DIgital_slice_image

ECMWF helps users to drive a regional chemistry model with CAMS data

Bojan Kašić, Xiaobo Yang, Miha Razinger, Milana Vučković

 

The Copernicus Atmosphere Monitoring Service (CAMS) produces daily forecasts of pollutants, aerosols and greenhouse gases across the globe. This makes it an ideal candidate for driving a regional chemistry model such as WRF-Chem, the US Weather Research and Forecasting (WRF) model coupled with chemistry. In view of the high level of interest among CAMS users in running the WRF-Chem model with CAMS data, including from universities in Britain and Ireland, ECMWF has carried out some tests and has shared the insights gained with the CAMS user community on the Copernicus Knowledge Base (http://copernicus-support.ecmwf.int/knowledgebase).

%3Cstrong%3EWRF-Chem%20example%20plot.%3C/strong%3E%20Initial%20nitric%20acid%20(HNO%3Csub%3E3%3C/sub%3E)%20concentration%20on%20the%20WRF-Chem%20grid%20at%20model%20level%205,%20interpolated%20from%20CAMS%20data%20and%20plotted%20in%20Metview.
WRF-Chem example plot. Initial nitric acid (HNO3) concentration on the WRF-Chem grid at model level 5, interpolated from CAMS data and plotted in Metview.

To drive the WRF-Chem model, the first step is to create initial and lateral boundary conditions from a global meteorological dataset. ECMWF has tested this with ERA5, the latest global weather and climate reanalysis from the Copernicus Climate Change Service (C3S). Surface and pressure level meteorological fields, which are mandatory for WRF initial and lateral boundary conditions, were retrieved from the C3S Climate Data Store (CDS) using the CDS API.

Next, the initial and lateral boundary chemistry conditions are needed to drive the chemistry part of the model. ECMWF’s Web API was used to retrieve CAMS chemical data on model levels, including aerosols, global reactive gases, volatile organic compounds, temperature and surface pressure. However, currently WRF cannot directly process chemistry data from global models and consequently additional tools are required to interpolate chemical variables to the WRF model grid.

WRF-Chem driven by CAMS data has previously been run as part of the Air Quality Model Evaluation International Initiative (AQMEII) phase 2 project, for which ECMWF provided boundary conditions. To pre-process the CAMS data for WRF-Chem, this group developed a set of utility tools to convert CAMS output into the format used in the MOZART chemical transport model. The same utility tools are available to CAMS users as the group has kindly agreed to share them with our user community. This means that the MOZART package, which is freely available from the US National Center for Atmospheric Research (NCAR), can be used to ingest the CAMS chemical data into the WRF-Chem boundary condition files. Once the WRF-Chem simulation is complete, users have the option to directly visualise the output in NetCDF format using ECMWF’s Metview package.

There is thus a complete set of tools available to drive WRF-Chem simulations using CAMS data as initial and lateral boundary conditions. It is expected that a more sophisticated and fully automatic software package will be made available in the future. This work in response to user requirements is part of ECMWF’s efforts to make sure that CAMS is a user-driven service.