The use of levoglucosan for tracing biomass burning in PM_2.5 samples in Tuscany (Italy)

Abstract

Levoglucosan was present in all samples and its concentrations showed a pronounced annual cycle with maximum levels in the cold season. The annual percentage of ratios of levoglucosan to OC ranged from 0.04 to 9.75% evidencing a major contribution of biomass burning to the aerosol OC during the winter. In the urban-background site, OC was strongly correlated with EC in winter, suggesting that the major fraction of OC was generated as primary particles along with EC. A background levoglucosan component showed that biomass burning was continuously taking place in all the investigated sites. The biomass burning contribution to the Tuscany aerosol was made up of a background component and an additional component during winter probably due to wood burning for domestic heating.

Introduction

Particulate matter (PM) of both natural and anthropogenic origin, is known to cause adverse effects on human health (Harrison and Yin, 2000; McDonald et al., 2004; Metzger et al., 2004) and visibility (Tao et al., 2009), but also to play an important role in local and regional air quality, as well as to represent a critical factor in climate change (Novakov and Penner, 1993; Kanakidou et al., 2005) and solar radiation (Crutzen and Andreae, 1990; Andreae, 1993). Many countries in the European Union, and many other countries worldwide, have adopted air quality standards for PM₁₀ and PM_2.5 in ambient air and started monitoring programmes to identify and quantify the sources that would be the most effective to control in order to reduce PM emissions.

Biomass burning, associated with both open and domestic fires, is a significant source of aerosol particles to the atmosphere, and its contribution to the carbonaceous aerosol in urban sites (Fine et al., 2001; Lanz et al., 2008) has been recognised to be substantial.

Levoglucosan (1,6-anhydro-β-d-glucopyranose) is considered a highly specific tracer for biomass burning aerosols (Simoneit et al., 1999; Puxbaum et al., 2007), because it derives from the pyrolysis of cellulose (present in every type of wood) and hemicellulose at high temperature (> than 300 °C) (Zhang et al., 2008). It is a perfect molecular marker because it is relatively stable in atmosphere and emitted in large amounts (Schkolnik and Rudich, 2006). The use of levoglucosan as biomass burning tracer has been preferred to the conventional water soluble potassium, since recent studies have evidenced that potassium has other significant sources, i.e. meat cooking, waste incinerators, coal usage (Duan et al., 2004; Wang et al., 2007; Caseiro et al., 2009), whereas levoglucosan is specific to the combustion of cellulose containing materials and no other sources are currently known to exist. Retene was also proposed as a tracer for wood smoke, however there are several problems with this option because retene is found primarily in coniferous woods and it is a semi-volatile compound with a high vapour pressure, thus making it difficult to quantify (Nolte et al., 2001). Major known sources of levoglucosan are fuel wood combustion in the winter season and agricultural, garden waste burning and forest fires, in particular in Mediterranean countries, especially in the summer (Puxbaum et al., 2007). The wildfires in Tuscany are not usually very extensive, an average area of 3.3 Ha and 0.9 Ha in 2009 and 2010, respectively, is reported (www3.corpoforestale.it).

In the present study, we report levoglucosan concentrations determined in PM_2.5 samples, collected in three different sites in Tuscany (Italy) during an annual sampling campaign (from March 2009 to April 2010). Together with levoglucosan, aerosol mass concentration and organic (OC) and elemental (EC) carbon were measured and their seasonal variations and correlations were also investigated.