Classification of Weather Patterns over the East Asia Region Using Clustering Analysis

In medium-range forecasting, the dependence of ensemble prediction data is high. However, there is a time limit for the weather forecasters to analyze all the ensemble data in detail. To utilize the ensemble data effectively in medium-range forecasting, representative weather patterns over the East Asia region in this study are defined. The k-means clustering analysis is applied to classify the weather patterns, objectively. Daily Mean Sea Level Pressure (MSLP) anomaly of the European Centre for Medium-Range Weather Forecasts (ECMWF) ReAnalysis-Interim (ERA-Interim) during 1981 ~ 2010 (30 years) is used as input data. To define the number of clusters and optimal study area calculate the Explained Variance (EV) and Explained Cluster Variance (ECV) values.In this study, the number of clusters and optimal study area is defined by thirty (k=30) and 20∼60°N, 100∼150°E.30 representative weather patterns with physical features such as frequency, wind, and pressure are summarized. Weather pattern #1 occurred all seasons, but it was about 56% in summer (June ~ September). The relatively rare occurrence of weather pattern (#30) occurred mainly in winter. In the future, these results will be used as the ensemble classification reference data for the scenario-based ensemble medium-range forecast.

Acknowledgements: This work was funded by the Korea Meteorological Administration Research and Development Program "Research of convergence technology of analysis and numerical model for severe weather" under Grant (1365003081).