Extending ECMWF's Forecast Postprocessing Options by Offering a New Python Framework

For over 30 years, ECMWF has provided tools and software libraries to allow users easy access to ECMWF forecasts, enabling fast post-processing and visualisation of the data. Users have the choice to use specific libraries or to use Metview, a workstation/batch system that combines the functionality of all these lower-level libraries in a high level interface and a set of tools that work together. Metview's unique strength is its close connection to the MARS archive system, allowing easy access to its over 200PB of data.

In the past year ECMWF worked closely with third-parties to offer the strengths of Metview through a new Python package. The work sees a merger of the high-level Macro language of Metview (already very similar to Python) with many existing technologies from the Python community. The new package centres around the standard Python packages xarray and pandas, and work has been done to extend the xarray package to map GRIB data into the CF data model. The new python package has been successfully tested within the new Copernicus Climate Data Store (CDS) and is able to extend the CDS Toolbox, also implemented in Python.

The presentation/poster will give an overview of the new possibilities this Python interface offers to users, and the work necessary to achieve this, including contributions to third-party packages which will benefit the wider MetOcean community.