Chaotic Signature in Power Spectrum and Recurrence Quantification of Dynamical Behaviour of Multivariate Time Series

Abstract

This study investigates the presence of nonlinear dynamics and chaotic behaviour in air temperature, atmospheric pressure, and relative humidity using data collected from Lagos, Nigeria. Power spectral density analysis revealed an aperiodic nature with possible non-linear processes governing the time series. Recurrence quantification analysis (RQA) was employed to quantify the determinism and chaoticity within the data. Results suggest that all three variables (relative humidity, air temperature, and atmospheric pressure) exhibit evidence of both deterministic and chaotic behaviour. Deterministic behavior was highest for air temperature, followed by pressure and then humidity. Conversely, chaoticity was highest for relative humidity, followed by pressure and then air temperature. These findings suggest complex underlying dynamics within the troposphere, potentially influenced by the region's convective nature and intense precipitation events. The observed determinism indicates some level of predictability, particularly for air temperature, while the chaoticity highlights the inherent complexity of atmospheric processes