Global top-down smoke aerosol emissions estimation using satellite fire radiative power measurements

Biomass burning occurs seasonally in most vegetated parts of the world, consuming large amounts of biomass fuel, generating intense heat energy, and emitting corresponding amounts of smoke plumes that comprise different species of aerosols and trace gases. Accurate estimates of these emissions are required as model inputs to evaluate and forecast smoke plume transport and impacts on air quality, human health, clouds, weather, radiation, and climate. Emissions estimates have long been based on bottom-up approaches that are not only complex, but also fraught with compounding uncertainties. Fortunately, a series of recent studies have revealed that both the rate of biomass consumption and the rate of emission of aerosol particulate matter (PM) by open biomass burning are directly proportional to the rate of release of fire radiative energy (FRE), which is fire radiative power (FRP) that is measurable from satellite. This direct relationship enables the determination of coefficients of emission (Ce), which can be used to convert FRP or FRE to smoke aerosol emissions in the same manner as emission factors (EFs) are used to convert burned biomass to emissions. We have leveraged this relationship to generate the first global 1° × 1° gridded Ce product for smoke aerosol or total particulate matter (TPM) emissions using coincident measurements of FRP and aerosol optical thickness (AOT) from the Moderate-resolution Imaging Spectro-radiometer (MODIS) sensors aboard the Terra and Aqua satellites. This new Fire Energetics and Emissions Research version 1.0 (FEER.v1) Ce product has now been released to the community and can be obtained from http://feer.gsfc.nasa.gov/, along with the corresponding 1-to-1 mapping of their quality assurance (QA) flags that will enable the Ce values to be filtered by quality for use in various applications. The regional averages of Ce values for different ecosystem types were found to be in the ranges of: 16–21 g MJ−1 for savanna and grasslands, 15–32 g MJ−1 for tropical forest, 9–12 g MJ−1 for North American boreal forest, about ~24 g MJ−1 for Russian boreal forest, and 18–26 g MJ−1 for Russian croplands and natural vegetation. The FEER.v1 Ce product was multiplied with FRP data to generate smoke TPM emissions, which were compared with equivalent emissions products from three existing inventories. The smoke TPM emissions results from FEER.v1 showed higher and more reasonable estimates than those of two other emissions inventories that are based on bottom up approaches and already reported in the literature to be too low, but portrayed an overall reasonable agreement with those of another inventory based on a hybrid method that includes the top-down approach, thereby suggesting that top-down approaches may hold better promise and need to be further developed to accelerate the reduction of uncertainty associated with fire emissions estimation in air-quality and climate research and applications. Based on analysis of data covering the period of 2004–2011, FEER.v1 results show that ~65–85 Tg yr−1 of TPM is emitted globally from open biomass burning, with a generally decreasing trend over this short time period. The FEER.v1 Ce product is the first global gridded product in the family of "emission factors", that is based essentially on satellite measurements, and requires only direct satellite FRP measurements of an actively burning fire anywhere to evaluate its emission rate in near real time, which is essential for operational activities, such as the monitoring and forecasting of smoke emission impacts on air quality.