Thunderstorm Tracking and Monitoring on the Basis of Three Dimensional Lightning Data and Conventional and Polarimetric Radar Data.

Kurzfassung

The aim of this work is to assess the benefit of total-lightning information as independent data source for thunderstorm tracking and short-term prediction (nowcasting) of storm evolution. Special focus has been laid on the three-dimensional lightning information and the in-cloud and cloud-to-ground discrimination provided by the lightning detection network LINET. The reliability of the lightning information and its usability for nowcasting purposes have been tested both separately and in combination with other data sources which are commonly used for thunderstorm nowcasting. The new thunderstorm tracker ec-TRAM (tracking and monitoring of electrically charged cells; Meyer et al. (2009)) has been developed to identify, track, and monitor thunderstorms in high temporal and spatial resolution by combining the information of independently tracked convective ground-precipitation cells and lightning-cells to new cell objects. The algorithm builds on the autonomously operating routines rad-TRAM (tracking and monitoring of radar cells; Kober and Tafferner (2009)) and li-TRAM (tracking and monitoring of lightning cells). The latter has also been developed within this work. The new tracking algorithm has been tested based on a thunderstorm data set of more than 500 storm tracks which were recorded by ec-TRAM in southern Germany during summer 2008. It is found that the newly composed cell objects comprehensively describe simple as well as complex thunderstorm structures and the cell tracking method of ec-TRAM proves to be more coherent and stable in comparison with the tracking performances of rad-TRAM and li-TRAM. For two selected thunderstorms the time series of cell parameters monitored by ec-TRAM have been complemented with three-dimensional polarimetric radar data and satellite data to assess how the temporal evolution and parameter correlation of total lightning strokes, hydrometeor formation, ground precipitation patterns, and cloud top temperature can be used to estimate the storm state and predict its development. The parameter evolutions are found to be consistent with the current state of knowledge. A principal life-cycle scheme can be identified for the cell parameters on large time scales. The stronger fluctuating short-term parameter evolutions are found to refl ect the momentary storm dynamic. Based on the lifetime diagrams several warning parameters for subsequent storm events can be suggested. Significant cell parameter correlations, which can be parameterized, are also found in statistical analyses over the complete data set. Strong positive correlations are found between cell extension, discharge frequency, and in-cloud discharge height. Two cell regimes, sharply separated at a specific cell characteristic, can clearly be identified in all correlation diagrams. Interpreted on the basis of previous studies and in terms of the current state of knowledge, it seems most likely that the two cell-regimes refl ect the storm characteristics of different storm organization forms. The parameterized correlation curves could then be used as cell parameterizations in operational nowcasting tools to predict the dynamic evolution, duration, and danger potential of a storm, provided that the storm system can be classified. Finally, it can be concluded that this study demonstrates the usability and the promising potential of total-lightning data as reliable and independent data source for future nowcasting tools.