FUTURE CHALLENGES

Internationally, the increasing demand for accurate weather and environmental prediction has led to significant attention being given to investments in numerical weather prediction. This is particularly true in the European community and in the United States. The international community has provided Canada with important collaborative opportunities. However, Canada is responsible for prediction of significant weather within our own boundaries, including our environmentally sensitive Arctic.

Ongoing and growing demands on the NWP research program include:

• Improved accuracy for small-scale events: NWP R&D activities will require increased model resolution, better vertical extension, and improved physics in view of improving forecast accuracy at all time and space scales in support of MSC’s severe weather warning system.

• Ensemble prediction: The non deterministic aspect of weather prediction at all time and space scales needs to be quantified with ensembles.  This will lead to increased usefulness of forecasts. Furthermore,  such tools are necessary to meet the growing demand of probabilistic forecasts for operational decision making systems for economical, agricultural and industrial management.

• Environmental Prediction: Environmental Prediction (EP) provides linkage between the land, ocean and atmosphere. The science behind environmental prediction plays a role in helping us understand and respond to: environmental issues (e.g., environmental emergencies, the decline, invasion and adaptation of species), economic issues (e.g., reducing loss of forests due to fire,  insect and disease infestations), social issues (e.g., mapping flood plains, providing design specifications for river diversions and dams to reduce the impact of floods), and political issues (e.g., climate change, bulk water exports, the ozone hole problem).

• Four recent examples of this are: 1)MSC has been asked to seek policy direction for a federal road weather service program as an abatement strategy for road salt use for Canada; 2) there is great pressure to have hydrologic-atmosphere coupled model to predict river flow in view of assessing flood and epidemic risk; 3) farmers will need hourly forecasts of winds, precipitation and temperature because of the growing use of microfarming techniques (irrigation, pesticide spraying); 4) the shipping industry need to know which routing will lower their fuel costs as they enter the ice-covered Gulf of St-Lawrence in winter.
  

The diversity of applications in EP will require significant resources for coupling of the MSC NWP atmospheric model (GEM) to other environmental models and decision-making tools.

• Performance Measurement: NWP systems have become much more complex over the years in terms of modelling, computing, and in the diversity of data coming from the observation network. Performance measurement is an increasingly important part of the research program at Meteorological Research Branch (MRB).  This  involves validating a NWP system by running assimilation cycles with a series (ensemble) of forecasts. These sensitivity studies are intended to estimate the impact of each suite of changes on dynamics, physics and observations, including the optimization of the observation network (maximum impact with a minimum of resources).
• Modern data assimilation system closely builds upon the driving NWP model, especially 4D data assimilation. RPN has achieved in the last decade with GEM, the formulation and implementation of a unified modelling and data assimilation strategy (with no international equivalent) to tackle with optimal resources these challenging new demands.
  
  

RPN has now arrived at an important crossroad in its R&D strategy with GEM:

1. One modelling component is focused on medium and long range weather prediction and involves global modelling with appropriate physics and variational data assimilation package.

2. The second is focused on very fine space scale and short time nested modeling at the regional/urban level.

3. The third activity involves the coupling of the atmospheric models with a variety of other physical and biological models (environmental prediction).

The first activity is essential to support the future of the Canadian global data assimilation research plan that is essential to the AEPD NWP operational weather MSC warning program. The first activity was, until the present time, the main focus of research at RPN. The two other research activities are growing under the pressure of the new mandate of the MSC. The success of the second activity depends dramatically on the quality of the first activity, since these new regional modelling strategies will be essentially driven by the AEPD global NWP system. Even if MSC is at the international R&D forefront in regional modelling, there is still many unanswered questions related to this topic that need to be addressed. Note that significant research progress in these two branches of activities will result in important impacts on the third activity, environmental prediction. The future of a scientific based environmental prediction system lies in the framework of a state-of-the-art NWP system.