GIFS-TIGGE
TIGGE-LAM Panel
April 2008
TIGGE LAM STATUS
April 2008
Tiziana Paccagnella
TIGGE LAM STATUS
April 2008
1.Archiving of TIGGE LAM output parameters: 1st phase
1.1.Ongoing steps and outcome of the Pretoria meeting
2.Archiving of TIGGE LAM output parameters: 2nd phase
2.1.Ongoing steps and outcome of the Pretoria meeting
3.Initial and Boundary Conditions
3.1.Link with the SRNWP Interoperability project
3.2.Ongoing steps and outcome of the Pretoria meeting
3.3.ICs & BCs availability and archiving
3.4.Initial and boundary conditions related to TIGGE Phase 1: Dataset for research. (some set of archived fields available during selected specific periods).
4.Relocation of LAM EPS systems and GIFS
4.1.Ongoing steps and outcome of the Pretoria meeting
5.Link with the Beijing RDP(B08RDP)
5.1.Ongoing steps and outcome of the Pretoria meeting
6.Tigge LAM Scientific Issues
7.TIGGE LAM meeting
8.Contacts and cooperations
Appendix 1: Provisional list of TIGGE LAM output parameters
Table 1:TIGGE LAM output parameters: single level
Table 2:TIGGE LAM output parameters: pressure levels
Output Fields Format
Appendix 2: ARKIMET short description
This document presents a quick review of the TIGGE LAM activities and discussions carried on by e-mail and during the first preliminary TIGGE LAM meeting held on the 4th of October 2007
in San Lorenzo de El Escorial – Spain, aside of the EMS/ECAM conference.
This report includes comments and decisions come out from the discussions during the TIGGE WG workshop held in Pretoria in March 2008.
1.Archiving of TIGGE LAM output parameters: 1st phase
The final set-up of the TIGGE LAM archive will consist of a decentralized archiving system hosted by some Regional Centres. This is something to be designed and harmonized with the Phase 2 of TIGGE now under planning.Coherently with the TIGGE directives, and starting from the TIGGE (global) list of output parameters, a list of output parameters has been compiled for the TIGGE LAM systems and it is reported as Appendix 1 to this document.This list must be considered provisional also due to some changes we are doing related to the future coding in GRIB2 format.
During the last months, it was decided that during this first phase of our TIGGE-LAM activity, it would be a valuable step forward to organize the archiving of some of the output parameters generated by the existent LAM EPS systems. For this reason some single level parameters have been labelled as HP (High Priority) sincethey are the most commonly used and verified. These HP parameters are:
- Mean sea level pressure
- 10m U-velocity
- 10m V-velocity
- Wind Gust
- Surface air temperature
- Surface air dew point temperature
- Total precipitation (liquid + frozen)
- Convective & large scale precipitation
In this first phase of TIGGE-LAM, to make easier the use and exchange of these fields, we agreed to ask the three TIGGE Archiving Centres (NCAR, ECMWF and CMA) for archiving also these LAM EPS HP parameters.Due to the regional nature of LAMs, the outputs coming from the different systems will be archived in one of the three Centres following a Geographical/Regional competence principle.The metadata info will be collected and accessible on the TIGGE-LAM web site at ARPA-SIM in Bologna (
We agreed about the proposal to archive the output fieldson a regular lat/lon grid at 0.1° x 0.1° hor. res. The full resolution fields “should “be available at the originating centres. This last issue will be further discussed and better addressed.
ECMWF, NCAR and CMA have been recently contacted through the TIGGE WG to evaluate the feasibility of the proposal. Philippe Bougeault and Steve Worley already expressed the willingness of ECMWF and NCAR to support TIGGE LAM in respect to this. De Hui Chen will communicate the position of CMA as soon as possible.
The inclusion of the HP TIGGE LAM parameters in the TIGGE archives is hold to some conditions like:
the volume of data is small compared to the volume of TIGGE global
the additional work for the archiving centre can be considered minimal.
the TIGGE LAM data providers must commit to use the TIGGE formats, and in particular all the metadata
each new EPS system receives an unambiguous name agreed across all partners
……
1.1.Ongoing steps and outcome of the Pretoria meeting
The TIGGE WG appreciated the work we did about the first phase of our archiving system implementation.We are now working on the specifics related to GRIB2 coding together with Manuel Fuentes. The request of some of the necessary infos was sent to the LAM EPS groups. GRIB2 coding is still under implementation for almost all the groups.
Tiziana will contact again the different groups to have more clear indications aboutthe volume of the data they will make available for archiving and the time they need to provide these fields in GRIB2.
About the data Policy, the Panel proposes to follow the same rules adopted by TIGGE but making available the data with a delay of 24 hours instead of 48 hours. The proposal did not find any objection from the TIGGE WG. As regards the formal aspects, a letter that should be send by the LAM EPS producers to the archiving centres. This letter will be based on the analogous letter written by the TIGGE data providers. Philippe will provide the template and Tiziana will circulate it within the Panel for comments and contributions.
2.Archiving of TIGGE LAM output parameters: 2nd phase
As mentioned in the previous section, there is a general consensus about going toward the organization of a decentralized TIGGE LAM archiving system hosted by some Regional Centres. Due to the many common aspects of this system with the future TIGGE Phase 2 archive, Tiziana proposed to establish a sub-group focused the development, implementation and management of the Distributed Archives.
This group should work together with the people in charge of the archiving in TIGGE.
2.1.Ongoing steps and outcome of the Pretoria meeting
The TIGGE WG is in favour of establishing this Focus Groupon archiving & user web interface. This focus group will start exchanging info and material by email. Tiziana proposed that they will meet together in a specific workshop in concomitance with the first plenary TIGGE LAM meeting to be organized, hopefully, during autumn this year (2008). See the paragraph below on the 1st TIGGE LAM meeting.Tiziana will contact the Panelto form this Focus Group with experts coming from the Panel itself or suggested by the Panel Members; this group will involve also experts nominated by the TIGGE WG.
During the workshop in Pretoria Tiziana showed few slides about ARKIMET (see Appendix 2), the archiving system under development at ARPA-SIM. This system, thought as a simple system but fitting some of the requirements of TIGGE LAM, is based on Free Software with GPL license and could be easily implemented in the regional centres which are not provided with another archiving system.
TIGGE LAM needs to find some volunteers to host the regional archives but it is difficult somebody can accept without knowing what this implies. This is why it is really important to speed up the work of the Focus Group to quantify the workload.
CPTEC volunteered to become regional archiving Centre and the SA Weather Service will evaluate this possibility when the workload will be better defined.
The Panelwill identify other centres with the support of the THORPEX Regional Committees.
3.Initial and Boundary Conditions
The concept of GIFS (Global Interactive Forecasting System) is a perfect framework to implement a system where each LAM could be coupled with each Global Model allowing the maximum flexibility both in the research phase and in the operational implementations.It is clear that this vision must deal with the many problems related to the huge amount of coordinated work to be devoted to this task.During this year of discussions two options were evaluated about the possible formats to provide initial and boundary conditions:
- Initial and Boundary conditions provided on a standard 3D grid:
on a lat-lon grid at horizontal resolutions adequate to the original model resolution;
on pressure levels with a number of vertical levels high enough to minimize the loss of information;
in terms of standard and pre-defined physical variables.
Advantages:
it is clear, well-defined;
stable in time;
LAM groups need to make major changes in the model pre-processing only once; minor conversions needed.
Drawbacks:
loss in accuracy due to double vertical interpolation
- Initial and Boundary conditions provided on the global models original “computational grid” (in the vertical).(note: here I assume that we all agree about going through horizontal interpolation)
Advantages:
Negligible work from the Global Models side
No loss of information due to double vertical interpolation
Drawbacks:
a lot of work must be dome by the LAM side to deal with all the different global model vertical coordinate systems.
In this case the development of a TIGGE COUPLER SW should be the more comprehensive way to meet the different requirements. Apart from the human resources necessary to develop this COUPLER, another critical aspect is clearly associated to the maintenance of this SW due to the necessity to update it every time each global or limited area model is changed.
A third option was also depicted as an expansion of option B where each of the providing centres could be asked to develop a Toolbox to interpolate from their original model grid to one or more standard formats.
At the end of last year it was decided to postpone the discussion after the exchange of some opinions with the group involved in the preparation of the SRNWP project on Interoperability leaded by UK Met Office and discussed during a specific workshop hosted at ECMWF in January this year (14 and 15 January 2008).
3.1.Link with the SRNWP Interoperability project
The Interoperability project was presented as one of three projects toaccompany the new SRNWP Programme(started in January 2008) with a dedicated financing. The other two projects are:
European multi-model LAM EPS
Model Verification
The Interoperability project was approved and will be funded starting this year. The other two projects are in “stand-by” also considering that the Interoperability project will develop the tools required by these other two projects. The MET OFFICE organized the workshop at ECMWF to discuss some aspect of the project which is now under rewriting to be presented and approved in few months.
The original programme proposal identified 6 deliverables, based on consideration of the following European NWP systems – Global: Arpege, IFS, UM, GMEand 4 LAM model systems: UM, HIRLAM, COSMO, ALADIN:
D1: A report documenting the standard output format and including a list of parameters for which the standard output format is applied.
D2: A report documenting the standard observational data format.
D3: Requirements and Specifications for the adaptor software:
This document includes the identification of the methods that can be used for implementing the adaptors, and for maintenance of the software in connection with the consortia. It must be agreed by all groups involved.
D4: Four adaptors that transform the output from every LAM to the standard output format. This includes the software as well as the documentation.
D5: Enhancements to existing software tools, that enable all LAMs to process data from the four available GMs. This includes the software as well as the documentation.
D6: Enhancements to existing software tools, that enable all LAMs to process data from the other LAMs. This includes the software as well as the documentation.
During the ECMWF workshop it has been decided to concentrate (in the new proposal) the efforts on tasks that can be fulfilled in the two year time-frame e.g. those associated to D4. More complex tasks like those associated to D5 and D6 (during the workshop Tiziana pointed out the importance of these two deliverable in the TIGGE LAM context) will be at least addressed to reach consensus on the specifics of the dataset to provide initial and boundary conditions.
The option that got quite a general consensus ( but of course everything is under reshuffling by UKMO during the finalization of the project proposal) was the definition of a standard output for the global models but keeping all the original information in the vertical and so avoiding detrimental effects due to double interpolation.
3.2.Ongoing steps and outcome of the Pretoria meeting
Referring to the ICs and BCs issue, Tiziana presented to the TIGGE WG what is reported in the preceding two sections stressing the importance of a strong cooperation with the Interoperability project. This is very important also considering that many of the involved centres are playing on both sides.
It is clear we have a problem of time schedules: the Interoperability project will use the first two years to define the standards. This means that products will be available in about four years from now (this is a subjective extrapolation). This time-range is critical for TIGGE.LAM.
The outcome of the Pretoria meeting is:
- We need to clarify how many groups are in favour to have available Initial and Boundary conditions on many pressure levels even being aware of the possible detrimental effect due to the required double-interpolation. During the TIGGE meeting in Pretoria we had two group asking for this option: CPTEC and WeatherSA.
- In the meantime we will be in close contact with the SRNWP Interoperability Programme trying to converge as faster as possible to the definition of a common format that can be acceptable also from the TIGGE partners outside Europe.
Tiziana will contact again the different LAM EPS group to have a precise and clear picture of the different positions. She will also contact UKMO and the SRNWP Expert Team on System Aspects to inform them about the outcome of the discussion and to confirm the willingness of. TIGGE LAM to be in contact with them and to contribute as much as possible to the technical discussion.
3.3. ICs & BCs availability and archiving
Preliminary proposal (coming out from the meeting in Spain):
- Temporary archive of daily “real time” data.
These data should be stored at the producing centres for a very short time from 3 to 10 days depending on the restrictions/possibilities of each producing centre.
- Permanent archive on selected periods (e.g. during specific projects, Forecast Demonstration Projects, Research Demontration Projects, etc.etc.).
These data should be stored at the producing centres. Another option could be to include these archives in the overall TIGGE archive centres.
- Permanent archive for Special Cases related to severe events.
As above. These data should be stored at the producing centres. Another option could be to include these archives in the overall TIGGE archive centres.
This proposal seems to reach a good general consensus; of course the physical implementation of these archives is secondary to the definition of the IC/BCs format.
3.4.Initial and boundary conditions related to TIGGE Phase 1: Dataset for research.(some set of archived fields available during selected specific periods).
From the discussion in Pretoria it came out that some groups could be interested to ICsBCs dataset during specific periods (around 30 days per year) in the same format of the TIGGE Phase 1 data: Pressure Levels (increased up to about 40 levels and 1° horizontal resolution).
If there are groups interested to this possibility, this proposal will be evaluated by each Global Data Provider.These datasets could be used by scientists to evaluate the impact of extracting Boundary Conditions through a double-interpolation procedure and, looking at the results, some useful guidelines could be available to better focus the final request of the IC and BCs dataset.
Tiziana will contact the different LAM EPS group about this point.
4.Relocation of LAM EPS systems and GIFS
The concept of “Relocatable Systems ” was revisited and re-modulated during the discussion within the TIGGE LAM Panel. It was pointed out the importance of employing modelling systems which have been scientifically tested in the different target regions with the associated severe events. It is important that the steering group ensures to support society with good science. It must be avoided to rush too quickly to operational implementations not supported by the necessary experimentations.
During the meeting in Spain, the discussion indicated the solution of selecting Specific Relocatable Systems to be implemented on Specific Relocation Regions. Different regional ensemble centres might act like WMO RSMCs (Regional Specialised Meteorological Centres) for different areas of the world.In the first phase of TIGGE-LAM these implementations will be fully supported by the Regional Specialized Centre; no investment from the local Institutions, which are supposed to be users of such systems, will be asked or required.