Announcements

9/May/2022

Reminder of impacts of ECMWF's Data Handling System (DHS) move to Bologna

As work to move ECMWF's DHS to Bologna progresses, the dates when users of C3S seasonal forecast data might experience disruptions are approaching.

• Two short dry-run tests successfully completed in April have paved the way for another, longer reduced service period in the coming weeks, lasting for 5 to 7 days; this will then be followed by the 4-week-long reduced service period during the actual move of the DHS infrastructure. The official announcements about the dates of these interruptions can be found in this link. This information will be updated as the work progresses so it is highly recommended you "watch" that page (see "Watch" button in the top right corner of the page).
• The expected impacts on the C3S seasonal forecast datasets are as follows:
  • Data available during both future reduced service periods: real-time forecasts with start dates after March 2022 from all providers and hindcast data from MetOffice system=601 (produced on-the-fly) with start dates from April.
  • Availability of other data is NOT guaranteed.
    • Please note this includes hindcast data relevant to the start dates around these periods of reduced service, from providers other than the MetOffice. The recommendation for users is to ensure they have downloaded all hindcast data they may need well in advance of the start of these periods of reduced service.

11/ February/2022

• As it happens every year, label "system" for the UK Met Office contribution to C3S seasonal forecast activity will be updated to a new value (system=601) starting from March 2022 real-time forecasts.
  • Please note that due to the production of hindcasts "on-the-fly", the CDS keyword 'system' does not have the same behaviour as for the other contributors. More info in the section "Versioning of forecasting systems contributing to C3S" of the Description of the C3S seasonal multi-system page.

3/January/2022

Changes to operational forecasting systems from Japan Meteorological Agency (JMA)

• A little later than previously announced,  CPS3 hindcasts for February start dates are now available at the CDS.
  • Description of this new forecasting system can be found here.
• Starting with February 2022 real-time forecast, JMA's contribution will be be from CPS3 forecasting system, while the contribution from CPS2 forecasting system will be discontinued.
• In this new system the horizontal resolution of the JMA data at C3S is upgraded to 1.25 x 1.25 degrees

  • NOTE: This is a different grid compared to the rest of C3S forecasting systems. You can retrieve a version of the data interpolated to the C3S 1-degree grid using the following CDS API keywords in your retrieval:

    CDS API retrieval keywords for interpolation to C3S 1-degree grid

    "grid" : "1/1",
    "area" : "89.5/0.5/-89.5/359.5",

8/December/2021

Changes to operational forecasting systems from Environment and Climate Change Canada (ECCC)

• As previously announced,  GEM5-NEMO hindcasts for December start date are now available at the CDS.
  • NOTE: this hindcast data will be only available via CDS-API until the official release of December 2021 forecast (12 UTC on the 13th December). See details and updates here.
  • Description of this new forecasting system can be found here.
• Starting with the December 2021 real-time forecast, ECCC contribution will be changed to CanSIPSv2.1 composed of CanCM4i and GEM5-NEMO forecasting systems. More details here

15/November/2021

Planned changes to operational forecasting systems:

• Environment and Climate Change Canada (ECCC) contribution to C3S will update one of its seasonal forecasting system to a new version, GEM5-NEMO, from December 2021
  • Hindcast data for December start dates from this new system will be available at the CDS during the week commencing on the 6th of December.
  • Real-time forecast data for December 2021 will be published on 13th December at 12GMT.
  • With the introduction of the new system, GEM5-NEMO, the contribution from the current GEM-NEMO system will be discontinued
• Japan Meteorological Agency (JMA) contribution to C3S will update their forecasting system to a new version, CPS3, from February 2022
  • Hindcast data for February start dates from this new system will be available at the CDS on January 2022 (date to be confirmed). 
  • Real-time forecast data for February 2022 will be published on 13th February at 12GMT
  • With the introduction of the new system, CPS3, the contribution from the current CPS2 system will be discontinued.

12/October/2021

An issue with sporadic anomalous values in "2m dewpoint temperature" from Météo-France System8 has been identified. The details can be found in the section "Fixed errors" under the code E9 of our "Known issues" page

• It only affects hindcast data from member 3 for October 2009, both high frequency data (seasonal-original-single-levels) and monthly statistics (seasonal-monthly-single-levels).
  • A very reduced amount of grid points and time steps present erroneous values of 1.e+20
• The data described might be affected by this issue if retrieved after its publication on 16th September 2021 and before the completion of the re-archival of the amended version (see details in "Known issues" page)

29/September/2021

An error has been identified in the "total cloud cover" data from JMA CPS2 forecasting system. The details can be found in the section "Not fixed errors" under the code E8 of our "Known issues" page

It affects all available data for that variable (daily data, monthly data and anomalies) both for real-time forecasts and hindcast dates.

Due to a bug during the data ingestion to the C3S archive the values of this parameter have been divided by 100.

6/July/2021

Changes to operational forecasting systems:

• As previously announced, Météo-France system 8 hindcasts for July start date are already available at the CDS.
  • NOTE: this hindcast data will be only available via CDS-API until the official release of July 2021 forecast (12 UTC on the 13th July). See details and updates here.
• Description of this new forecasting system can be found here.

23/May/2021

Changes to operational forecasting systems:

• Météo-France contribution to C3S will come from their new seasonal forecasting system, System8, from July 2021
  • Hindcast data for July start dates from this new system will be available at the CDS during the first week of July.
  • Real-time forecast data for July 2021 will be published on 13th July at 12GMT

13/May/2021

Changes to operational forecasting systems:

• Seasonal forecasts from Environment and Climate Change Canada (ECCC) 's prediction system, CanSIPSv2, are included in the C3S multi-system from May 2021.
  • This contribution is composed of two different forecasting systems: GEM-NEMO and CanCM4i. More details here.
    • Data for all CDS datasets (daily/subdaily, monthly and anomalies) will be made available independently for those two forecasting systems.
    • Graphical products labelled ECCC will be created as a multi-system combination of these two individual forecasting systems.
    • Graphical products for the C3S multi-system: The two ECCC forecasting systems are first combined and their combination is then treated like each one of the other contributing forecasting centres (so that each centre contributes to the C3S multi-system with equal weight).
  • Hindcast data for May start dates from these new systems is already available at the CDS.
  • Real-time forecast data for May 2021 will be published on 13th May at 12GMT.

11/Mar/2021

A data gap in MetOffice's GloSea6 forecast data has been reported as "Data gap G5" in C3S Seasonal Forecast known issues

• No data is available from MetOffice GloSea6 (system=600) for the following start dates: 22 and 23 February 2021
• An issue with assimilation of sea-ice data was reported by the MetOffice affecting forecasts initialised on 22 and 23 February 2021. Due to the potential impact of that wrong sea-ice initialisation throughout the forecast, those start dates haven't been processed by C3S.

2/Mar/2021

The integration into C3S seasonal forecast multi-system of ECCC forecasting systems, CanCM4i and GEM-NEMO, announced earlier (19th/January/2021) to happen with the release of March 2021 real-time forecasts has been delayed to April 2021.

16/Feb/2021

Changes to operational forecasting systems:

• MetOffice contribution to C3S will come from their new seasonal forecasting system, GloSea6, from March 2021
  • Hindcast data for March start dates from this new system is already available at the CDS.
  • Real-time forecast data for March 2021 will be published on 13th March at 12GMT
  • A change in the labelling of 'system' versions for the MetOffice has been implemented. GloSea6 data will be initially labelled as system=600. More details here.

19/Jan/2021

Upcoming changes in the C3S seasonal forecast service planned to be implemented for the release of the March 2021 real-time forecast (13th March at 12GMT):

• A new forecasting system will be introduced from the MetOffice contribution to C3S, GloSea6. The on-the-fly hindcasts of the current system GloSea5-GC2-LI (GRIB system keyword=15) for the February nominal start date are the last ones to be published from that forecasting system.
• A new contribution from ECCC (Environment and Climate Change Canada) will be added consisting of their two forecasting systems CanCM4i and GEM-NEMO

More details about these planned changes will be soon notified. Please note this is just an early warning about the plans for implementation of those contributions and changes to that schedule are still possible.

14/Dec/2020

With the publication of the December 2020 C3S seasonal forecasts (released on the 13th December at 12UTC) a couple of new additions have been made to the catalogue of graphical products shown at https://climate.copernicus.eu/charts/c3s_seasonal/:

• Time series of seasonal forecasts and their corresponding hindcast for the zonally averaged u-component of wind at 10hPa
  

• Evolution of probabilities for quartile-bounded categories (Percentiles 25 and 75) of the area-averaged monthly-mean sea-surface temperature anomaly computed over specified regions of the Tropical Pacific (the NINO 1+2, 3, 3.4 and 4 areas) 
  

26/Oct/2020

An issue in the CDS infrastructure affecting some seasonal forecast retrievals has been identified. It will be tracked as error (E7b) in the seasonal forecast known issues page.

This error prevents users from getting all available members from CFSv2-v20110310 (NCEP CFSv2) data. This issue affects retrievals done after the release of the October 2020 forecast, which happened at 12UTC on the 13th of October 2020.

It affects all data available in the high-frequency (daily/subdaily) datasets:

• seasonal-original-single-levels
• seasonal-original-pressure-levels

Members with start time at 06h,12h and 18h are not being retrieved from the archive by the CDS and just members with start time at 00h are returned.

This is currently under investigation and it will be announced here once the issue has been resolved.

13/Oct/2020

With the release of the October 2020 real-time forecast (at 12 UTC on 13th October) some additions will be introduced to the list of available forecasting systems. Specifically:

• Contribution from JMA CPS2 system will be included
• Upgrade of CMCC contribution, to the new operational seasonal forecasting system SPSv3.5 (system=35)

Additionally, with the release of the November 2020 real-time forecast (at 12 UTC on 13th November), another change of forecasting systems will be included. Specifically:

• Upgrade of DWD contribution to the new operational forecasting system GCFS2.1 (system=21)

Details about hindcast dates available can be found here. The complete list of parameters available for each forecasting system can be found here.

13/June/2020

• With the release of June 2020 real-time forecast (at 12UTC on the 13th June) some changes have been introduced to the geographical regions available at C3S seasonal forecast graphical products (https://climate.copernicus.eu/charts/c3s_seasonal). Specifically:
  • A new area called "Pacific Islands" has been added to cover in detail that region of the world
  • The area called "South America" has been amended, so it accurately covers the whole South American sub-continent.
  • The area called "Tropics" has been amended to be centred over the International Date Line (meridian 180º)

All these changes have been introduced starting from June 2020, and previously produced plots haven't been affected by them.

15/May/2020

• The issue affecting availability of NCEP CFSv2 members in the high-frequency (daily/subdaily) datasets reported as "known issue" E7 (C3S Seasonal Forecast known issues) has been now fixed and all NCEP CFSv2 members are now available for users to retrieve.

12/ May/2020

• As it happens every year, label "system" for the UK Met Office contribution to C3S seasonal forecast activity has been updated to a new value (system=15) starting from May 2020 real-time forecasts
  • Please note that due to the production of hindcasts "on-the-fly", the CDS keyword 'system' does not have the same meaning as for the other contributors. Instead, it is just an indexing label that gets changed once per year.
  • More info in the "Summary of available data" and in the "Seasonal forecasts and the Copernicus Climate Change Service" pages

25/March/2020

• An error in the retrieval of high-frequency data (daily/subdaily) from NCEP seasonal forecast data (NCEP CFSv2) data has been detected. This has been documented as "known issue" E7 (C3S Seasonal Forecast known issues)
  

  It affects all data available in the high-frequency (daily/subdaily) datasets:

  • Seasonal forecast daily data on single levels (seasonal-original-single-levels)

  • Seasonal forecast daily data on pressure levels (seasonal-original-pressure-levels)

  Members with start time at 06h,12h and 18h are not being retrieved from the archive by the CDS and just members with start time at 00h are returned.

11/February/2020

• An error in the processing of surface solar radiation variables for NCEP has been reported (Known issue E6 in the "not fixed issues" section here)
  • The parameters "surface net solar radiation" and "surface solar radiation downwards" (and their counterparts in the monthly statistics and anomalies datasets) have been accidentally swapped with each other in the case of
    
    • Forecasts with nominal start dates from October 2019 to February 2020 (both included)
    • Hindcasts with nominal start dates October, November, December, January and February
  • A fix has been already introduced in the operational production of C3S data; starting from nominal start date March 2020 real-time forecast data (and its corresponding hindcasts) will not be affected by this.
  • A recovery process for the affected data already published will be put in place. Details about when this will become available will be announced here.

8/January/2020

• The recovery process for missing monthly and anomalies data for NCEP contribution has been now finished. More details in known issues E4.a1 and E4.a2 ("fixed issues" section here).

13/December/2019

• With the release of December 2019 C3S seasonal forecasts a new encoding for monthly data has been implemented allowing users to get the real initialization date of each individual ensemble member from lagged start forecasting systems.
  More information here.
  Related information about known issues here (items E4.a1 and E4.a2)

2/December/2019

• A data issue has been recently reported affecting CMCC contribution data for start dates between March and September. The impact of those errors is quite limited in general, but it can be noticeable for "geopotential height" variable in pressure levels.
  More details can be found in known issues E5a and E5b here.

29/November/2019

• The recovery process for wrong snowfall data from CMCC contribution has been now finished. More details in known issues E3a, E3b, E3c and E3d ("fixed issues" section here)

7/November/2019

• The recovery process for the wrong snowfall data from CMCC contribution has started for the November start date hindcasts (i.e. 1993 to 2016). More details can be found in known issues E3.a ("not fixed errors" section here)

11/October/2019

• All the relevant documentation for the new forecasting systems to be released with October 2019 real-time forecasts (Météo-France System7 and NCEP CFSv2) is now available.
• Due to technical reasons, the publication of monthly mean data for NCEP has been delayed, but the full set of high-frequency data will be available as expected in the CDS. This has been recorded as known issue E4 (see details here in the "not fixed errors" section)
  
  • As a side effect, publication of updated MetOffice hindcasts for November nominal start date (system=14) is also delayed.
  • The current estimate is that all the above-mentioned delayed data would be made available on the week commencing on 4th November.

22/August/2019

• An upgrade to Météo-France contribution (change of operational forecasting system from System 6 to System 7) will be introduced from October 2019 nominal start date. Documentation for the new system here.
• A new data stream will be added to the C3S seasonal forecast multi-system in October 2019. This data comes from NCEP's Climate Forecast System version 2 (CFSv2). Documentation for this system here.

5/April/2019

• An upgrade to MetOffice contribution (change of forecasting system version to GloSea5-GC2-LI) has been introduced from May 2019 nominal start date.