An investigation of the tipping term as a prognostic tool in short-range thunderstorm forecasting

A nowcasting system, GANDOLF, was operated by the Met Office and the University of Salford during the Sydney 2000 Forecast Demonstration Project (FDP). The purpose of this project was to test and evaluate the current state-of-the-art in nowcasting technology, and initial verification by the University of Salford has shown GANDOLF's performance to be poor. The system uses a conceptual model of the life cycle of an idealised convective cell to forecast the evolution of radar-observed storms in an object-oriented fashion. Analysis has shown the implementation of the model to be inadequate due in part to its reliance on relatively low resolution NWP model forecasts rather than the use of conceptual modelling itself. The advent of very high resolution three dimensional wind fields produced by four dimensional variational analysis (4DVAR) modelling techniques offers a way of circumventing many of the problems inherent in GANDOLF, and such data have been obtained for the cases during which GANDOLF was operational. These were produced by the National Center for Atmospheric Research's VDRAS system, which was also operated during the FDP, as part of the three-dimensional US Autonowcast system. The wind data allowed the calculation of time series of the three-dimensional tipping term field, which, it is suggested, can be used as a prognostic tool in automated convection nowcasting. This paper summarises the theoretical background of the tipping term in the context of severe storms, and presents the results of a number of tests performed on the data. The conclusions of these tests indicate that the tipping term may be a useful prognostic tool. Observations of the increasing asymmetry in the tipping term couplet predicted by theory as convection reaches its maximum intensity may offer a way of improving objective analysis of the life cycle of cells.