Hailstone detection based on time series association rules

Hailstone is one of main meteorological disasters and it is difficult to forecast effectively. In this paper, a new hail echo detection method based on time series association rules was proposed and a hail echo automatic detection system was built. Compared to the existing methods, the test results have shown that this method eliminated hail alert failure completely and also reduced the proportion of false alarm, raising the accuracy of hail forecast. The system was able to assist forecasters to make correct hail forecasts and provide reference to other disastrous weather forecast.