Abstract:
This paper applies a variety of statistical methods for exploring relationships between time series of ground-level ozone concentrations, aerosols (in terms of number-concentrations of particles larger than 0.5 µm) and meteorological parameters. The time sequences were obtained automatically and simultaneously (records of the every hour averages) at a coastal station near Ahtopol in Bulgaria for the period May 10 – July 27, 2002. Introductory statistical analyses suggested a number of potential interrelationships between ozone concentration variations and some meteorological parameters. There were also indications of aerosol-particle concentrations showing some connections with wind speed and wind direction, and also with solar radiation; the latter relationship however likely to be a result of the partly radiation-driven pattern for wind speed and direction. These introductory findings warranted the use also of Mathematical Spectral Analysis (MSA) in analysing the time series, so as to show the usefulness of MSA for meteorological applications. The conclusions based on the power spectra, the cross-amplitude spectrum, the absolute coherency spectrum and the phase spectrum are confirmed by the diurnal cycles. The power spectrum of the observed data as well as the crossamplitude, absolute coherency and phase spectra were calculated for hourly meteorological parameters, aerosols and ozone data. The present article contains the calculations of the spectra and their interpretation. Among conclusions is that diurnal variations of meteorological parameters well explain diurnal variation of ozone concentrations, but that the daily aerosol concentrations variations are less affected by diurnal variability of meteorological parameters except solar radiation.