The aim of the Symposium is to examine the capabilities, opportunities and requirements for improved forecasts in the 0-6 hour timeframe. Special emphasis has been placed on forecasts of high impact weather such as summer convective hazards (heavy rain, hail, lightning, high winds), as well as winter weather events (snowstorm, blizzards, etc.) and other hazardous conditions (fog, low ceiling and visibility, etc.). Since the 0-6 hour period spans the timeframe where it is known that both traditional Nowcasting techniques and NWP can contribute useful information, both approaches are covered at the Symposium as well as methods that combine Nowcasting and NWP.
There is a significant opportunity to improve forecasts of high impacts weather in the 0-6 hour timeframe. Traditional nowcasting techniques, based on extrapolation, rapidly lose skill in the first hour or so due to the difficulty in predicting system evolution whilst NWP models exhibit very little predictive ability in the first six hours due primarily to the problem of spin-up. Hence, there is a gap approximately between one and six hours, where our current predictive ability is lacking.
The Symposium organizers believe that the time is ripe to make significant progress in this area. In recent years there have been important advancements in a number of fields that are required to make progress. The ability to observe the four-dimensional structure of the atmosphere has increased dramatically in the last decade with the introduction of new remote-sensing platforms such as Doppler radar, Next-generation weather satellites, wind-profilers, etc. New data assimilation techniques, such as 3DVar/4DVar and the Ensemble Kalman Filter (EnKF), have been developed and are being tested in order to assimilate these non-traditional data sources into numerical weather prediction models. The grid spacing of NWP models is now approaching a horizontal resolution at which moist convection can be explicitly resolved. Verifying these forecasts, which have high intermittency in both time and space, is an important research topic; significant progress has been made recently in developing new verification statistics that aim to give a truer measure of forecast skill than traditional r.m.s. scores.
More traditional Nowcasting techniques will continue to play a central role in very short-range weather forecasting, for a number of reasons. Firstly, unless the dynamical balances in a convective storm are specified very accurately, it is difficult for a numerical forecast in the first few hours to improve over extrapolation techniques (in which the balances are implicit). Secondly, the CPU expense of the data assimilation and NWP techniques mentioned above make it unlikely that any numerical forecast can be made and disseminated to the end user in less than an hour (at least for another decade or so). Hence, Nowcasting techniques, which are not as computationally intensive, have a very important role to play in this timeframe.
And let's not forget the role of the human! Forecasting significant weather can hardly be left to purely automated systems, because the risks and consequences are too high for neglecting human expertise. One of the goals of the Symposium will be to identify the optimal mix of human and machine in the preparation of very short range forecasts of hazardous weather.
However, the techniques described above will not be sufficient alone to ensure progress in the area of shrt-range weather forecasts. Basic understanding of the dynamics and four-dimensional structure of precipitating weather systems is required to put together all of the pieces of a successful forecast system. For example, high-resolution NWP will require a more accurate representation of picrophysical processes inside clouds and precipitation as well as more refined subgridscale turbulence parameterization than those used currently for operational NWP. Hence, basic studies of the dynamics of precipitating weather systems and boundary layer processes are required for making progress in very-short-range weather forecasting. These studies are also essential for providing conceptual models in oriented-object approaches in Nowcasting.
Finally, research into the problem of very short range forecasting will be for naught unless the needs of the end-user are considered. It is probably fair to say that many end-users do not have much experience with such short-range weather forecasts, hence understanding user needs and tailoring forecasts for effective utilization are essential. Furthermore, it should be remembered at the Symposium that because of vast disparities in technical infrastructure, the needs and requirements of the end-user varie immensely from country to country.
This Symposium has been designed to address many of the issues raised above. Towards that goal, the Symposium has been divided into eight sessions, namely :
- Basic understanding of atmospheric processes in the 0-6 hour timeframe
- Techniques
- Data assimilation
- Observational needs
- Applications
- Very short range NWP
- Operational systems and the role of the human in the very short range forecasting
- Verification
The Symposium organizers wish you a productive and enjoyable time in Toulouse!
|
Andrew Crook, chair of the Scientific Committee
|
|
Stéphane Sénési and Pascal Brovelli, chairs of the Local Organizing Committee
|
|
Tom Keenan, chair of the WWRP Nowcasting Working Group |
|
