Using Self Organizing Maps to Investigate Summer Synoptic Climatology in North China Area
Analysis of the synoptic climatology over the North China Area promotes a better understanding of the relationships between atmospheric circulation and surface elements. Therefore, investigation on synoptic climatology features is conducted based on the classified synoptic types of North China using the summer ERA 40 MSLP data from 1958 to 2002 which are treated by self organizing maps (SOMs) algorithm. 36 different synoptic patterns are identified, which can be divided into 4 kinds of extreme types such as strong north high and south low, strong west high and east trough, strong northwest low and southeast high and westward extension of strong eastern high, as well as transitional types between them. Study of spatial characteristics indicates that the symmetry of synoptic types in 2D self organizing maps reflects the general characteristics of synoptic climatology over the North China Area, whereas the asymmetry represents the unique features. Study of temporal characteristics shows that synoptic situation of the study area is relatively more stable with stronger high and low systems, or with dominant high system, and vice versa. By referring to the interannual analysis, it is apparent that 6 patterns are dominated by remarkable linear trends. Finally, analysis is performed on precipitation distribution characteristics of the corresponding synoptic patterns. It shows that the rainfall in different parts of North China is contributed by distinct synoptic types, and slight changes in surface circulation can produce drastically different locations of precipitation center, while the topographic effect will further amplify the difference. This study uses a more comprehensive dataset and higher temporal resolution data than most of the past studies to quantitatively investigate the summer synoptic types in North China, resulting in the expansion of synoptic climatology research over the study area. Therefore， results here can be applied to develop the identification technology of the typical weather process in numerical weather predication, and to be the basis of regional climate scenarios research.