The consequences of a large number of fire events and the relative burned area that occur each summer in the Mediterranean Basin are crucial for land degradation, air quality, and weather and climate changes. One of the first relevant effects of the burning activity is the emission of pollutant species (e.g. 10 and 2.5 mm particulate, oxides of nitrogen, sulphur dioxide, etc.) and greenhouse gases, which have been recognized as an important issue by the Kyoto Protocol: the dominant fraction of vegetation fire emissions is released as CO2 and CO, being responsible for about 90–95% of the total carbon emitted (Andreae and Merlet, 2001). Model approaching can contribute to appraise the fuel consumption and the resultant emissions although some uncertainties affect these estimates due to the inaccuracy of the quantitative input data and the spatial variability of fuel characteristics that influence smoke production. In this context, more comprehensive and accurate data inputs would be of valuable help for  predicting and quantifying the source and the composition of fire emissions.

The objective of this study is to estimate the emission of dioxide and monoxide carbon, as well as other pollutant species, from several Mediterranean maquis fires occurred in Sardinia island (Italy) during 2007. We used the USDA First Order Fire Effects Model (FOFEM) to assess the amount of consumed fuel and its estimated emissions in both flaming and smoldering combustion.

In order to achieve realistic fire emission estimates, FOFEM input fuel load data were surveyed in several representative maquis fuel types correspondent to those combusted, and parameterized in custom fuel models. Fuel model maps, obtained from supervised classification of remotely sensed images, were finally used to better quantify and spatialise the emission information.

The results show that the use of both appropriate fuel data and fine fuel maps is crucial to attain reasonable simulations of fuel consumption and smoke emissions. The FOFEM outputs and the derived smoke emission maps can be useful for emission source models coupled with dispersion models and decision support systems.