1Institute of Environmental, Chemical and Pharmaceutics Sciences, Department of Earth and Exact Sciences, Federal University of São Paulo, Brazil
2Institute of Physics, University of São Paulo, Brazil
3Department of Sanitary and Environmental Engineering, Federal University of Juiz de Fora, Juiz de Fora, Brazil
4Max Planck Institute for Chemistry, Biogeochemistry Department, Mainz, Germany
Abstract. In this study, we examine the spectral dependence of aerosol absorption at different sites and seasons in the Amazon Basin. The analysis is based on measurements performed during three intensive field experiments at a pasture site (Fazenda Nossa Senhora, Rondônia) and at a primary forest site (Cuieiras Reserve, Amazonas), from 1999 to 2004. Aerosol absorption spectra were measured using two Aethalometers: a 7-wavelength Aethalometer (AE30) that covers the visible (VIS) to near-infrared (NIR) spectral range, and a 2-wavelength Aethalometer (AE20) that measures absorption in the UV and in the visible. As a consequence of biomass burning emissions, about 10 times greater absorption values were observed in the dry season in comparison to the wet season. Power law expressions were fitted to the measurements in order to derive the Ångström exponent for absorption, defined as the negative slope of absorption vs. wavelength in a log-log plot. At the pasture site, about 70% of the Ångström exponents fell between 1.5 and 2.5 during the dry season, indicating that biomass burning aerosols have a stronger spectral dependence than soot carbon particles. Ångström exponents decreased from the dry to the wet season, in agreement with the shift from biomass burning aerosols, predominant in the fine mode, to biogenic and dust aerosols, predominant in the coarse mode. The lowest Ångström exponents (90% of data below 1.5) were observed at the forest site during the dry season. Also, results indicate that low absorption coefficients were associated with Ångström exponents below 1.0. This finding suggests that biogenic aerosols from Amazonia may have a weak spectral dependence for absorption, contradicting our expectations of biogenic particles behaving as brown carbon. Nevertheless, additional measurements should be taken in the future, to provide a complete picture of biogenic aerosol absorption spectral characteristics from different seasons and geographic locations. The assumption that soot spectral properties represent all ambient light absorbing particles may cause a misjudgment of absorption towards the UV, especially in remote areas. Therefore, it is recommended to measure aerosol absorption at several wavelengths from UV to near IR to accurately assess the impact of non-soot aerosols on climate and on photochemical atmospheric processes.