Sensitivity of hourly Saharan dust emissions to NCEP and ECMWF modeled wind speed

Abstract

The impact of atmospheric aerosol concentrations on climate remains uncertain. This is due to the lack of understanding of the physical processes involved and also to uncertainties on the data used to model them. In the case of modeled dust emissions, the key parameters are the surface wind speed and the soil characteristics. Dust emissions are issued from inhomogeneous land surfaces and depend on local changes in wind speed. Nevertheless, their representation in models is crude since the local variability of the emission is not taken into account due to the model resolution. This simplified physical representation is problematic since the first emission process, saltation, is triggered by a threshold on the wind speed value. For most model studies, the global meteorological fields used are issued either from NCEP or ECMWF models, but the impact of the forcing was never quantified. In this study, we examine the emission fluxes variability based on the selected meteorological data set. The difference between the two emission fluxes can reach a factor of 3. This bias is not systematic and is highly variable on the timescale of an hour within different areas over western Africa. The impact of the sandblasting scheme is also studied; the variation arising by using two different schemes do not exceed 20%.