An improved k-means clustering algorithm is proposed after analyzing the disadvantages of the traditional k-means algorithm. The cluster centers are initialized by combining the sample mean and standard deviation, the optimal cluster centers are searched by the hybridizing particle swarm optimization and traditional k-means algorithm, and the criterion function is improved during the iteration process to search the optimal number of clusters. The theory analysis and experimental results show that the improved algorithm not only avoids the local optima, also has greater searching capability than the traditional algorithm. This improved algorithm is used to analyze the morphology of the ridge sail (the upper surface of ice ridges). The comparison with the measured data shows that the influences of the geographical locations and the growing environments on the formation of ice ridges can be perfectly reflected by the clustered results.