Tuesday, 11 February 2003
Statistical methods for nowcasting thunderstorm rainfall
For many hydrological applications, especially flash flood warning and urban drainage management a good short period forecast of heavy rainfall is required. However, the deterministic nowcasting of thunderstorm motion and development has proved a difficult problem in meteorology due to the short spatial and temporal continuity of convective systems. Methodologies that use extrapolation cannot provide information that can be applied for such hydrological applications in a robust manner. In this paper we demonstrate the use of a statistical method of nowcasting. Our methodology utilizes an approach that efficiently parameterizes spatio-temporal dynamic models in a hierarchical framework. Furthermore, this approach can easily incorporate additional information to aid in the nowcast. Finally, we note that the approach accounts for the uncertainty in the prediction and provides relevant distributional information concerning the nowcast. Case studies are presented that show the effectiveness of the technique and its potential for hydrological use.