Inversion of particle size distributions based on wavelet analysis

Abstract

ABSTRACTIn the Earth’s radiative budget, the role aerosols play in climate change depends on their particle size distribution. To address disadvantages linked to obtaining particle size distributions using traditional inversion algorithms, we propose a novel method combining wavelet analysis and Tikhonov regularization to solve the Fredholm integral equation of the first kind. This method can retrieve aerosol particle distributions, using aerosol optical thickness (AOT) from a CE-318 sun photometer. Experiments were carried out to verify the information retrieved using this new algorithm. AOT and particle size distributions were analysed and retrieved during various weather conditions, including dust storms, haze, fog, thunderstorms, and sunny skies. In addition, the seasonal average optical thickness and aerosol particle size distributions for the Yinchuan area in China were studied over a recent four-year period. Our results showed that particle size distributions were bimodal during dusty conditions, but unimodal during all other weather conditions studied. Moreover, our results indicate that air quality improved over two recent years.