Monitoring and modelling forest fire emissions toward an accurate inventory development

Forest fires play a crucial role in the natural ecosystems, with both negative and positive impacts on all biosphere components and reverberations on different scales, from local to global. For example, forest fires pose a significant risk of reversal of carbon stored in the biomass through vegetation combustion, soil degradation, organic matter consumption, and decrease of biodiversity. Among the primary effects of forest fires, the production of a remarkable amount of greenhouse gases and solid particulate matter is one of the most important. CO2 and CO are the dominant fractions released by fires. The wide variety of pollutants released include other greenhouse gases [methane (CH4), nitrous oxide (N2O)], photochemically reactive compounds [e.g., nonmethane volatile organic carbon (NMVOC), nitrogen oxides (NOX)], and fine and coarse particulate matter (PM). All these emitted compounds interfere significantly with atmospheric budgets and air quality at local, regional, and even global scale. Under global warming, climate driven increases of wildfire extent and severity might exacerbate the current situation.

Over the last decades, the demand for fire emission (FE) quantification increased due, in principal, to two main causes. The first is related to the amplified need for GHG emissions accounting by governments and companies. The second is the increased need for key inputs to be supplied to analyse and model better the impact of vegetation fires on atmospheric conditions and air quality. A crucial point to support FE inventories and modelling has been recognized in up-to-date, accurate, and consistent FE estimates. Typically they are affected by a number of errors and uncertainties depending on multiple and interdependent factors (fuel characteristics, burning efficiency, fire type, weather, and geographical location). Improvements were made possible through new advances in remote sensing, experimental measurements of emission factors, and the use of biogeochemical and fuel consumption models. Despite these advances, the quality of the inventories is still hard to assess, since the methodologies, input data, and assumptions vary strongly between the inventories. In general, the comparison between emission estimates from many inventories is recognized to be a helpful method to identify uncertainties.

From this context, this work aims firstly to presents the contemporary state of the art in FE monitoring, modelling, and inventory development in Italy. Several publicly available inventories were compared in terms of FE historical trends, seasonality, and spatial distribution. The results pointed out that FE estimates are affected by large uncertainties mainly due to the differences in capturing the burned area, despite the improvements achieved by remote sensing. In addition, significant inconsistencies derive from the assessment of fuel characteristics. Finally, the work briefly presents the development of an integrated methodology to assess the estimation of greenhouse gas emissions through the coupling of fire behaviour, risk exposure model, and semi-physical fire emission models.