Numerical Dust Models - Mineral Dust: A Key Player in the Earth System - page 209

9.7 Regional Dust Models

Several regional (or mesoscale) models of the full cycle of dust emission, transport
and deposition exist for key regions, including the Sahara (e.g. Nickovic and
Dobricic 1996 ; Pérez et al. 2006 ; Heinold et al. 2007 ; Menut et al. 2009 )and
Asia (e.g. Shao and Wang 2003 ; Uno et al. 2006 ). Such regional-scale models that
operate with horizontal grid resolutions of 10-50 km were mostly developed for
the purpose of providing operational dust forecasts and are suited for simulation of
individual dust storm events (see Chap. 10 ) . Regional models are also useful for
comparisons and interpretation of in situ dust observations, made during intensive
field experiments. While the topography, soil conditions and small-scale extreme
wind events are described with more detail in regional compared to global models,
there are still uncertainties in describing the dust source area, surface conditions in
the source area and changes in size distribution and chemical composition during
transport.
A main purpose of regional dust models is the prediction of dust storms.
For East Asia and the Mediterranean region, such forecasts are compiled in the
World Meteorological Organization's Sand and Dust Storm Warning Advisory and
Assessment System (SDS-WAS; http://www.wmo.int/pages/prog/arep/wwrp/new/
Sand_and_Dust_Storm.html ; see also Chap. 10 ) . There, daily dust forecasts are
provided by more than 15 organizations in different geographic regions, integrating
research and user communities and including regional as well as global forecast
models. Such regional dust forecast models are also utilized in the context of field
experiments (like Fennec 2011 or the Saharan Mineral Dust Experiment (SAMUM)
2006-2011), as they are designed to simulate specific dust events and provide results
for simulated dust concentration, deposition fluxes and aerosol optical thickness
at scales that are suitable for comparison with in situ field observations. In turn,
such dust models provide necessary information for planning measurements (e.g.
research flights) during such field studies.
Differences in dust emission fluxes between different models can be due to the
use of different soil datasets as input data (Laurent et al. 2010 ) or the use of different
emission schemes (Kang et al. 2011 ). In addition, even at model grid resolutions of
10-50 km, meteorological features that impact on dust emission still need to be
parameterized, which leads to discrepancies due to the parameterization choices.
Darmenova and Sokolik ( 2007 ) point out that the choice of the planetary boundary
layer scheme considerably influences the computed emission fluxes and vertical
transport of dust. Intercomparison studies of regional-scale Asian dust models (Dust
Model Intercomparison study (DMIP); Uno et al. 2006 ) and an intercomparison of
models simulating a dust event in the Bodélé Depression (Todd et al. 2008 ) highlight
considerable differences in the modelled dust distributions. In Todd et al. ( 2008 ),
the performance of five regional-scale dust models was compared for a 3-day dust
event over the Bodélé Depression in Chad simulating dust distributions during the
BoDEX field experiment in 2005. While the models reproduce the diurnal cycle
in surface winds, the peak wind speeds are often too low in the models leading to