Analysis of particulate emissions from tropical biomass burning using a global aerosol model and long-term surface observations

Abstract

Abstract. We use the GLOMAP global aerosol model evaluated against observations of surface particulate matter (PM2.5) and aerosol optical depth (AOD) to better understand the impacts of biomass burning on tropical aerosol over the period 2003 to 2011. Previous studies report a large underestimation of AOD over regions impacted by tropical biomass burning, scaling particulate emissions from fire by up to a factor of 6 to enable the models to simulate observed AOD. To explore the uncertainty in emissions we use three satellite-derived fire emission datasets (GFED3, GFAS1 and FINN1). In these datasets the tropics account for 66–84 % of global particulate emissions from fire. With all emission datasets GLOMAP underestimates dry season PM2.5 concentrations in regions of high fire activity in South America and underestimates AOD over South America, Africa and Southeast Asia. When we assume an upper estimate of aerosol hygroscopicity, underestimation of AOD over tropical regions impacted by biomass burning is reduced relative to previous studies. Where coincident observations of surface PM2.5 and AOD are available we find a greater model underestimation of AOD than PM2.5, even when we assume an upper estimate of aerosol hygroscopicity. Increasing particulate emissions to improve simulation of AOD can therefore lead to overestimation of surface PM2.5 concentrations. We find that scaling FINN1 emissions by a factor of 1.5 prevents underestimation of AOD and surface PM2.5 in most tropical locations except Africa. GFAS1 requires emission scaling factor of 3.4 in most locations with the exception of equatorial Asia where a scaling factor of 1.5 is adequate. Scaling GFED3 emissions by a factor of 1.5 is sufficient in active deforestation regions of South America and equatorial Asia, but a larger scaling factor is required elsewhere. The model with GFED3 emissions poorly simulates observed seasonal variability in surface PM2.5 and AOD in regions where small fires dominate, providing independent evidence that GFED3 underestimates particulate emissions from small fires. Seasonal variability in both PM2.5 and AOD is better simulated by the model using FINN1 emissions. Detailed observations of aerosol properties over biomass burning regions are required to better constrain particulate emissions from fires.

Near-field emission profiling of tropical forest and Cerrado fires in Brazil during SAMBBA 2012

“…Scaling factors can range from 1.5 to 5, representing a significant underprediction of aerosol abundance in the atmospheric column. The EF values presented here are generally either similar to or lower than previous values reported for this biomass burning environment, suggesting that the EFs used in models are not responsible for the underestimate of AOD over tropical South America; several modelling studies have been undertaken during SAMBBA (Archer-Nicholls et al, 2015;Reddington et al, 2016;Pereira et al, 2016;Johnson et al, 2016) and have required scaling of their emissions to match in situ and satellite measurements. Consequently, scaling emission factors to match observations implies that the discrepancy lies elsewhere if it does relate to emissions (e.g.…”

“…We recommend that comparisons of techniques be made in the future to assess the size of any such potential biases. Our calculated EFs do not indicate that the scaling of emissions that is required within global and regional numerical models to reproduce in situ and satellite aerosol concentrations over Brazil (Kaiser et al, 2012;Tosca et al, 2013;Archer-Nicholls et al, 2015;Reddington et al, 2016;Pereira et al, 2016;Johnson et al, 2016) is related to underestimates in EFs used in emission inventories.…”

Tropospheric Ozone Assessment Report: Assessment of global-scale model performance for global and regional ozone distributions, variability, and trends

“…Some recent inventories have started to use satellite retrievals of fire radiative power instead of burnt area which appears superior at least in some world regions (Kaiser et al, 2012). However, a comparison of five global biomass burning emission inventories based on different satellite fire or burned area products showed a large range of 365-1422 Tg CO emissions for the year 2003 (Stroppiana et al, 2010;Reddington et al, 2016). Different estimates of biomass burning emissions are also highly variable in the spatial patterns, temporal variability, and long-term trends (Schultz et al, 2008).…”

Estimation of black carbon emissions from Siberian fires using satellite observations of absorption and extinction optical depths

“…Our estimate for the integral emissions over the fire season considered is a factor of 2.2 larger than the corresponding estimate based on the GFED4 data. Based on comparison of satellite-derived and simulated AOD, several previous studies showed evidence that OC emissions provided by the GFED inventory may indeed be underestimated in different regions of world, including Siberia (see, e.g., Petrenko et al, 2012;10 Tosca et al, 2013;Konovalov et al, 2014;Reddington et al, 2016), although it was also argued (Konovalov et al, 2015;2017b) that models may underestimate AOD due to inadequate representations of the BB aerosol aging processes. So it is possible that a part of the differences between our optimal estimate of the OC emissions and the corresponding GFED data may compensate for some missing processes (e.g., involving the formation of SOA due to oxidation and condensation of semi-volatile organic compounds) in our model.…”

“…CC BY 4.0 License. Konovalov et al, 2014;Reddington et al, 2016). In this study, the MODIS AOD data were used to constrain OC emissions and to optimize the calculated AOD values.…”