【摘要】 运用空间数据挖掘中的聚类方法(K-均值法和CLARANS法),对影响夏季青藏高原上中尺度对流系统(MCS)东移的动力场空间分布特征进行了研究,得到了有利于MCS移动和传播的动力学条件。研究成果不仅为预测高原上MCS的移动趋势,揭示其发生、发展的机理提供了一种新的方法和思路,且对于提高长江流域暴雨和强对流天气预报的准确性具有重要的实际应用价值。更多还原

【Abstract】 Clustering methods, i.e., the K-means method and CLARANS method, are used to analyze the spatial distribution features of dynamical field influencing Mesoscale Convective Systems(MCS) move out of the Tibetan Plateau eastward from June to August 1998, and the dynamical conditions favoring the trajectories and propagations of MCS are obtained. The results indicate that the changes of the spatial distribution features of vorticity field, divergence field and vertical wind speed field are helpful to predict the development of MCS. They offer a new method to discuss the regulations of MCS trajectories and the causes of MCS generation and development. The results also reveal that it is of great value in improving the precision of heavy rainfall and intensive convective weather forecasting in the Yangtze River Basin.更多还原