This paper proposes an online performance monitoring method for weather radars based on ground clutter characteristics to evaluate radar system status and enhance the stability and reliability of observation data. The method identifies stable ground clutter regions by exploiting the spatiotemporal persistence of clutter and the high occurrence probability of strong echoes. An empirical cumulative distribution function is then constructed, with the 98th percentile extracted as the monitoring indicator, and a relative calibration bias metric introduced to quantitatively characterize system deviations. A case study with the Jingzhou Radar demonstrates that this method can effectively capture stable clutter features under various weather conditions, different ranges, and anomalous propagation scenarios. Further application of the method to multiple radars in Hubei Province verify its reliability in detecting system performance biases and data anomalies, confirming its feasibility for online monitoring of weather radar performance. Therefore, this method is a strong support for improving the long-term stability of radar observations.