Faits saillants

1

The Regional and Urban Numerical Weather
Prediction and Operational Long Range Plan of the
Meteorological Service of Canada
Dr. Gilbert Brunet Meteorological Research
Branch Meteorological Service of
Canada Environnment Canada Challenges in Urban
Meteorology A Forum for Users and Providers 21
September 2004 Aknowledgement Desgagné, Bélair,
Mailhot and Roch
2
Outline of the talk

• Who we are
• Multi-scale meteorological modeling
• Numerical weather prediction now, in one year and
  ten years
• Two examples of importance for urban area
• -Hurricanes
• - Urban meteorology and air quality
• Future RD challenges

3
Who we are

• Canadian Meteorological Centre and Meteorological
  Research Branch is the
• Canadian equivalent to US NOAA NWS NCEP and US
  Navy FNMOC for numerical weather prediction
• Canadian equivalent to LLNL NARAC for multi-scale
  atmospheric transport and dispersion modeling
• Equivalent centres within the World
  Meteorological Organization Washington (USA),
  Bracknell (UK), Toulouse (FR), Melbourne (AU),
  Tokyo (JP)

4

Uniform resolution Variable resolution
Hydrostatic Nonhydrostatic Global
Limited-area Distributed memory -----------------
---------------- 3D Var Data Assimilation 4D
Var Data Assimilation Ensemble Kalman Filter
Operational forecast Emergency response
Volcanic ash Air quality Stratospheric ozone
Wave model Coupling to oceanographic
simulations etc


5
Improving Hurricane Forecasting
Hurricane Juan 28 September 2003 Halifax
Hurricane climatology
6
Grid of GEM for global Numerical weather
Prediction (33 km horizontal resolution)

7
Instantaneous precipitation rate (mm/hr) for the
Operational GEM model A 5 day animation
(20/01/2002 to 25/01/2002) (HR100km, TR45 min.)
8
Instantaneous precipitation rate (mm/hr) for the
Meso-Global GEM model A 5 day animation
(20/01/2002 to 25/01/2002) (HR33km, TR 15 min.)
9
Canada-Japan Collaboration

• 25 time more powerful than the IBM cluster at
  the
• Canadian Meteorological Center (CMC)
• Available in 510 years from now at CMC

10
Modelling the Full Lifecycle ofHurricane Earl
(Sept 1998)at 1km Resolution with MC2 Model
(Canadian equivalent to US MM5)
964 hPa
ET Phase
985 hPa
Tropical Phase Class2 Hurricane
September 1998 Classified as a very active TC
period
11
-Humidity at 350m height is shown over Gulf of
Mexico for the first 12 hours of the
simulation.-Only 1 of the simulation domain is
shown!
12
From CCRS
REPRESENTATION of URBAN SURFACES in
Meteorological Service of Canada ATMOSPHERIC
MODELS
13
Operational Representation of Urban Surfaces at
the Canadian Meteorological Center (CMC) New
Opportunities
In the higher resolution convective scale models
that are on the verge of being operationally
implemented at CMC, it will become increasingly
important to correctly represent physical
processes over urban surfaces.
This is not the case in the short and
medium-range weather forecast systems currently
operational at CMC, in which even large urban
areas (e.g., 50 km x 25 km) would have a
negligible impact on the atmospheric circulations
produced by the models.
GLOBAL medium-range forecasts
Urban area does not even cover a single grid
point of the model !
? 100 km
Urban area covers 2 model grid points.
REGIONAL short-range forecasts
? 25 km
LOCAL convective scale models
Urban area covers a large number of points
(50x251250)
? 1 km
RPN
14
As the resolution increase, you need to consider
new details such as.
What you need is more than a high resolution
topography Need to develop a new Physics
parameterization scheme (Town Energy Budget or
TEB, details in the surface characteristics
e.g. heat, momentum and moisture fluxes)
15
The Joint Urban 2003 Experiment
Atmospheric dispersion study 28 June to 31 July
2003
Include the following meteorological
measurements 22 surface met stations 6 surface
energy budget stations 2 CTI windtracer lidars 2
radiosonde systems 4 wind profiler/RASS systems 1
FM-CW radar 3 ceilometers 9 sodars Oklahoma
mesonet NEXRAD radars of the US weather service
RPN
16
Other Cases and Collaborations
Assessment of role and impact of TEB in Canadian
urban environments

• For example
• Cold weather cases with snow (e.g., Montreal in
  January)
• Other cases opportunity to use the Multi-city
  Urban Hydro-meteorological dataset (MUHB) with
  Prof. Tim Okes group at U. British-Colombia
• Developping an operational system for Vancouver
  in view of the 2010 Winter Olympics.

RPN
17
GEM is an ideal tool for multi-scale atmospheric
transport and dispersion problems, including
urban scale
18
Urban atmospheric transport and dispersion tools

• We have started a project based on GEM to develop
  and validate an integrated, state-of-the-art,
  high-fidelity multi-scale modeling system for the
  accurate and efficient prediction of urban flow
  and dispersion of CBRN materials.
• Development of this proposed multi-scale modeling
  system will provide the real-time modeling and
  simulation tool to predict injuries, casualties,
  and contamination and to make relevant decisions
  to minimize the consequences based on a
  pre-determined decision making framework.

19
Conclusion

• For 2004-2005 the RD strategy in collaboration
  with CMC, regional weather services and Canadian
  Universities
• Global NWP with a MESOGLOBAL GEM (35km) with a
  lid at the stratopause (.1mb) with the Regional
  GEM physics package
• A 4D-Var data assimilation system with increasing
  new asynoptic and satellite data
• An Ensemble Prediction System (EPS) will be
  delivered with a comprehensive physics and
  initial condition perturbations approach
• - A comprehensive unified EPS RD and
  operational Long Range Plan has bee initiated
  with the National Weather Service. Ribbon Tying
  Ceremony 16-18 November, 2004, NCEP, Camp Springs
  

20
Conclusion

• Improved our Regional (15km-10km) and Local
  (3km-1km) NWP system with applications to urban
  area problems
• Collaborating with CMC, REGIONS and Canadian
  Universities and other partners for Environmental
  Prediction (coupling GEM with chemistry,
  hydrology and ocean)