Timelines of Numerical Weather Prediction
2005
- January 12
- Data assimilation in the Ensemble Forecast System: Optimum
Interpolation is replaced by the Ensemble Kalmann Filter
method
- March 15
- 4D-Var analysis implemented in the Global data assimilation
system.
2004
- May 18
- New GEM regional model at 15 km. Bring satellite data
assimilation in regional analysis to same level as in global
analysis
2003
- Febuary 24-28
- New GCM II model for seasonal forecasts
- September
- Public access to model output in GRIB format via
Weatheroffice
2001
- January 11
- Implementation of a new regional analysis on "eta"
coordinates. Addition of TOVS and ACARS/AMDAR data
2000
- June 14
- Implementation of a new global analysis on "eta"
coordinates. The 3D-Var data assimilation is performed directly
on the GEM model 28 eta vertical levels. Implementation of new
GEM version to increase the level of activity in the
model.
- September 27
- Improvements to the global analysis: addition of TOVS and
ACARS/AMDAR data
1998
- January 22
- Improvements of the operational force-restore land-surface
scheme
- March 11
- New snow analysis
- September 15
- Regional GEM's hor. resolution increases from 35km to 24
km. The number of vertical levels stays 28. Improvements of
some parametrizations of the physics processes.
- October 14
- The SEF model (T199/21L) is replaced by the GEM model -
global configuration. The global GEM resolution is 0.9º
and 28 levels.
1997
- February 24
- The RFE model is replaced by GEM regional. No regional
analysis cycle. The regional GEM model is initialized with the
global analysis on 16 pressure levels.
- June 10
- The regional GEM model is initialized with the global
analysis on 21 sigma levels instead of the analysis on pressure
levels
- June 18
- The OI is replaced by the 3D-VAR scheme.
- September 24
- Re-introduction of the regional analysis cycle. (with
12-hour spin-up). The resulting regional analyses are used to
initialize the regional GEM.
- November
- Installation of the experimental GEM 15 km run (HIMAP)
1995
- June 14
- Global SEF model with increased resolution from T119 to
T199L21
- July 26
- A new surface layer treatment in EFR as well as a soil
humidity analysis that compensates for the surface humidity
biases of the model (error feedback of the surface dew point
forecast)
- August 9 & 30
- Two versions of HUMSAT were implemented in quick
succession
- September 19
- Improved snow analysis (changes to the albedo calculation
over snow)
- December 13
- Increase of the resolution of the RFE model: from
50km/25levels to 35km/28 levels
1994
- July 21
- New sea surface temperature analysis running once a day
based on the analysis of anomalies with respect to climatology.
New climatology fields were implemented at the same time
- September 21
- New 3-D upper air objective analysis was implemented
(versus previous split horizontal and vertical scheme):
includes additional level at 925 hPa and improved treatment of
SATEMS. Temperature are diagnosed for the height analyses
1993
- June 16
- New higher resolution global system with HUMSAT. Model
increased from T79 to T119 with the same number of vertical
levels (21). Physics is considerably changed with KUO, shallow
convection and new radiation scheme
- November 3
- New 50km/25 level EFR with new physics package (similar to
Spectral SEF except for minor differences). Output is kept on a
variable grid. HUMSAT is used in the regional assimilation
cycle
1992
- March 18
- New ice analysis based on SSM/I data; new snow depth
analysis on 400x200 global gaussian grid. Optimized spectral
model about 12% more efficient. Changes in the warming in the
upper stratosphere; new analysis of humidity when temperatures
are below 0C.
- Dec 16
- New NEC SX-3/44 replaces Cray. EFR 100km has increased
levels to 23 and a slightly modified physics package (including
a new radiation and shallow convection). A regional analysis
with a 12-h spin-up was implemented.
1991
- March 12
- Global spectral model (SEF T79/21 levels) replaces the
spectral hemispheric model; new global data assimilation
system; global surface mean sea level and surface air
temperature analyses
- August 1
- New spectral model that removes warm west coast bias
1990
- March
- Reduced bias in EFR installed on XMP-416
- July 4
- New sea surface temperature on 400X200 on global gaussian
grid
- November 8
- Changes to the trial field model as well as the spectral
forecast model to correct a problem with levels below the
surface
1989
- January 25
- New OA with new horizontal and vertical correlation
functions statistics
1988
- Febuary 23
- Improved vertical resolution of EFR to 19 levels
- June 15
- Global OA operated in hemispheric mode. Analysis of
residuals
- October 31
- Trial field model is the spectral T89/20L model
- Dec 7
- 100km resolution EFR model with microtasking
1987
- Febuary 11
- Optimization of the EFR (10% faster)
- April 12
- Conversion to XMP-28 with 152.4KM resolution and implicit
non-linear normal mode initialization
- July 21
- OA improvements: extended search of best 24 in nearest 40
observations on 180x45 gaussian grid
- October 27
- Code optimization and improvements to the treatement of
water vapour fluxes and virtual temperature
1986
- Febuary 20
- More efficient global code - highly vectorized
- April 22
- Operational implementation of the EFR model (Regional
Finite Element/Éléments Finis Régionaux)
at 190km resolution, 15 vertical levels and spectral normal
mode intialization
- October 23
- Expansion of core window of EFR;
- reduction in maximum grid length spacing to 228.6km
- Dec. 17
- Spectral model improvements. Installed on Cray XMP-28,
later in 1987
1985
- April
- Improved V9 spectral on 130X32 gaussian grid
1984
- June
- Spectral model with triangular 59 (T59) truncation and 15
levels (15L) in the vertical installed on the Cray 1S
1983
- May
- Improved version 9 of the spectral model
1981
- Fall
- Spectral R29 13L model (gaussian grid 96X32)
- New objective analysis was introduced. Normal mode
initialization
1978
- January
- Optimal interpolation (OI) used in an objective analysis on
a gaussian grid
- December
- Spectral R29 10L model, version 8 physics
1977
- March
- Spectral R29 5L model
- V7 physics
1976
- Febuary 18
- Spectral model, 20 waves, with rhomboidal truncation (R20)
and 5 levels (5L) in the vertical. Featured version 6 (V6)
physics.
1975
- mid
- A Baroclinic 7 level primitive equation model was installed
on Cyber 7600.
1968
- September
- A Baroclinic 4 level model was installed on an IBM
360-65.
- "Kruger" type analysis was being used.
1962
- fall
- A barotropic model was run on a Bendix machine.
- First operational forecast 500hPa chart was done in July
1963.
1950
- mid
- A barotropic model using the Fjortoff method
- late
- A barotropic model was run on the IBM 650 at McGill
University.
- Model was run on an experimental basis only.
2004
- January 7
- Conversion to the new IBM supercomputer of the regional and
global forecast systems
- May 18
- 15 km Regional model implantation.
- New satellite data TOVS and GOES for
assimilation.
2003
- June 19
- Corrections to TOVS AMUS-B data assimilation for the
regional model
2002
- December 10
- TOVS assimiliation in global and regional models
2001
- January 11
- Implementation of a new regional analysis on "eta"
coordinates. Addition of TOVS and ACARS/AMDAR data
- June 7
- ATOVS-1B assimilation in global and regional analysis
systems
- September 11
- ISBA surface scheme implantation in the GEM regional
- December 11
- New data set for 3D-Var assimilation (ACARS,AMDAR, 5 more
channels TOVS and SATWINDS of GOES)
- New variational quality control scheme
1998
- September 15
- GEM (24km, 28 levels) / REGIONAL analysis 3D-VAR (12-hour
spin-up)
- October 14
- global GEM (0.9°/28 levels) replaces the SEF model
(T199/21 levels>
1997
- February 24
- GEM (35km, 28 levels) initialized from global analysis OI
3D (without spin-up)
- June 18
- 3D-Var global analysis
- GEM initialized from 3D-Var global analysis (no
spinup)
- September 24
- Regional 3D-Var analysis implemented with 12-hour
spin-up
1996
- December 13
- RFE (35km, 28 levels) / REGIONAL analyse OI 3D (12-hour
spin-up)
1995
- June 14
- SEF model T199L21
1994
- September 21
- 3D global analysis
1993
- June 16
- SEF model T119L21
- November 3
- RFE (50km, 25 levels) / REGIONAL analysis OI 2D (12-hour
spin-up)
1992
- December 16
- RFE (100km, 23 levels) / REGIONAL analysis OI 2D (12-hour
spin-up)
1991
- March 12
- RFE (100km, 19 levels) / global analysis OI 2D
1988
- February 23
- RFE (152.4km, 19 levels) / OI hemispherical analysis in
2D
- RFE (100km, 19 levels) / OI hemispherical analysis in
2D
- global analysis with SEF T79L21 (implemented on Mars
12, 1991)
1987
- April 12
- RFE (152.4km, 15 levels) / OI hemispherical analysis in
2D
1986
- April 22
- RFE (190km, 15 levels)/ OI hemispherical analysis in
2D
Evolution of the CMC's global
forecasting system since 1950
2005
- March 15
- 4D-Var global analysis is implemented. The analysis fits
data along the time axis as well as in the three spatial
dimensions.
2004
- January 7
- Conversion to the new IBM supercomputer of the regional and
global forecast systems
2003
- June 19
- Corrections to TOVS AMUS-B data assimilation for the
regional model
2002
- December 10
- TOVS assimiliation in global and regional models
2001
- June 7
- ATOVS-1B assimilation in global and regional analysis
systems.
- December 11
- New data set for 3D-Var assimilation (ACARS,AMDAR, 5 more
channels TOVS and SATWINDS of GOES)
- New variational quality control scheme
- Corrections to the global model for better wind treatment
in highly mountainous terrain
2000
- June 14
- Implementation of a new global analysis on "eta"
coordinates. Implementation of new GEM version to increase the
level of activity in the model.
- September 27
- Improvements to the global analysis: addition of TOVS and
ACARS/AMDAR data
1998
- October 14
- global GEM model 0.9°/L28
1997
- June 18
- 3D-Var analysis
1995
- June 14
- global spectral model SEF T199/L21
1994
- July 21
- 3D analysis
1993
- June 16
- global spectral model SEF T119/L21 with new physic
1991
- March 12
- global spectral model SEF T79/L21 with global analysis
1985
- April
- spectral model T59/15L with improved V9 physics version,
gaussian grid 130x32
1988
- June 15
- global objective analysis in hemispheric mode
1984
- June
- spectral model T59/15L
1983
- May
- spectral model R29/13L with V9 physics
1981
- December
- spectral model R29/13L on gaussian grid 96x32
1978
- January
- Optimal Interpolation analysis
- December
- spectral model R20/10L with V8 physics
1977
- March
- spectral model R29/5L with V7 physics
1976
- February 18
- spectral model R20/5L with V6 physics
1975
- September
- baroclinic model with 7 levels
1968
- September
- baroclinic model with 4 levels
1950-1960
- barotropic model
Created :
1997-02-24
Modified :
2005-04-07
Reviewed :
2002-12-18
Url of this page : http://www.msc.ec.gc.ca /cmc/op_systems/historical_e.html
The Green LaneTM, Environment Canada's World Wide Web Site.
|