Investigating chaotic features in solar radiation over a tropical station using recurrence quantification analysis

Abstract

The use of solar energy for power generation and other uses is on the increase. This demand necessitate a better understanding of the underlying dynamics for better prediction. Nonlinear dynamics and its associated tools readily lend itself for such analysis. In this paper, nonlinearity in solar radiation data is tested using recurrence plot (RP) and recurrence quantification analysis (RQA) in a tropical station. The data used was obtained from an ongoing campaign at the Federal University of Technology, Akure, Southwestern Nigeria using an Integrated Sensor Suite (Vantage2 Pro). Half hourly and daily values were tested for each month of the year. Both were found to be nonlinear. The dry months of the year exhibit higher chaoticity compared to the wet months of the year. The daily average values were found to be mildly chaotic. Using RQA, features due to external effects such as harmattan and intertropical discontinuity (ITD) on solar radiation data were uniquely identified.