Description of System 4
The resolution of the atmosphere model is increased from TL159 with 62 vertical levels to TL255 with 91 vertical levels. The model top is increased from 5 hPa to 0.01 hPa.
The NEMO ocean model replaces the previous HOPE ocean model. NEMO is used in the ORCA1 configuration, which has an approximate 1 deg x 1 deg resolution with equatorial refinement. The ORCA1 grid has a somewhat complex structure that cannot yet be described in GRIB. NEMO output is in netCDF, and the netCDF files are archived in ECFS. There are no plans at present to make System 4 ocean data available in MARS.
The ensemble sizes are slightly increased. The real-time forecasts will now have 51 members (previously 41). The re-forecasts will have 15 members uniformly for all start months (previously 11 members).
The period of the re-forecasts has been extended to include the last 5 years of operational running of System 3, and covers the 30 years from 1981 to 2010. This is longer than the 1981-2005 25 year period of the previous system, and should allow slightly better calibration of the forecasts and better assessment of their skill. It also means that anomalies will be relative to a standard 30 year period.
The standard forecast length (for both forecasts and back integrations) remains at 7 months. However, the issue date of the forecasts will be brought forward from the 15th to the 8th of the month. This gives a forecast range of 6.75 months from the release date of the forecasts.
Four times a year, from the Feb, May, Aug and Nov starts, 15 members (out of 51) of the forecast ensemble will run to 13 months. This will allow an "ENSO outlook" to be given. The re-forecasts for these start months also have 15 ensemble members that extend to 13 months. Note that these extended runs remain classed as experimental rather than operational, and data from them will not be available via dissemination.
Major scientific changes for System 4
- A new ocean model (NEMO) and ocean data assimilation system (NEMOVAR). These changes give improvements to the mean state and to SST forecast skill in the East Pacific and Tropical Atlantic oceans.
- Change from IFS cycle 31r1 to 36r4, including many improvements to the model physics, and giving a generally much improved model climate (although not for equatorial Pacific winds).
- Improved representation of the stratosphere and processes such as the QBO, including volcanic aerosol and radiatively interactive ozone.
- Improved land surface model (HTESSEL) and land surface initialization methodology.
- Revision of stochastic physics, with the new SPPT3 scheme giving substantially bigger spread to the SST forecasts.
Tests show that the prediction skill for Pacific equatorial SSTs (i.e. El Niño variability) is measurably improved in S4 in the East Pacific, but fractionally degraded in the west Pacific. Prediction skill in the Atlantic is notably improved. Changes in skill for atmospheric variables are less precisely measurable with the limited test samples available. However, ACC skill measures show substantial and consistent improvement in tropical scores. NH scores (poleward of 30N) do not seem to be improved in winter, but show noticeable improvements in other seasons, based on analysis of the first 18 years of re-forecast. Full information on skill will be available prior to operational implementation.
System 4 performance
Preliminary information on the performance of System 4 is available in the presentation to the Forecast Products Users' Meeting.
More comprehensive documentation of the performance is in preparation.
Re-forecasts and pre-operational data
Production of the re-forecasts is complete, and cover the period January 1981 to December 2010. The re-forecasts have also been created for January to April 2011 to bridge the gap to the start of the real-time forecasts.
Pre-operational forecast data have been generated from May 2011, and are now available. Although the pre-operational system runs a few days ahead of System 3, data will not be released earlier than for S3. Pre-operational data will eventually be available via the dissemination system. Note that pre-operational data is not intended for operational use, but is to enable users to configure and test their processing and application software.
For the November 2011 forecast, we expect to change the status of Systems 3 and 4 such that System 4 becomes the operational forecast system. When this change occurs, MARS access to the data from Systems 3 and 4 will remain unchanged. However, System 4 will be the official ECMWF forecast, and will be produced on the System 4 operational schedule. System 4 data will be disseminated as EXPVER=0001, and dissemination of System 3 data will cease. ECMWF derived products, such as the web plots and the multi-model combinations, will all be based on the operational forecast system only.
We intend to continue to run System 3 up to and including February 2012 in a post-operational mode, whereby the data will become available in MARS according to the old System 3 operational schedule. This is to ease the transition process for non-operational users who need more time to adjust their experimental applications to the new data. However, such users are cautioned that the post-operational system will have a reduced level of support, and the availability of System 3 data is not guaranteed. Data from the old system will not be available via dissemination or the GTS. Experimental applications should be migrated to System 4 as soon as practicable. Operational applications should only ever use the operational system.
Data access
Data are archived in the existing multi-model streams, with ORIGIN=ECMF:
The monthly mean anomalies , which are only produced for the real-time forecasts, are archived with appropriate parameter definitions in STREAM=MSMM, instead of STREAM=MMSA. Wave model output will continue to be archived as for System 3.
All production data (re-forecasts, pre-operational and operational real-time forecasts) are archived as EXPVER=0001, SYSTEM=4. If atmosphere data is retrieved on the archived grid, then the resolution will differ from that of System 3.
An example showing how to modify a MARS request for System 3 data according to the details given above in order to retrieve the equivalent System 4 data is given in the table:
| MARS request - System 3 | MARS request - System 4 |
|
|
|
The longer range integrations (out to 13 months) are archived separately from the 7 month integrations, and are accessed by specifying METHOD=3 instead of METHOD=1. The first 7 months of METHOD=3 data for each extended integration is a simple copy of the corresponding METHOD=1 data.
A small number of additional surface fields are archived, together with output from a few additional pressure levels. This is largely related to additional physics in the model and the improved representation of the stratosphere. Limited model level data is still provided for a subset of integrations, for use as forcing for dynamical downscaling of the seasonal forecasts. The annual range forecasts keep the same extensive set of monthly mean data as in System 3, but have a reduced set of daily data. This is also true for some of the new pressure levels in the seasonal integrations – monthly means are available, but 12 hourly data are not. However, where 12 hourly pressure level data is available in System 3 seasonal runs, it will also be in System 4.