Prediction support system for lightning flash based on case-based retrieval

In order to prevent the damages caused by natural phenomena, such as, rain, cloud, lightning and squall, we use meteorological radars for the observation. In recent years, the demand is increasing for more accurate information for natural phenomena, since it is becoming more important to analyze the observation data for the prediction. We propose a prediction support system based on case-based retrieval for lightning flash. The system watches the change of thundercloud, and retrieves the similar cases of meteorological features. The retrieval results are shown with similarity for present meteorological circumstances.