Abstract
In order to improve the accuracy of strong rainfall nowcasting for a very short term usually of one hour, a new rainfall nowcasting method based on the image processing technique is developed. The key technique in this study are to evaluate the deformational errors between observed and predicted rainfall areas obtained by Japan Meteorological Agency (JMA) and to minimize errors of the JMA rainfall nowcasting in real time based on its evaluation. As a result of verification experiments for intense rainfall events, it is concluded that our new method remarkably improves the accuracy of rainfall nowcasting by considering complicated deformation processes of rainfall area, and the predictive value for one hour ahead forecasted by the new method is at least 10% greater than that by the original JMA rainfall nowcasting.
