Clustering of high dimensional dynamic data is challenging problem. Within the frame of big data analysis, the computational effort needed to perform the clustering task may become prohibitive and motivated the construction of several algorithms or the adaptation of existing ones, as the well known K-means algorithm. One of the critical problem in k-means, k-menoid, k-means or other clustering algorithms required to pre-assigned no. of k which cannot detect non-spherical clusters. With the existing RLClu algorithm needs users to pre-assign two minimum thresholds of the local density and the minimum density-based distance. Clustering is the process of data classification when none prior knowledge required for classification. To overcome these problems STClu clustering algorithm is proposed. In this algorithm a new metric is defined to evaluate the local density of each object, which shows better performance in distinguishing different objects. Furthermore, an outward statistical test method is used to identify the clustering centers automatically on a centrality metric constructed based on the new local density and new minimum density-based distance. Dynamic clustering is an approach to get and extract clusters in real time environments. It has much application such as, data warehousing, sensor network etc. Therefore there is need of such technique in which the data set is increasing in size over time by adding more and more data.