Articles | Volume 17, issue 11
https://doi.org/10.5194/acp-17-7111-2017
https://doi.org/10.5194/acp-17-7111-2017
Research article
 | 
15 Jun 2017
Research article |  | 15 Jun 2017

Impact of the choice of the satellite aerosol optical depth product in a sub-regional dust emission inversion

Jerónimo Escribano, Olivier Boucher, Frédéric Chevallier, and Nicolás Huneeus

Abstract. Mineral dust is the major continental contributor to the global atmospheric aerosol burden with important effects on the climate system. Regionally, a large fraction of the emitted dust is produced in northern Africa; however, the total emission flux from there is still highly uncertain. In order to reduce these uncertainties, emission estimates through top-down approaches (i.e. usually models constrained by observations) have been successfully developed and implemented. Such studies usually rely on a single observational dataset and propagate the possible observational errors of this dataset onto the emission estimates. In this study, aerosol optical depth (AOD) products from five different satellites are assimilated one by one in a source inversion system to estimate dust emission fluxes over northern Africa and the Arabian Peninsula. We estimate mineral dust emissions for the year 2006 and discuss the impact of the assimilated dataset on the analysis. We find a relatively large dispersion in flux estimates among the five experiments, which can likely be attributed to differences in the assimilated observation datasets and their associated error statistics.

Download
Short summary
Top-down estimates of mineral dust flux usually rely on a single observational dataset whose observational errors propagate onto the emission estimates. Aerosol optical depth from five satellites are assimilated one by one into a source inversion system over northern Africa. We find a relatively large dispersion in flux estimates among the five experiments, which can likely be attributed to differences in the assimilated observational datasets and their associated error statistics.
Altmetrics
Final-revised paper
Preprint