Université du Québec à Montréal

Montréal (Québec) Canada

Corresponding author’s address:

René Laprise

Department of Earth Sciences

Université du Québec à Montréal (UQAM)

B. P. 8888, Stn Centre-Ville

Montréal (Québec)

Canada H3C 3P8

laprise.rene@uqam.ca

Global atmospheric General Circulation Models (GCMs), coupled with land-surface, ocean and sea-ice modules, are the most sophisticated tools for understanding the physical processes responsible for the maintenance of the climate system, its natural variability and anticipated changes. Such mathematical models of the climate components are very demanding on computer resources. For this reason, GCMs are limited to coarse computational meshes; typical horizontal resolutions are seldom better than about 500 km, and the reliable scales of such models are even larger than this. Most environmental, societal and economic impacts resulting from anticipated climate changes, however, are associated with processes operating on much finer scales. There thus exists a demonstrated need for "downscaling" GCM results. The only realistic approach to produce climate projections at spatial scales of a few tens of kilometers is to limit the domain over which high resolution is required. The concept of Regional Climate Models (RCMs) consists in applying to climate simulations a time-honoured numerical weather prediction technique of "nesting" a limited-area, high-resolution regional model within a global, coarse-mesh GCM.

The Canadian RCM (CRCM) is based on the highly efficient fluid dynamics kernel of the MC2 (Mesoscale Compressible Community) model, developed by the late André Robert and colleagues of the Cooperative Centre for Research in Mesometeorology (CCRM). This kernel is coupled with detailed physical parameterisation packages developed at the Canadian Centre for Climate Modelling and Analysis (CCCma). The latest version of CRCM includes the state-of-the-art land-surface scheme CLASS developed at the Atmospheric Environment Service (AES). An extensive suite of diagnostics tools has also been designed to analyse the vast amount of simulated regional climate data, following the modular approach used at CCCma for the global model.

Two 5-year integrations of the CRCM nested with Canadian GCM simulations corresponding to current and doubled greenhouse-gas concentrations have recently been completed and analysed. These simulations were performed over a region covering western Canada on a 45-km grid mesh using a 15 min timestep. It is noteworthy that such a long timestep makes the CRCM 3 to 5 times faster compared to other RCMs. Preliminary analysis of these high-resolution simulations reveals a much more realistic distribution of precipitation over complex topography in British Columbia (paper by Laprise et al. submitted to Atmosphere-Ocean).

The CRCM is undergoing the following developments and improvements:

• Coupled mixed-layer ocean and thermodynamic sea-ice modules are being adapted for use within RCM.
• A coupled dynamical regional ocean model is also currently being tested in order to study complex air-sea interactions in the Gulf of St-Lawrence.
• Tropospheric chemistry modules have been implemented in the CRCM in order to study tropospheric ozone episodes in the Windsor-Québec City corridor.
• The transport and size segregation of aerosols is being developed for the study of climate feedbacks between aerosols and radiation in the Arctic as part of the Northern Aerosols Regional Climate Modelling (NARCM) project.
• The CRCM is also used to study the energy and water budgets over the Mackenzie River basin as part of the Canadian GEWEX MAGS project.
• The CRCM serves as driver for studying energy and mass exchanges between the atmosphere and land surface within the context of the CLASS project.

RCMs can also serve as "intelligent interpolator". From the prescription of the time evolution of a small set of coarse-grain atmospheric variables (winds, temperature, pressure and atmospheric water vapour) at the perimeter of its regional domain, RCMs generate a set of coherent fine-grain atmospheric and surface fields, as well as fluxes within its domain. RCMs can thus generate difficult (or impossible) to observe fields, such as precipitation amounts in data sparse regions. This tool can serve to create synthetic or surrogate data bases to feed environmental impact, adaptation and management models.

The CRCM project is now entering the model validation phase. This validation of regional simulations constitutes an important aspect of the project since the model’s success at reproducing a variety of atmospheric phenomena lends confidence in its ability to correctly simulate altered climate conditions. In this validation phase, the CRCM will be tested over a wide range of situations. The CRCM has successfully simulated tropical easterly waves, the types of disturbances that occasionally develop into hurricanes. A version of CRCM extending into the middle atmosphere is being tested to study gravity wave production, vertical propagation and momentum deposition which is so important for the understanding of dynamics where the ozone layer is located. Composite satellite measurements of Earth’s surface parameters are used to validate surface processes in CRCM. Archived radar data are also being used in order to compare precipitation statistics with those derived from CRCM simulations. Progress along these lines have been reviewed at the latest RCM Workshop described below.

The development and validation effort of the CRCM is carried out collaboratively within AES, CCCma, Forestry Canada and in Canadian Universities, including Victoria, Toronto, York, Sherbrooke, McGill and UQAM. The CRCM project has been financed by Environment Canada, the Canadian Climate Research Network (NCRC) through the Canadian Institute for Climate Studies (CICS), the NSERC Strategic Grant programme, the Québec FCAR programme and through internal grants of UQAM.

On February 27 and 28 1997 the Canadian RCM group hosted its annual workshop at UQAM. With over sixty participants, this workshop brought together from Canada and abroad a number of students and experts on regional climate modelling and applications to impact and adaptation studies.

The two-day workshop consisted of some 14 oral presentations including 3 key-note lectures by guests speakers, 13 posters and 4 working group sessions. (Numbers beside participants names refer to their institution, as listed below.) The invited talks were the following:

• Regional climate modelling in BALTEX, by D. Jacob (8)
• Nested model studies of elevation and lake effects on the surface climate change signal, by F. Giorgi (9)
• Observational evidence of climate changes in Switzerland, by M. Beniston (7)

A wide range of topics were also covered in the other oral presentations:

• Hydrological cycle in Mackenzie River basin, by M. MacKay (1)
• A Canadian RCM/CLASS simulation over the HAPEX-MOBILHY area, by E. Chan (1)
• Coupled mixed-layer and sea-ice modelling with the Canadian RCM, by S. Goyette (2)
• The influence of land surface fluxes on the atmospheric response of a RCM over Alaska, by A. H. Lynch (3)
• The Northern Aerosols Regional Climate Modelling (NARCM) project, by L. Barrie (1)
• Sea-salt modelling in the NARCM project, by S.-L. Gong (1)
• Modelling of tropospheric chemistry, by D. Plummer (5)
• Climate scenario simulations over western Canada with the Canadian RCM, by R. Laprise (11)
• Ecological modelling in climate models, by B. Bass (1)
• Report on Workshop held at NCAR on Model Data Users needs, by C. Hakkarinen (6)

Two posters sessions were held during extended lunch periods to allow participants to interact with Graduate Students, Research Assistants and Faculty Members involved in research activities using the CRCM. The following posters were presented:

• Adjustment following a prescribed diabatic heating, by A. Caya, P. Zwack and René Laprise (11)
• Sensitivities of the one-dimensional Kain-Fritsch convection scheme, by D. Paquin and R. Laprise (11)
• The influence of lateral boundary conditions on the fine scale features developed in CRCM simulations, by D. Caya, S. Biner and R. Laprise (11)
• Regional Scale Variability of the climate in RCM Model Simulations, by H. Côté, R. Laprise, M. Giguère and G. Bergeron (11)
• Sensibility experiment on the influence of different sea ice conditions in Hudson Bay on the atmosphere, by P. Gachon and P. Zwack (11)
• Forest Fire Indexes Climatology with RCM: 1xCO₂ and 2xCO₂, by A. Arif and J.-P. Blanchet (11)
• A simulation of the effects of Gulf of Mexico sea surface temperature anomalies using the Canadian Regional Climate Model, by Y. Shao, C. Lin (4) and R. Laprise (11)
• A physical based numerical model to study the potential release of greenhouse gases by hydroelectric reservoirs, N. Barrette, R. Laprise and M. Lucotte (11)
• Towards the application of the CRCM to tropospheric ozone climatology, by V. Bouchet, E. Torlaschi, R. Laprise (11) and J. McConnell (5)
• Simulation of a radiatively active tracer: Kuwait oil fire case, by J. S. Fontecilla and J.-P. Blanchet (11)
• Lamb disturbances in the RCM-MC2 model: explicit resolution Vs parameterization, by C. Thurre and R. Laprise (11)
• Fine-scale modelling of air flow and tracer dispersion over Mexico City, by L. Spacek and M. Giguère (11)
• Water budget: explicit Vs diagnostic methods, by G. Bergeron, A. Frigon and R. Laprise (11)

Four parallel working group sessions were held to discuss specific topics in restricted groups under the lead of a chair:

• Sensitivity of RCM to lateral boundary conditions, chaired by D. Jacob (8)
• Data needs for impacts studies, chaired by M. Beniston (7)
• Hydrological Applications/Processes for RCM’s, chaired by M. MacKay (1)
• Data and strategies for RCM validation, chaired G. Morneau (10)

This year’s Workshop attracted a lots of interest with a participation exceeding previous years’ attendance. Next year it is planned to hold a joint workshop with modellers and applications and impact researchers, possibly in early April 1998. Those interested in receiving information about the Canadian RCM project or forthcoming workshops should contact Prof. René Laprise, principal investigator, Canadian RCM project (laprise.rene@uqam.ca).