Date

Mar 06, 2014

Attendees

Goals

  • TIGGE/TIGGE-LAM time-series archive set up

Discussion Items

TimeItemWhoNotes
1.location list 

There is a general agreement to use GTS synop stations for both TIGGE and TIGGE-LAM datasets

Regarding additional point of interest for high resolution European TIGGE-LAM forecasts only 4 proposals were received from partners during last half a year:

  1. Italy - to include some selection of stations from their own high density networks (around 500 locations which will cover mainly Northern Italy, with special attention to the Po basin)
  2. DWD - to add some wind mast stations (around 40 already provided), but there is a link to new parameters requirement which would have to be archived
  3. Lawrence Wilson - to consider e.g. European high density networks at least for precipitation (MAP project)
  4. MetOffice - they will provide some additional locations within the next 48 hours after the meeting

It was mentioned that DWD request should be further discussed during the 1st half of 2014 as currently it's supported by one partner only and it would mean a significant change to the current archive set up and operational running (see below @2).

2.

parameter list

 

It was confirmed that so called High Priority parameters will be archived in the first phase.

Only DWD proposed to add some new parameters related to near future renewable energy applications. Their proposal is to add e.g. direct or diffuse radiation or low/medium/high cloud cover. For wind mast stations some wind data from lowest models' levels would have to be newly provided as well to be able to compute the values at the wind mast height for given location.

It was mentioned that it would be excellent to have access to some observation database with data for the chosen locations. It seems to be hardly possible at the moment because of current data policies on countries' level. It might be an option to have access at lest to the freely available data from synop stations.

3.method for getting the model forecast for given location Confirmed that the 4 closest points should be OK.
4.archival format Not discussed, just announced that NetCDF will be used and technical work is in progress.
5.AOB There was a question when and how the time-series archive will be created. The current estimate is that the re-processing of TIGGE/TIGGE-LAM datasets will start in May/June  this year earliest. The re-processing itself of all data since 2006 will take probably a few weeks.

Action Items

  • Tiziana/Andrea will provide the selected list of Italian precipitation stations from their high density network
  • Ervin will investigate if we can retrieve easily locations of  European precipitation stations from high density networks which are used at ECMWF on regular basis

Original meeting announcement

Dear colleagues,

You're reading this e-mail because you belong to some group(s) below:

  • TIGGE-LAM archive data providers
  • TIGGE LAM Panel members
  • SRNWP expert team on verification
  • SRNWP Expert Team on EPS and Predictability
  • GIFS-TIGGE group

In the last August the first e-mail about proposed time-series archive set up was sent. As we haven't received too many reactions we would like to organize a webex conference on the same topics the next week. If you are interested please read following information or pass this e-mail to any other appropriate personnel who might like or need to join us.

Please vote at doodle poll below about your availability for the webex the next week 4.-6.3. (Tue-Thu)
http://doodle.com/wh83tyx4imke69dm

Please be aware of the fact that you will need a laptop with windows or mac installed to be able to join i.e. there will be no possibility to join by phone and only limited chance to join from linux environment (32 bit linux only works with webex to our experience).

================================================
Meeting agenda

One of the ECMWF's tasks in the GEOWOW project is about creation of EPS time-series archive derived from TIGGE/TIGGE-LAM datasets for selected locations.

Please find below a link to a short document summarising our opinions  what can be stored and methods which can be used to derive the point values from TIGGE/TIGGE-LAM forecasts:
https://confluence.ecmwf.int/download/attachments/22904893/GEOWOW-WP4-Tigge-TimeSeries-Plan-20130815.doc?api=v2

Based on information from the document our current proposals are:

  1. location list - option 1 in the chapter 2  => GTS synop locations + selected special points of interest (to be defined)
  2. parameter list - the 1st and some of the 2nd order parameters in the chapter 6 => e.g. all TIGGE-LAM parameters listed at https://confluence.ecmwf.int/display/TIGGE/Archive+specification
  3. method for getting  the model forecast for given location - a variant of the option 1 in the chapter 3 => archive halo of 4 surrounding nearest model grid point values sorted according to their distance to the point location + some additional information (land-sea mask, orography, distance from the point location) 
  4. archival format - NetCDF

There is a need to agree on  final archive set up as it cannot be changed easily afterwards (we are planning to re-process all TIGGE data since 2007, it means a huge amount of data).

 

The most important topics of the meeting should be the first two points about location and parameter's lists. Especially we'll need to know your proposals about any special point locations which you would like to add to the currently planned station list (synop stations).


For the latest up-to-date information and other documents please refer to the ECMWF TIGGE-LAM time-series web site:
https://confluence.ecmwf.int/display/TIGGE/TIGGE+EPS+time-series+archive
All important information and proposals will be regularly updated there and your comments which can be added directly to every web page  are always highly appreciated.

 

  • No labels