Global mapping of maximum emission heights and resulting vertical profiles of wildfire emissions

Abstract. The problem of characteristic vertical profile of smoke released from wildland fires is considered. A methodology for bottom-up evaluation of this profile is suggested and a corresponding global dataset is calculated. The profile estimation is based on: (i) a semi-empirical formula for plume-top height recently suggested by the authors, (ii) satellite observations of active wildland fires, and (iii) meteorological conditions evaluated for each fire using output of the numerical weather prediction model. Injection profiles of the plumes from all fires recorded globally from March 2000 till November 2012 are estimated with a time step of 1 h. The resulting 4-dimensional dataset is split into daytime and nighttime subsets. The subsets are projected onto a global grid with a resolution of 1° × 1° × 500 m, aggregated to a monthly level, and normalised by total emissions in each vertical column. Evaluation of the obtained dataset was performed in several ways. Firstly, the quality of the semi-empirical formula for plume-top computations was evaluated using updated MISR fire Plume Height Project data. Secondly, the upper percentiles of the profiles are compared with an independent dataset of space lidar CALIOP. Thirdly, the results are compared with the distribution suggested for AEROCOM modelling community. Finally, the inter-annual variations of the calculated profiles are estimated.