2014-03-28                                                                                                                                             Per Undén



Notes from the UERRA kick-off meeting / 1st General Assembly, Met Office, Exeter, 26-27 March 2014



      1. Introduction and general points


The UERRA web site is at www.uerra.eu and the Agenda and the slides of the presentations can be found there. Also a wiki system and a document exchange will be at www.wiki.uerra.eu.

The Coordinator, Per Undén and the Financial officer, Madeleine Benderyd, gave an overview of the project, deliverables and financial rules and the reporting.

The partners and WP leaders presented the plans in the respective Work Packages.

The Advisors, Adrian Simmons and Blaz Kurnik gave comments after the presentations. The reviewer enrolled by REA, Mikko Strahlendorff gave comments after the sessions of the respective WPs.

Comments by the Advisors:

Adrian Simmons emphasized that UERRA should make sure components that are mature are put forward for the Copernicus CC services, in spite of the project still running for 4 years.

Action: UERRA management team to discuss timescales for transition of components to Copernicus at next management telecon.

For the work of WP1 in recovering observations, one should take a wider view and not reduce effort on the first half of the century. It is essential for the global climate services and future regional RA may go further back in time.

Sensitivity to boundary conditions should be tested, even if it is considered to be relatively small. Different data sets have been discussed for the lateral boundary conditions, of differing quality and resolution.

Action: WP2 partners to consider running sensitivity studies to impact of alternative sources of LBCs e.g. ERA-SAT replacement, JRA-55, etc for deterministic, ERA-CLIM2, 20-CR for probabilistic reanalysis.

Air quality and chemistry will need to be included in future regional RA, as in the global, and coupled RA.

The WP3 plans and background from CORE-CLIMAX looked very useful.

Overall there was a good dialogue at the meeting and, even though the deliverables ae many, they will be helpful.

It is important to involve users and one can start with just a few.

Blaz Kurnik stressed the importance of supplying EEA with input of indices and input for climate services that EEA collates and summarizes. CIBs are important and to have as quickly as possible based on events. E.g. the drought indicators are important.

Mikko Strahlendorff stressed that the large investment of 3.8 billions going into Copernicus in total and that the end result of adaptation must pay off.

It is important to lift any new data which are used and relate it to GEO and WIS. Meta data should be supplied to WIS.

In UERRA the recovered observations should go into MARS to make sure that the other WP s use them and in general that WP s work together.

The downstream hydrological models to be used in UERRA give a simple evaluation but that is not enough, even if the application is useful.

      1. Points from the presentations and discussions.

 See also the presentations on uerra.eu.


 There are choices to be made how to prioritise filling gaps of observations. A survey will be made with the partners.

 Gridding of observations will be made also at different resolutions.


Met Office production runs will not begin immediately as the first two years will be dominated by the development of new capabilities not used in EURO4M; direct assimilation of six-hourly precipitation accumulations, regional ensemble-variational data assimilation, regional variational bias correction, and optimization of lateral boundary condition perturbations. It is envisaged that an early, deterministic (4DVAR) production run may be possible by the end of year 2. However, this depends on the availability of the ERA-SAT replacement as an alternative source of LBCs to the current ERA-Interim (assuming sensitivity tests indicate superior quality from the higher-resolution ERA-SAT replacement). Soil moisture can come from HADISST2. The ensemble spread inflation may vary with time due to varying observation numbers.

 Action: DB to clarify timescales for ERA-SAT replacement run with ECMWF.

An issue was found with the BUFR coding of SYNOP (cloudiness) which seems to be wrong (how the SYNOP code was interpreted). E-OBS 24h precipitation data will also be used.

 University of Bonn will develop and run a hybrid ensemble filtering and nudging system over the CORDEX 11 km domain. Coupling may come from the NCEP 20CR RA. There will be perturbed observations in the ensemble nudging step and virtual PBL observations. MARS observations will be used. There is a choice of physics and it can be discussed with WP2 and 3 on which parameters are most important.

SMHI will set up and run HARMONIE 3D-VAR with structure functions and large scale control. A 5 year period will be run with two physics packages. The area encompasses CORDEX but the projection is different.

The downscaling near surface RA will use input from SMHI HARMONIE bur passing through the ALADIN model. The uncertainty will vary a lot in space due to availability of observations. Different data bases will be used and QC is very important. One needs to consider change over time of vegetation and there are certain data sets of Leaf area index and vegetation employed in CMIP5.

The issue of getting observations from WP1 into WP2 through MARS at ECMWF was discussed and actions to follow this up will be taken in UERRA.

The evaluator pointed out the importance of the scientific validation of the RA systems. The different time scales in the WP s and the release of E-OBS data need to be considered.



All diagnostic information and methods from WP2 will be input for WP3 and they will be surveyed. Also correlations from WP1 are of interest. Then there will be focus on the common set of agreed diagnostics. E.g. obs- background and analysis statistics are obvious ones from WP2.

The question of which information is independent or not will be decided on but it is difficult. Uncertainty in the observations also needs to be included.

The hydrological discharge models have errors too and may give differing signals and this should be checked.

The user requirements from the CORE CLIMAX survey is very useful also here.

The question of which parameters from the RA can be relied on, like snow and SW radiation and soil moisture compared with satellite derived ones needs to be considered.

How well we do in extreme events will be important to check.

KNMI (and EDI) will be more involved towards the end, to synthesize the results.



The archiving of a least the most common parameters will be done in MARS at ECMWF and using current and coming developments.