Revisions during 2003

14 January 2003 Intoduction of Cycle 25r4. This version includes changes to many aspects of the data assimilation and forecasting system:

1. Revised multi-incremental (T95/T159) 4D-Var algorithm, including interpolation of high resolution to low resolution trajectory
2. Use of Omega-equation and non-linear balance in the Jb, together with statistics from ensembles of 4D-Var assimilations
3. Revised Jc (more selective to gravity waves)
4. Assimilation of GOES WV radiances, MODIS winds, more HIRS channels and SAR ocean-wave data
5. Direct assimilation of SSM/I radiances in 4D-Var
6. Improved cloud-scheme numerics, cleaner code, and bugfixes;
7. Revised cloud physics
8. Revised convection scheme with new type and cloud top/base algorithms; checks all levels up to 700hPa for initiation of deep convection.

In addition, the RNORM parameter scaling EPS initial perturbations was reduced from 2.0 to 1.6 as a consequence of the data assimilation statistics retuning. This effectively keeps the EPS spread at the same level as in the current operational version.

The impact of these changes has been found to be meteorologically positive during the test period over a wide range of areas and parameters. Tests conducted during the summer have shown a clear reduction of the forecast errors generated over North America and advected towards Europe at this time of the year. The new model version removes convective potential instability more effectively and realistically, as was found in several case studies. Stratosphere forecasts have also been improved.

The new model cycle is used throughout the Centre's forecast suites, the T511 main run, the EPS and the BC Optional Project. The atmospheric model identification number in the GRIB headers changed from 201 to 202.

Model numbers for the Global wave model (113) and the European waters wave model (213) remained unchanged.

4 March 2003 Intoduction of Cycle 25r5. The ECMWF operational forecast production was moved to the IBM supercomputer which comprises two identical Cluster 1600 systems. Each cluster has thirty IBM p690 enterprise UNIX Servers. All operational products are now provided from the IBM computer system. The corresponding model Cycle is 25r5 which includes several code optimisation for the IBM machine.

29 April 2003 Intoduction of Cycle 26r1. This is a technical change only and involves the writing of observations directly into the Observation Data Base (ODB).

The new model cycle is used throughout the Centre's forecast suites, the T511 main run, the EPS and the BC Optional Project. The atmospheric model identification number in the GRIB headers changed from 202 to 203.

7 October 2003 Intoduction of Cycle 26r3. Changes from the previous operational version (Cycle 26r1) include:

1. a new formulation of the humidity analysis (modified background-error covariances, corrected calculation of background-errors for SSMI that uses FASTEM emissivity model over sea);
2. new data streams (AIRS from Aqua, AMSU-B, AMSU-A from Aqua, Japanese wind profilers, Meteosat5, GOES9 and GOES12 WV clear sky radiances, GOES12 winds and MIPAS ozone-profile retrievals);
3. monitoring of ENVISAT data: SCIAMACHY, GOMOS, and MIPAS;
4. a new linear radiation scheme in 4D-var, a new radiation sampling (HALO) and a new aerosol climatology in the full model;
5. a relaxation of the convective mass flux limiter for long time steps (used for the EPS and monthly forecasts);
6. new model parameters (UVB, CAPE, photo-synthetically active radiation, freak waves). Details of the new parameters are:
   • Atmospheric model (Local GRIB table: 128):
     • 57 UVB Downward UV radiation at the surface w m**-2
     • 58 PAR Photosynthetically active radiation at the surface wm**-2
     • 59 CAPE Convective available potential energy J kg**-1
   • Wave model: (Local GRIB table 140):
     • 252 WSK Wave surface kurtosis -
     • 253 BFI Benjamin-Feir index -
     • 254 WSP Wave spectral peakedness s**-1

   For further information please refer to ECMWF Technical Memorandum 366. The fields for these new model parameters are written into MARS and are available through dissemination from all suites and for all default post-processing time steps.

• The change to the new cycle was transparent for users apart from a change in Model identification number for the atmospheric model in GRIB headers, which was set to 121.
• Note: after reaching the end of the defined range of possible identifications numbers in cycle 26r1 with 203 it was agreed that the numbers will be recycled.
• The wave model identification numbers were not changed and continue to be 113 for the Global Wave Model and 213 for the Mediterranean Wave Model.
• Verification results gathered to date from the experimental suite and from several months of research experiments indicated that, in general, a significant and consistent improvement in the upper air (including humidity) forecasts can be expected from this new cy26r3.