| Home > Products > Data Services > ECMWF system > Brief history > |
Brief history of the ECMWF analysis and forecasting system
| The Operational ECMWF system | ||||
|
| See also | ||
|
A brief history of the ECMWF forecasting system from 1985 to the current date follows.
On 1 May 1985 the forecast model cycle is 22. The T106 model became operational. The spectral representation in the horizontal is truncated at wavenumber 106. The vertical resolution was not changed. Note that before 1 May 1985 for the first four months of the TOGA period ECMWF still operated the T63 model with 16 levels in the vertical.
On 13 May 1986 the forecast model cycle is 26. The 16-level model was replaced by a 19-level model. The three extra levels were introduced in the stratosphere.
In 1987 the ECMWF data assimilation system consisted of a multivariate optimal interpolation analysis, a non-linear normal model initialisation and a high resolution spectral model which produced a first-guess forecast for the subsequent analysis. Data were assimilated every 6 hours.
In 1987 the ECMWF forecast model used a spectral formulation in the horizontal, with triangular truncation at total wavenumber 106, a vertical coordinate with 19-level resolution which was terrain-following at low levels. The comprehensive physical parametrisation schemes included shallow and deep (Kuo) convection, a radiation scheme which allowed interaction with model generated clouds and the diurnal radiative cycle.
On 17 September 1991 the next major change in model resolution
was introduced into operations as cycle 39 with the T213 model on 31 levels.
An article from ECMWF Newsletter describing these changes is given
in pdf 
(380KB)
file.
On 30 January 1996 ECMWF introduced a 3-dimensional variational
(3D-Var) analysis scheme. 3D-Var is a new code for the analysis of model-level
values of temperature, vorticity, divergence and specific humidity, and
surface pressure. Minor changes to the forecast model were implemented
at the same time.
An article from the ECMWF Newsletter describing these changes is given
in pdf (232KB)
file.
On 25 November 1997 The first version of a four-dimensional variational data assimilation system (4D-Var) was introduced. It is based on an evolution of the previous 3D-Var system with a 6-hour cycling.
An article from the ECMWF Newsletter describing these changes is given
in pdf (36KB)
file.
On 1 April 1998 the model spectral resolution was increased from T213 to TL319.
On 9 March 1999 Introduction of 50 level model (Cy19r2). This
version has 50 levels in the vertical instead of 31 in the operational
version, most of the extra resolution being in the stratosphere between
150 hPa and 0.1 hPa (the current operational model top level is at 10
hPa). Horizontal spectral and physical truncations are kept unchanged
(TL319). The model resolution in the Ensemble Prediction System
(EPS) will continue to be TL159L31.
An article from the ECMWF Newsletter describing these changes is given
in pdf (784
KB) file.
On 12 October 1999 Introduction of cy21r4 with 60 levels in the vertical for the reference model (T319) and 40 levels in the EPS (T159).
An article from the ECMWF Newsletter describing these changes is given in pdf
(204
KB) file. On 12 September 2000 Introduction of Cy23r1. Implementation of 12 Hour 4D-Var.
21 November 2000 The model resolution was upgraded from TL319
to TL511 in the deterministic mode, and from TL159
to TL255 in the ensemble mode (EPS).
An article from the ECMWF Newsletter describing these changes is given
in pdf (332
KB) file.
 |
21.01.2002 |
 |
© ECMWF |
