Abstract 081

Abstract ID: 081

The technical development of the S2S database

Lead Author: Richard Mladek
ECMWF, United Kingdom

Keywords: s2s, database

Abstract: Since the S2S project start in 2013 one of the key role of ECMWF has been to create and maintain a dedicated data archive of selected meteorological fields. Based on success of TIGGE, the similar key principles have been followed (common data format GRIB 2; common terminology, parameters, units model levels; common tools for data discovery and mass downloads; systematic handling of new model changes and production incidents; production of user statistics etc).

The added complexity of S2S archive comes from heterogeneity of the participating model outputs because of different reforecast methods used (fixed vs on-the-fly) or different model runs frequency. To be able to archive all required parameters, some new data format proposals had to be submitted to WMO Expert Teams (e.g. GRIB 2 templates for new soil or ocean parameters). Also many ECMWF’s tools has had to be updated (ecCodes, MARS, Web-API) to be able to handle and serve S2S data.

In January 2023, the archive size containing outputs from 12 global weather prediction models is more than 200 TiB, with 1845 active users. The amount of data delivered to users is approaching 1.4 PiB. In our presentation, among other things, we will show the most popular parameters and the most frequent topics of scientific papers based on the S2S data. All gained knowledge of user needs and the best ways how to satisfy them should help in the future developments of the S2S database.

Manuel Fuentes (ECMWF)
Frederic Vitart (ECMWF)