Spatio-temporal Analysis Of China's Maritime Crude Oil Flow Based On AIS Big Data

Energy cooperation is a key area of "The Belt and Road Initiative",while crude oil is an important part of energy cooperation.As the world's largest crude oil importer,China imports most of its crude oil by sea.Therefore,analyzing the spatio-temporal changes of China's maritime crude oil flow is of great significance to national energy security.However,the public crude oil flow data is mainly based on statistical data,which is relatively coarse in time and space.Furthermore,it is difficult to provide the maritime crude oil flow at a specific time,a specific route,and a specific port.As a new type of real-time big data source,the Automatic Identification System(AIS)can extract China's high spatio-temporal resolution of maritime crude oil flow,providing data support for the national energy transportation security and energy strategic reserves.There are still many shortcomings in the existing research on the calculation of maritime crude oil flow.The crude oil flow obtained from statistical data has low temporal and spatial resolution.The crude oil flow calculation model based on the gravity equation method is not mature enough and the accuracy of result is low.Although the method based on AIS big data can extract marine crude oil flow with high spatio-temporal resolution and rich information,there is no simple and effective calculation method.Furthermore,compared with the existing research based on statistical data,the analysis of maritime crude oil flow based on AIS big data can explore the current situation of China's crude oil imports from a deeper and broader level,and more accurately grasp its temporal and spatial characteristics.In this paper,we use AIS big data,take China's maritime crude oil transportation as the research object,and use time series decomposition,network analysis,spatial clustering,community division,mathematical statistics and other methods,to carry out a spatiotemporal analysis of China's maritime crude oil flow from the national and port level.The research content mainly includes the following three aspects:(1)Propose a calculation method of maritime crude oil flow based on the crude oil tanker load condition.This method includes data preprocessing and maritime crude oil flow calculation,which can extract maritime crude oil flow with high spatio-temporal resolution at the national and port level.Data preprocessing includes crude oil tanker data screening,"dirty data" cleaning,ship load information mapping,and ship sharing Maritime Mobile Service Identity situation identification.Maritime crude oil flow calculation includes discrimination of the crude oil tanker load condition based on k-means clustering algorithm,crude oil tanker voyage division considering the whole life of the ship,and crude oil flow statistics based on voyage load condition.By comparing the maritime crude oil flow extracted in this paper with the crude oil trade volume provided by Joint Organization Data Initiative(JODI),it is found that the correlation coefficient,R2,between the two is 0.82,showing they are highly correlated.Since there exists the situation that non-island countries transfer crude oil through land,rail,pipeline and other methods,and finally realize their own crude oil import and export through ports in other countries,there is a spatio-temporal shift between the maritime crude oil flow in this paper and the JODI data.However,their total crude oil flow is relatively stable.The crude oil flow in this paper is 12%lower than the JODI data in terms of crude oil import and export volume.This is because the data in this article only counts the country's maritime crude flow while the JODI data counts the country's crude oil flow transported by all means.All of these show the reliability of the calculation method proposed in this paper.(2)Analyze the spatio-temporal characteristics of China's maritime crude oil flow at the national level.In terms of time series analysis,it is found that China's annual maritime crude oil flow is close to Clarkson Research data,and has the similar trend,which verifies the accuracy of China's maritime crude oil flow.Due to the U.S.military activities in the Persian Gulf and the outbreak of Middle East Respiratory Syndrome epidemic around the world in 2015,China's maritime crude oil flow has two obvious troughs.Ignoring these troughs,China's maritime crude oil flow is showing a trend of rising to stable changes.In terms of spatial distribution,China's main maritime crude oil suppliers are Angola and Middle Eastern countries.In the directed crude oil transportation network with countries as nodes,due to the characteristics of China importing more crude oil but exporting less,China's degree centrality and weighted degree centrality are very high,while the closeness centrality and betweenness centrality are low.Furthermore,spatio-temporal analysis were carried out for crude oil tankers of different sizes and countries to which they belonged.It is found that The larger the crude oil tankers are,the more crude oil is imported to China,and the more consistent the trend is with China.Also,the larger crude oil tankers have more crude oil suppliers,and the main crude oil suppliers are farther away from China.The higher the voyage proportion of the country that the crude oil tankers belong to is,the more crude oil is imported to China,and the more the trend is consistent with China.Except for the crude oil tankers belonging to Kuwait,which mainly travel between China and Kuwait,the crude oil tankers belonging to other countries has no special national preference for crude oil transportation.(3)Analyze the spatial differentiation of China's maritime crude oil flow at the port level.In terms of difference analysis,China's ports have formed four port groups in Bohai Sea,East China Sea,Taiwan Strait,South China Sea through spatial clustering.The maritime crude oil flow of the northern port groups is higher than that of the southern port groups.The time series trends and the spatial distribution of main crude oil suppliers in these port groups are similar,while the port structure and crude oil supply structure in these port groups cluster are different.In terms of network analysis,in the undirected crude oil flow network with ports as nodes,the degree centrality and weighted degree centrality of Chinese ports have a power function relationship.Since the crude oil flow between ports in this network does not distinguish the direction,Chinese closeness centrality ranked first port is higher than China in the world rankings.After community division,Chinese port is divided into four communities with discrete and intersecting spatial distribution.The main community is connected to the Middle East,the secondary community is connected to the surrounding areas of China,and the remaining communities play a supporting role.In terms of comparative analysis,we compare the crude oil flow network above and the undirected crude oil tanker flow network with ports as the node.It is found that the centrality of these two networks has a strong correlation,and is very similar in value and ranking.In the crude oil tanker flow network,the Chinese ports are divided into three communities.The main community include almost Chinese ports and is connected to the Middle East and China's surrounding areas.This is quite different from the community division result in the crude oil flow network.