Temporal multiscaling characteristics of particulate matter PM10 and ground-level ozone O3 concentrations in Caribbean region

- First multifractal analysis for PM10 and O3 pollutants in Caribbean region.
- Nonstationary and nonlinear nature of pollutants was analyzed in the fully developed turbulence framework.
- Multifractal lognormal model was performed.
- Intermittency parameter was estimated by a lognormal model.

A good knowledge of the intermittency of atmospheric pollutants is crucial for air pollution management. We consider here particulate matter PM10 and ground-level ozone O3 time series in Guadeloupe archipelago which experiments a tropical and humid climate in the Caribbean zone. The aim of this paper is to study their scaling statistics in the framework of fully developed turbulence and Kolmogorov's theory. Firstly, we estimate their Fourier power spectra and consider their scaling properties in the physical space. The power spectra computed follows a power law behavior for both considered pollutants. Thereafter we study the scaling behavior of PM10 and O3 time series. Contrary to numerous studies where the multifractal detrended fluctuation analysis is frequently applied, here, the classical structure function analysis is used to extract the scaling exponent or multifractal spectrum ζ(q); this function provides a full characterization of a process at all intensities and all scales. The obtained results show that PM10 and O3 possess intermittent and multifractal properties. The singularity spectrum MS(α) also confirms both pollutants multifractal features. The originality of this work comes from a statistical modeling performed on ζ(q) and MS(α) by a lognormal model to compute the intermittency parameter μ. By contrast with PM10 which mainly depends on puffs of Saharan dust (synoptic-scale), O3 is more intermittent due to variability of its local precursors. The results presented in this paper can help to better understand the mechanisms governing the dynamics of PM10 and O3 in Caribbean islands context.

108