This dataset contains the 2000-member original validation test case for OpenIFS@home as presented in Sparrow et al 2021, GMD.

These are results from an ensemble of forecast simulations of the tropical cyclone Karl from September 2016, studied during the NAWDEX field campaign. This cyclone underwent extratropical transition and intensified in mid-latitudes to give rise to an intense jet-streak near Scotland and heavy rainfall over Norway.

The OpenIFS@Home data comprises output from the European Centre for Medium Range Weather Forecasts (ECMWF) OpenIFS model run under the BOINC framework and managed by http://climateprediction.net/ at the University of Oxford. OpenIFS is derived from ECMWF's IFS operational forecast model. The dataset consists of a 2000 member ensemble of the ECMWF OpenIFS cycle 40 release 1 model, run on the OpenIFS@home volunteer framework. The spatial resolution of the dataset is approximately 125km at the equator (corresponding to a spectral resolution of T159) with 60 vertical levels

The model output data format is WMO GRIB. Single level parameters and surface fields use GRIB-1 format whilst multi-level upper air fields use GRIB-2 format

To work with GRIB files requires the ECMWF ecCodes package. This software provides command line tools for linux/macOS for inspecting the contents of these files, such as listing parameters and meta-information

The model output files are on the native IFS grid, the so-called reduced Gaussian grid, rather than a regular latitude-longitude grid commonly used in plotting. The ECMWF metview application can be used to produce plots directly from the model output files without the need to interpolate first. Other applications exist, such as the Climate Data Operators (CDO) that can be used to first convert the GRIB model output files to netCDF for use in other plotting packages.

The data is arranged into subdirectories (one per ensemble member), each containing a zip file of the output GRIB data for that simulation. These follow the standard OpenIFS output format and represent 6 hourly snapshots throughout the forecast. The number in the filename denotes the hour of the forecast contained in each respective GRIB file