Having made a first set of data available in early 2022, ECMWF has now released a much larger open data set to the public, representing weather forecasts at a higher resolution and with a reduction in release times.
For example, medium-range data of the integrated forecasting system (IFS) is now provided at a resolution of 0.25° x 0.25° (28 x 28km), compared to the 0.4° x 0.4° provided before.
This is the resolution at which the Artificial Intelligence/Integrated Forecasting System (AIFS) is currently run. More information on the AIFS, including a link to access the data, is available on the AIFS machine learning data page. “We’re excited about the next step for the AIFS on the road to a full operational system, giving users data to test and use,” commented Mat Chantry, machine learning coordinator at ECMWF.
Alongside this, medium-range data of the latest beta version of the AIFS is also available at a resolution of 0.25° x 0.25°. More parameters will be available in the coming days, based on user feedback. These will enable limited-area modeling and machine learning initialization. Documentation can be found on the open data page for real-time forecasts.
“This change means open access to IFS medium-range forecast data at a much better resolution than before,” said Umberto Modigliani, deputy director of the forecasts and services department of the ECMWF.
Furthermore, ECMWF’s contribution to the subseasonal to seasonal (S2S) data set, which is part of a global scheme, is now available with a 48-hour delay rather than the previous three-week delay.
ECMWF provides daily global forecasts of atmospheric composition as part of its Copernicus Atmosphere Monitoring Service (CAMS) portfolio. These forecasts include all meteorological variables at 0.4° x 0.4°. Data for pressure levels and some surface fields of these forecasts were previously only available after a six-day embargo. They are now available in real-time, as they are for atmospheric composition products.
“Atmospheric composition and weather are inherently integrated, with interactions going in both directions. Having all meteorological fields available in the CAMS global forecasts supports a better and more comprehensive use of the data by many CAMS users,” stated CAMS deputy director Richard Engelen.
The delay on subseasonal to seasonal forecasts provided by ECMWF under a prediction project run by the World Weather Research Programme (WWRP) and the World Climate Research Programme (WCRP) has also been greatly reduced. ECMWF is one of many forecasting centers contributing to the project. Forecasts are now released with only a 48-hour delay, rather than the previous three-week delay. All S2S data providers are being encouraged to reduce the delay for their contribution.
“The reason this is important is that it will strongly increase the usefulness of the S2S database, making it possible to assess the value of the S2S forecasts near real time,” said ECMWF scientist Frédéric Vitart.
By increasing the amounts of open data at higher resolutions, ECMWF intends to provide societal benefit as these data sets satisfy the World Meteorological Organization Unified Data Policy core requirements, as well as making more data available for use in initiatives like Early Warnings for All (EW4All).
“Altogether, this is a significant upgrade of ECMWF’s open data offering, which marks an important milestone in our move towards open data,” said ECMWF data policy and partnership coordinator Emma Pidduck.
To find out more about ECMWF’s latest developments, click here.