Stockholm University Press, Tellus B: Chemical and Physical Meteorology, 1(61), 2009
DOI: 10.3402/tellusb.v61i1.16831
Stockholm University Press, Tellus B: Chemical and Physical Meteorology, 1(61), p. 307, 2009
DOI: 10.1111/j.1600-0889.2008.00387.x
Full text: Download
The regional dust model system LM-MUSCAT-DES was developed in the framework of the SAMUM project. Using the unique comprehensive data set of near-source dust properties during the 2006 SAMUM field campaign, the performance of the model system is evaluated for two time periods in May and June 2006. Dust optical thicknesses, number size distributions and the position of the maximum dust extinction in the vertical profiles agree well with the observations. However, the spatio-temporal evolution of the dust plumes is not always reproduced due to inaccuracies in the dust source placement by the model. While simulated winds and dust distributions are well matched for dust events caused by dry synoptic-scale dynamics, they are often misrepresented when dust emissions are caused by moist convection or influenced by small-scale topography that is not resolved by the model. In contrast to long range dust transport, in the vicinity of source regions the model performance strongly depends on the correct prediction of the exact location of sources. Insufficiently resolved vertical grid spacing causes the absence of inversions in the model vertical profiles and likely explains the absence of the observed sharply defined dust layers.