This study examined the quantification problem of a storm shape using the concept of equivalent ellipses. The equivalent ellipses of a storm event were estimated at every time step with respect to the several thresholds of rainfall intensity, which was also examined in terms of their size and number. In addition, the average equivalent ellipse was decided, and the confidence intervals of major axis, minor axis, and rotational angle were calculated to evaluate if the average equivalent ellipse could be the representative one. As results, the following results could be derived. First of all, the number of equivalent ellipses and the size of equivalent ellipses increase as the threshold increase. Secondly, the appropriate ratio of major and minor axises of equivalent ellipse is 2 : 1. Finally, the average rotational angle estimated with respect to several threshold rainfall intensities were all found not to be statistically different from that of all representative rotational angles.