R-ENS primarily focuses on delivering daily European-scale regional air quality monitoring and forecasting services. The services are based upon an ensemble of individual analyses forecasts performed at seven centres in Europe, recognised for their expertise and experience in providing routine air quality forecasts at the scale of Europe for several years. The R-ENS subproject also largely develops from the corresponding activities within the GEMS and PROMOTE projects.

 The chemical data assimilation and forecast systems used are: CHIMERE, EMEP, EURAD, LOTOS-EUROS, MATCH, MOCAGE and SILAM. Dossiers describing these systems (and the corresponding references) are available here; briefly, their current characteristics are:

The main activities of R-ENS are to operate uncer pre-operational conditions the seven individual suites, ensemble processing, as well as verification against surface observations in Europe. In addition, research is carried out, especially on:

• verification against observations and in particular issues of representativeness and effects of model resolution on skill scores;
• air quality modelling, in order to ensure continuous upgrade of the systems, taking in particular into account results from the verification as well as the subprojects R-EVA, and feedback from users;
• ensemble methods and multi-model ensemble processing, in order to deliver products of highest quality and estimate of inter-model variability as a measure of uncertainty;
• pollen and primary biogenic aerosol particle representation in Air Quality models and development of new products -this is a component of air quality that receives growing interest, given the widely recognised impacts on health and the fact that they can contribute significantly to particulate matter mass;
• data assimilation aspects are studied in the subproject R-EDA.

 The joint expertise of the R-ENS teams will also be used to provide training and teaching material for the users.

Services

All services are available on the dedicated daily regional products webpage. The page is regularly updated as products become available during the day. Ftp data services are also available; for access to numerical products, please contact the macc regional platform team indicating name, detailed affiliation and a few lines on the reasons for asking for access to the data.

The services operated daily are :

• 0-72h ensemble forecasts of key air pollutants (ozone, NO₂, PM₁₀, PM_2.5, SO₂ and CO) at the surface and in altitude (500m, 1000m and 3000m above ground level). They are in general available by 12 UTC.  Statistical products (“chemical EPSgrams”) are available later in the afternoon.
• 0-72h individual forecasts of key air pollutants (ozone, NO₂, PM₁₀, PM_2.5, SO₂ and CO) at the surface and at altitude (500m, 1000m and 3000m above ground level). Depending on the model, forecasts are available between 6 and 11 UTC.
• Hourly surface ozone ensemble analyses. They are available by 17 UTC.
• Hourly surface analyses for ozone and, depending upon the systems, other observed pollutants. They are available by 16 UTC.
• Verification of forecasts against surface sites in Europe. Different skill indicators are computed and appear on the website in the early afternoon for forecasts up to the day before.

 The hours indicated here correspond to the current pre-operational set-up. Hours for delivery will improve gradually as in particular observational data (needed for assimilation and verification) become available earlier, through on-going collaboration with the European Environment Agency (within the subproject D-INSITU). Also, it is expected to extend the forecast horizon by one day (up to +96h) in the course of 2011.

 How are services computed?

The services are built upon a set of seven state-of-the-art air quality models that share a few key commonalities :

• a dedicated European-scale (~8km) anthropogenic emissions inventory, which is developed in the context of the D-EMIS subproject;
• meteorological forcings from ECMWF weather forecasts based on 12 UTC daily (these forecasts have a horizontal resolution that matches well the one of the R-ENS chemistry and transport models);
• global chemical boundary conditions provided by the G-IDAS subproject, which operates the systems developed in G-RG and G-AER. Currently, only the boundary conditions for reactive gases are used in the daily operation of the R-ENS models; within the last year of MACC, the regional models will progressively include also aerosol boundary conditions;
• assimilation and verification using the same set of surface observations distributed in Europe (see the list of current data providers, resulting in hourly observations from around 1000 sites gathered on a daily basis).

 As described above, the individual model operations are run by a network of centres distributed in Europe. Individual model outputs are gathered and processed centrally to provide the R-ENS services. Differences seen between the outputs from the individual models thus solely come from differences in the regional assimilation and modelling system themselves (resolutions, formulation, parameterizations, …), not from differences in the key input data (emissions, meteorological forcings, chemical boundaries, observations assimilated): they really reflect the spread among models and uncertainties within leading air quality models in Europe. The same model configurations are used for the air quality assessment activities conducted within the R-EVA subproject.   

  Partners and main contributions