Construction and Analysis of Meteorological Elements Correlation Network

Cui-juan Fang,Feng-jing Shao,Wen-peng Zhou,Chun-xiao Xing,Yi Sui

DOI: https://doi.org/10.1007/978-3-319-59072-1_9

2017-01-01

Abstract:Analysis of the correlation between meteorological elements could help find climate changing patterns. In this paper, the time series of meteorological elements, such as pressure, temperature and humidity, are converted to a correlation network, in which nodes represent the correlation relation (state) between the two meteorological elements and edges represent the transformation between different states. By analyzing the topological properties of the correlation network (degree, strength, path, etc.), the correlation patterns between meteorological elements could be found. Empirical studies of Weifang with 9 years climate observation data show that the correlation network has a power-law distribution and sub-seasonal characteristics. The correlation between temperature and pressure are more strongly negative and it did not change significantly with the year went. The correlation shows a seasonal variation that more negative correlation in summer and the spring as follows.