Study On The Structural Characteristics And Evolution Of ENSO Temperature Network

Enso is a typical interannual large-scale air sea interaction phenomenon in the tropical Pacific,which can have an important impact on the global climate by changing the Walker circulation in the equatorial Pacific.It has always been a key topic in the field of climate research.In recent years,as a new tool,complex network has been widely used in climate research,and many important achievements have been made.Its research idea is to analyze the interaction and relationship between meteorological stations or grids from a macro perspective,and to indicate whether the basic units of climate system have approximate dynamic behavior characteristics through simple network structure parameters.Using climate network,this paper studies the impact of El Ni(?)o and La Ni(?)a events on global temperature.The main research contents are as follows:(1)Based on the global surface temperature data,the global temperature correlation networks corresponding to El Ni(?)o and La Ni(?)a events are constructed respectively,and the differences of their effects on the network topological structure properties are analyzed.The results show that,compared with the La Ni(?)a temperature network,the correlation between grid temperature series in El Ni(?)o temperature network is weakened,and the connectivity of the network is significantly reduced,especially in the tropical and southern temperate regions.The network connectivity of the eastern Pacific,Western Pacific,Indian Ocean and Atlantic Ocean in the equatorial region is relatively large,and the decrease in El Ni(?)o network is notable.They are also the four key regions leading to the differences of the structural characteristics of the two types of network.On this basis,the reason of the difference between the two types of network characteristics has been preliminarily discussed,with the increase of SST in Ni(?)o3.4 region,the SST in equatorial eastern Pacific,equatorial Indian Ocean and other areas rises,which strengthenes outgoing long wave radiation and convection activities,and the interaction between low latitude and mid-latitude areas,and the variance of air temperature changes in the north and south temperate regions increases.As a result,the correlation between the temperature series of the four key regions and the rest of the world is weakened,therefore the connectivity of the global grid temperature network is decreased.(2)By constructing the ENSO temperature correlation network in the cold state(La Ni(?)a month)and warm state(El Ni(?)o month),this chapter reveals the adjustment characteristics of the structural characteristics of ENSO temperature correlation network in the middle of 1970s,and explores its possible causes.The results show that the average connectivity of ENSO temperature correlation networks changed from cold state greater than warm state to warm state greater than cold state around the middle of 1970s,and this change is particularly significant in the tropical ocean region.In this period the temperature anomaly in the tropical East Pacific and other key areas corresponding to the cold state was adjusted from negative phase to positive phase,resulting in a significant increase in the probability of negative temperature covariance between different regions,a decrease in correlation coefficient and a corresponding decrease in network connectivity,while the temperature anomaly in the key areas corresponding to the warm state remained in a positive phase,with a small probability of turning negative temperature covariance and a slight increase in network connectivity,After the mid-1970s,the connectivity of the warm state network is greater than that of the cold state.By filtering out the warming trend,it is found that the global warming trend not only improves the connectivity within the global temperature network,but also makes the difference between El Ni(?)o and La Ni(?)a events on the connectivity of the temperature network more significant.(3)The temperature network parameters are used to evaluate the prediction performance of NCC_CSM model for El Ni(?)o and La Ni(?)a events.Using the structural characteristic parameter of temperature network connectivity,it is found that the spatial distribution characteristics of temperature network connectivity constructed by NCEP reanalysis observation data and NCC_CSM model data are consistent,but the connectivity of model temperature network is significantly higher than that of observation data.With the increase of advance prediction months,the prediction ability of model for El Ni(?)o and La Ni(?)a events gradually decreases,and the prediction ability of this model for La Ni(?)a events is significantly better than that for El Ni(?)o events.