Investigation of chaotic features of surface wind speeds using recurrence analysis

Research output: Journal Publications and ReviewsRGC 21 - Publication in refereed journalpeer-review

18 Scopus Citations
View graph of relations

Author(s)

  • Z.R. Shu
  • P.W. Chan
  • Q.S. Li
  • Y.C. He
  • B.W. Yan

Detail(s)

Original languageEnglish
Article number104550
Journal / PublicationJournal of Wind Engineering and Industrial Aerodynamics
Volume210
Online published4 Feb 2021
Publication statusPublished - Mar 2021

Abstract

Modelling and forecasting of wind speed are of essential importance in various science and engineering applications. However, due to the complex nonlinear dynamics that governing wind speed fluctuation, precise windspeed modelling remains a difficult task in practice. This paper investigated the usability of recurrence analysis for diagnosing hidden structures in wind speed dynamics, in which nonlinear dynamic analysis techniques, i.e., recurrence plot (RP) and recurrence quantification analysis (RQA), were implemented on the wind speeds measured at several surface stations in Hong Kong. The existence of chaos was identified via both phase space reconstruction and recurrence plot. The results indicated apparent site-to-site variability in wind speed dynamics, with the estimated RQA measures ranging from 4.694 to 5.859 for Lmax (maximal length of diagonal structures), 0.284–0.390 for DET (determinism), and 2.125–2.401 for ENT (entropy). Generally, TC station at hill-top tends to possess the largest RQA measures, whereas CCH station at an offshore island and TMS station at hill-top hold the lowest values, implying that the wind speed fluctuations at TC are likely to exhibit higher degree of complexity. Moreover, seasonal variability in wind speed dynamics was also evident, with winter season usually indicates the lowest RQA measures. The largest seasonal RQA measures, on the other hand, vary from site-to-site.

Research Area(s)

  • Chaos theory, Nonlinear dynamic analysis, Recurrence plot, Recurrence quantification analysis, Surface wind speed, Time series analysis