and documents will be updated on the official TIGGE-LAM web pages www.smr.arpa.emr.it/tiggelam (action TP)