A complex network representation of wind flows - INSU - Institut national des sciences de l'Univers Access content directly
Journal Articles Chaos: An Interdisciplinary Journal of Nonlinear Science Year : 2017

A complex network representation of wind flows

Abstract

Climate networks have proven to be a valuable method to investigate spatial connectivity patterns of the climate system. However, so far such networks have mostly been applied to scalar observables. In this study, we propose a new method for constructing networks from atmospheric wind fields on two-dimensional isobaric surfaces. By connecting nodes along a spatial environment based on the local wind flow, we derive a network representation of the low-level circulation that captures its most important characteristics. In our approach, network links are placed according to a suitable statistical null model that takes into account the direction and magnitude of the flow. We compare a simulation-based (numerically costly) and a semi-analytical (numerically cheaper) approach to determine the statistical significance of possible connections, and find that both methods yield qualitatively similar results. As an application, we choose the regional climate system of South America and focus on the monsoon season in austral summer. Monsoon systems are generally characterized by substantial changes in the large-scale wind directions, and therefore provide ideal applications for the proposed wind networks. Based on these networks, we are able to reveal the key features of the low-level circulation of the South American Monsoon System, including the South American Low-Level Jet. Networks of the dry and the wet season are compared with each other and their differences are consistent with the literature on South American climate.

Dates and versions

insu-03727075 , version 1 (19-07-2022)

Identifiers

Cite

Maximilian Gelbrecht, Niklas Boers, Jürgen Kurths. A complex network representation of wind flows. Chaos: An Interdisciplinary Journal of Nonlinear Science, 2017, 27, ⟨10.1063/1.4977699⟩. ⟨insu-03727075⟩
2 View
0 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More