Capturing and Analyzing Coherent Structures in Temporal Streamflow with Complex Networks

To analyze the coherent structures and the partitioning quality of the atmosphere, the network algorithm is introduced and applied to double gyre flow and Arctic stratospheric flow. By partitioning the network, each part consisting of regions with similar dynamic behaviors, including mass transport and mixing, can be identified from Lagrangian frame. First, a directed network is considered by using the trajectories of sampled particles, and Lagrangian flow network (LFN) is established by capturing the mass exchanges between pairs of grid points in the streamflow. Then, the Infomap method is used to detect well-connected regions. In particular, both coherence and mixing metrics can be served as indicators of transport, and therefore the spatial connectivity can be quantified. Further, the proposed efficient method and LFN are implemented successfully in the practical geophysical flow, and the dynamic behaviors of the flow in Arctic are analyzed in detail. By the example, the advantages of the ensemble solutions are evaluated as analyzing the structures of the flow field, and the extent of the impact on the community is assessed in the selection of the open domain. Also, the persistence of coherent structures of streamflow is verified over years. As the conclusion, it is shown that LFN is a novel and promising quantitative analysis method. The results obtained could gain the key understanding on dynamic behaviors in streamflow and the development of complex networks in the research of atmospheric spatial connectivity.