You are here
ECMWF is developing the pre-operational atmospheric service of the European COPERNICUS programme. The near-real-time production (analyses and forecasting) system is based on the Integrated Forecast System (IFS). In earlier versions, IFS was coupled to the chemistry transport model MOZART-3. Since 2014-09-18, C-IFS (composition IFS) with the on-line integrated TM5 chemistry is used.
The NRT stream produces analyses and forecasts of global stratospheric fields of O3.
Other reactive gases are available from the chemistry model. For comparison purpose we display also NOx (as NO+NO2 for CIFS-TM5 based model).
This succession of models and configurations is listed here by date of change in the resulting analyses:
- 2015-09-03, CIFS-CB05 g9rr based on C-IFS 41r1_CAMS: first implementation of the operational CAMS o-suite. The changes are described in the corresponding e-suite validation report and have no impact on stratospheric ozone.
- 2014-09-18, CIFS-CB05 g4e2 : this is a major switch, where CIFS-CB05 uses the photochemical module of TM5 which does not take stratospheric chemistry into account. The only product suitable for stratospheric composition is now ozone, thanks to data assimilation.
- 2014-02-24, IFS-MOZART fnyp : another change happens, e.g. IFS Cy40R2. The preparatory e-suite experiment was named fzpr. The final log_fnyp corresponds to that configuration.
- 2013-10-07, IFS-MOZART fnyp : several changes happen, e.g. IFS Cy38R2. The preparatory e-suite experiment was named fwu0
- 2013-01-07, IFS-MOZART fnyp : major change w.r.t. strato ozone: assimilated MLS data (NRT dataset) switches from v2.2 to v3.3. This fixes the vertical structure issues described by Lefever et al. (ACP, 2015)
- 2012-07-05, IFS-MOZART fnyp : started on originally with IFS Cy37R3.
- 2009-10-09, IFS-MOZART f93i : IFS Cy36R1 and assimilating (among other streams) the Aura-MLS v2.2 NRT ozone profiles. See log_f93i. The stratospheric ozone delivered by this experiment was validated in Lefever et al. (ACP, 2015).
A parallel stream built from preliminary experiments, know as the e-suite, is used for the next implementation of the o-suite.
The e-suite as displayed on this site uses:
- No data from 2017-02-01 at 0000UT
- CIFS-TM5 gkvv from 2016-04-01 at 0000UT see
- No data from 2015-09-03 at 0000UT
- CIFS-TM5 g9rr from 2014-10-01 at 0000UT see esuite
- No data from 2014-09-18 at 0000UT
- CIFS-TM5 g4e2 from 2013-12-01 at 0000UT see esuite_g4e2
- IFS-MOZART fzpr from 2013-08-01 at 0000UT see esuite_fzpr
Implementation of control runs (aka "forecast runs")
A control run is executed in parallel with the o-suite. It configuration is (should be) identical to the o-suite except for the absence of data assimilation. The goal is to better understand the impact of data assimilation in the o-suite products.
More details can be found on the page about NRT dissemination of global products