No description available.

The thesis deals with testing of radar-based hail detection methods and their use in hail climatology of Czechia. Data from C-band Doppler radars were used. There is described the testing of the selected algorithms in the first part of the thesis. The algorithms were chosen according to the data availability. On the basis of 25 well documented hail events, suitable threshold values were found to form hail criteria. Because of the hail events considered, the work concentrates on severe hail with diameter from about 2 cm. In addition, available data from Germany (Baden-Württemberg) were included in testing the detection criteria. The best performing criteria were used to form a new combined criterion (COMBI). The second part of the work applies the hail criteria, in particular the COMBI criterion. The performance of the COMBI criterion is shown in four case studies. Then the COMBI criterion was determined from the radar data of the six years 2007-2012. All the days from the months May to August were evaluated. The point maximum of about one day with severe hail was found for Czechia. For the most of the Czech territory the value was not higher than a quarter of a day in a year. In the seasonal distribution the maximum of hail days belongs to June. And the diurnal distribution showed the maximum of severe hail...

Thesis (MSc (Mathematical Sciences. Applied Mathematics))--University of Stellenbosch, 2006. / In this work we present a novel method to do measurement correlation between target observations made by two ormore radar systems. Some of the most common radar sensors available are those measuring only range (distance to the target) and bearing (azimuth angle). We use these measurements to determine the correlation between two di¤erent sensors observing the same target. As a by-product of the correlation algorithm, we nd a way to estimate the target height for a target observed by at least two radar sensors. The correlation method is expounded upon, where we discuss measurement correlation for moving targets. Targets are tracked using a Kalman Filter, and correlation is done between new observations and existing target tracks. Finally, the correlation algorithm is implemented in an interactive 3D computer simulation. Results obtained indicate a high success rate, with false correlations only obtained where sensor accuracy is the limiting factor.