We are pleased to announce the implementation of IFS Cycle 48r1, the first new scientific model upgrade in ECWMF's Data Centre in Bologna, for June 2023, with a three-month Release Candidate Phase starting in March. Precise dates will be given in due course.

As highlighted in a recent news article, Cycle 48r1 delivers a significant upgrade. Many of the scientific improvements are related to the enhancement of the ensemble forecasts, thus making them our main output in the medium range. Major technical and configuration changes relevant to end-users include:

  • Increase of the horizontal resolution of the medium-range ensemble from 18 km to 9 km
  • Daily runs of the extended-range ensemble as a separate system from 0 UTC to day 46 with 101 members
  • Two sets of re-forecasts, one for the medium range and one for the extended range, each running twice per week, on Monday and Thursday
  • Introduction of the Golomb-Rice coding compression method, as defined in the CCSDS recommended standard and implemented by the Adaptive Entropy Coding Library (AEC), as the default packing type for all gridded model output in GRIB 2 format

As an introduction to the new model cycle we have scheduled two webinars:

On 2 March, 9:00 UTC, Andy Brown, Director of Research, will introduce Cycle 48r1 and its performance (register here).

On 15 March, 15:00 UTC, Florian Pappenberger, Director of Forecasts, and Matthieu Chevallier, Head of Evaluation, will present the latest model verification, new products, and technical aspects like the access to test data and the handling of increased ensemble data volumes (register here).

All users of ECMWF forecast products are invited to take note of the details provided in the Cycle 48r1 Implementation page and test their processing chains as early as possible with the test data provided. To get notified of further updates please “watch” the implementation page. 

Please report any issues with this model upgrade and your tests via the ECMWF Support Portal.

ECMWF has today discovered an error that affects 2m temperature values depicted on its standard (web-based) 10-day meteograms ("10-day ENSgrams"). The "15-day ENSgrams" and "15-day ENSgrams with climate" are unaffected.

The error is as follows:

To derive 2m temperature values to plot on its meteograms ECMWF applies a height-based correction to model 2m temperature output, that is based on a fixed lapse rate (-6.5K/km) assumption. In short we compute the altitude difference between the site in question, and the nearby model orography, and then use this as a multiplier for the lapse rate adjustment. Due to different horizontal resolutions, the HRES orography dataset differs from the ENS orography dataset. The error is such that, for the 10-day meteograms, these orography datasets were accidentally switched around. This happened when we moved our operations to Bologna from Reading, on 18 October 2022. In relatively flat areas the impact on displayed 2m temperature is negligible, but in very mountainous regions where HRES and ENS orography heights generally differ by a large amount it can be substantial. There are further potential complications at coastal sites.

This error has been addressed quickly, and with effect from the 12UTC run of 4 January 2022 all meteogram 2m temperatures will be correctly computed.

 We sincerely apologise to all ECMWF users for the inconvenience that the error will have caused.

ECMWF is running an online survey that relates to future changes to the structure of its re-forecasts. This will help shape our re-forecast strategy.
If you use or have an interest in our re-forecasts, we strongly encourage you to complete the survey at the following link - it should not take more than 10 minutes:
The survey will close at 00UTC Monday 12th December.
Feel free to notify the above to interested parties within your organisation - i.e. anyone that works directly with ECMWF re-forecasts (for research, for forecasting or for other purposes).  
Thank you very much!
Forecast Department

ECMWF is providing some guidance on SYNOP surface pressure observation issues based on experience from the SEEMHEWS-A (South-East European Multi-Hazard Early Warning Advisory System) project. In this project, additional SYNOPs needed to be monitored prior to their inclusion in the set of in-situ observations processed at ECMWF. Within the context of Non-EUMETNET and Non-WMO NHMS (National Meteorological and Hydrological Service) observations, this required quality checks and guidance on how to improve SYNOP observations rejected by ECMWF IFS model. It was decided to create a troubleshooting page to support resolution of issues in such cases in which close collaboration between the SYNOP observation data provider and the ECMWF observation monitoring team is needed.

The new troubleshooting page can be accessed here.

ecPoint output upgraded

Beginning with the 12UTC runs on 23rd May 2022 the way in which ECMWF computes post-processed ecPoint output, for 12h rainfall, changed in several ways. The skill and reliability of the ecPoint forecasts should improve as a result. For more information on this upgrade, and some examples, see here.

Users of ECMWF forecast products are reminded that there will be some noteworthy changes in forecast behaviour when cycle 47r3 goes live on 12th October. Please check this web page for details: 47r3 Impact on Surface Weather representation. It has been updated today (11th October) with more information.

We are pleased to confirm the operational implementation of IFS Cycle 47r3 with the   

   06 UTC run, 12 October 2021

The upgrade will bring improvements to the assimilation and observations usage and a significantly improved physical basis for moist processes, necessary to facilitate further development of the Integrated Forecasting System (IFS) and future application at convection-permitting resolutions. The release candidate phase has been running in near real-time since 14 September. In a recent Webinar we explained the scientific background, its meteorological impact, and some new and revised products. Forecast users are recommended to read "47r3 Impact on Surface Weather representation", highlighting the changes with 47r3 in the characteristics of many frequently used forecast fields and products.

We would also like to remind you of a change to the “Master Tables Version Number”, affecting all parameters in GRIB 2, and encourage you to check your data processing with the test data provided as soon as possible ahead of implementation.

Detailed information can be found on the 47r3 implementation page. Please “watch” the page to get notified about further updates.      

The new IFS Cycle 47r3, expected to be the last upgrade before ECMWF's new data centre in Bologna will become operational, is scheduled to be implemented on 12 October 2021.

IFS Cycle 47r3 will bring improvements to the assimilation and observations usage and a significantly improved physical basis for moist processes, necessary to facilitate further development of the Integrated Forecasting System (IFS) and future application at convection-permitting resolutions.

The upgrade will bring more accurate upper air fields, particularly tropical winds, as well as precipitation in convective regimes. Several new convection-related products will be available and systematic errors are reduced for wind gusts and visibility for fog and precipitation

In a first webinar on  

    6 July 2021 at 8:00 UTC: Join the webinar on Zoom

Andy Brown, Director of Research, will introduce the new cycle.  

For more information and to keep up to date with further progress please follow and “watch” the 47r3 implementation page.    

On Tuesday 18th May ECMWF will release to users its new-look cyclone database products (products that feature front and cyclone locations and cyclone track information), following a collaborative initiative involving Met Office scientists. The changes are explained on this page in the Forecast User Guide.

The new IFS Cycle 47r2 has been successfully implemented with the 06 UTC run on 11 May 2021. The key improvements introduced with this upgrade, the change to single precision in the HRES and ENS systems and the increased vertical resolution of ENS, are summarised in Forecast upgrade innovates on single precision and ensemble resolution. Furthermore, new Tropical Cyclone tracks from the 06/18 UTC forecast cycles are also available as WMO essential products, including graphical products. More details can be found on the implementation page for IFS Cycle 47r2.

We hope you will benefit from our enhanced forecast products. Please let us have any feedback via our Support Portal.

The new Cycle 47r2 test system has now been running as release candidate in near real-time since 9th February, on track for implementation on 11 May 2021. Cycle 47r2 layers in ecCharts and ENS meteograms will become available on 6 April. The recommended versions of ECMWF software packages to handle 47r2 data have now also been announced.

In a final webinar on  

    16th April 2021 at 8:00 UTC - join the webinar

Director of Forecasts Florian Pappenberger will present the latest model verification, new Tropical Cyclone tracks, and technical aspects like the access to test data and the handling of the increase of ENS model levels in dissemination. As usual, a recording of this webinar will be made available.

Please mark the webinar date in your calendar and keep watching the implementation page for more details and any further updates!

The new IFS model cycle which includes single precision for HRES and ENS and (with the computational savings made) an increase from 91 to 137 levels in the ENS forecast, is progressing well. Next week (9th February 2021) users will be able to access test data. A week later we will be running a Q&A session for those users who make use of ENS model level data.

Updates to the BUFR templates are included in the new cycle and we encourage users to test their decoding software. Moreover, Tropical Cyclone products from the 06 and 18 UTC cycles will be available as WMO essential.

Please visit the implementation page for details on how to access test data, the new BUFR template, the Tropical Cyclone products, model performance etc. at

   Implementation of IFS Cycle 47r2

ECMWF informed all ecCharts users on the latest update to this service.

The updates include the ecCharts sharing functionality, a number of layers for ENS and extending range cumulative distribution functions (CDF) for temperature and precipitation.

Full details of this update are available at ecCharts updates - November 2020

Director of Research Andy Brown will introduce the new IFS model cycle (47r2) and its performance in the following two webinars:

Do not forget to add one of the dates to your calendar!

Additional details on the cycle and any new updates are available at Implementation of IFS Cycle 47r2.

A further webinar is planned for April 2021 and details on how to join will be provided closer to the event.

We are working on the new model upgrade (IFS cycle 47r2), expected to be the last cycle before ECMWF's computing centre is moving to Bologna. The upgrade includes single-precision for ENS (forecast up to day 46 and hindcast) and HRES (forecast) model. Moreover, the number of ENS model levels will be increased to 137 (currently 91) to bring ENS in line with the vertical resolution of HRES. A summary of the meteorological performance will be made available in November. We have scheduled webinars to introduce the cycle, two in December and one just before the implementation, planned for May 2021.

More details are available at Implementation of IFS Cycle 47r2.