A methodology is presented for the real time, automated identification,structure analysis, tracking and short range forecasting of storms with Doppler weather radar volume scan data. A “storm” is defined as a continuous region where the specified thresholds for both volumetric size and radar reflectivity are exceeded. The “centroid tracking” technique is employed to match the storms at one time with those at the following, and mergence and division have been taken into account. Based on the least square analysis on the time series of the central position and size of a storm, its new position and size within a short period can be predicted with linear extrapolation. The result shows that this method can be a useful tool for the identification, tracking and warning of severe convective weather.