Aerosol optical properties measured in Argentina: wavelength dependence and variability based on sun photometer measurements

This paper deals with the spectral dependence and time variability of Angstrom wavelength exponent scaling law (alpha), which is the spectral varying slope of the logarithmic relationship between aerosol optical depths (tau) and the wavelength (lambda). It is commonly used to retrieve intensive air masses optical properties such as aerosol size distribution from extensive quantities (tau) and Angstrom turbidity coefficient (beta). This spectral variation of alpha is studied at different wavelengths from measurements taken by ground-based sun photometer covering from near-infrared to ultraviolet range. We analyze the spectral measurement of aerosols optical depths at eight specific selected wavelengths from 340 to 1020 nm using the sun photometer measurements from AErosol RObotic NETwork (AERONET) from NASA. Data from the entire year 2000 were used from instruments deployed at two different sites covering the regions of Argentina as northcentral at Cordoba CETT (31.5S, 64AW) and "pampa humeda" at Buenos Aires CEILAP (34.5S, 58.5W). A new approach of Angstrom wavelength exponent spectral variation was developed to take into account with a more accurate precision the significant curvature appearing in the logarithmic relation between tau and lambda. Using the direct spectral solar radiation set, time series of Angstrom coefficient of turbidity and wavelength scaling law was computed with a day to day data base clustering with uncertainty lower than 0.01 in the optical depth reconstruction over the bulk sun photometer measurements. Temporal series of constant and spectral dependence of wavelength exponent scaling law and turbidity coefficient was derived and shown to vary in space and time. Different meteorological forcing for both sites was evidenced using a regression coefficient analysis to well assess the spectral dependence of wavelength exponent coefficient due to the different cumulating mode of particles and air masses origin at different sites. This spectral decomposition is a key issue in aerosols analysis of steady state and regional scale intrusion episodes with strong connection to their potential contribution of pollution episodes in air-quality problems on urban environment. (C) 2003 Elsevier Science Ltd. All rights reserved.