Geoscience Reference
In-Depth Information
6.8
Modelling
Accurate simulation of dust emission and transport is a large challenge for current
sentation of specific meteorological features will briefly be summarised. The key
process on the continental scale is the representation of the summertime monsoon
circulation and specifically the strength of the heat low. For northern Africa, many
coarse-resolution models struggle to represent the correct intensity and northward
propagation of the monsoon (e.g. Cook and Vizy
2006
), which is to some degree
related to accumulated effects of errors introduced through the parameterisation of
deep moist convection (Marsham et al.
2011
,
2013b
). In contrast, the synoptic-scale
dust outbreaks in winter and spring are usually satisfactorily reproduced (Shao et al.
2002
; Laurent et al.
2006
; Cavazos et al.
2009
;Shaoetal.
2010
; Alizadeh Choobari
et al.
2012a
). Synoptic-scale AEWs are also reasonably well represented in state-
of-the-art weather prediction models, although interactions with moist convection
and the land surface can lead to errors in both strength and propagation (Sander
and Jones
2008
;Shenetal.
2010
). However, systematic biases in the monsoon
circulation affect AEWs, too.
The largest deficits of current dust models are arguably associated with deep
moist convection and PBL processes. Accurately modelling haboob dust storms is
closely tied to a realistic representation of organised moist convection (Knippertz
et al.
2009b
; Reinfried et al.
2009
; Cavazos-Guerra and Todd
2012
; Marsham et al.
2011
). Practically all weather and climate models used for dust simulations have
grid spacings that require a parameterisation of deep convection. These are designed
to release convective instability within a model grid box and therefore do not
allow for the spatial separation of up- and downdrafts that are crucial to mesoscale
organisation and the formation of cold pools, which in reality can reach dimensions
of many grid boxes of a typical regional model (up to several 100 km). Other
issues are the sensitivity of convective initiation to topography and soil moisture and
the occurrence of convectively generated small-scale and short-lived wind events
that are not well represented by coarse-resolution models (Heinold et al.
2013
).
As a consequence, a key meteorological mechanism in areas like the Sahel is not
satisfactorily represented in most models (Fig.
6.10
) and sometimes not even in
analyses due to a lack of observations and difficulties of the model to generate the
parent convection (Knippertz et al.
2009a
).
While models often reproduce the large-scale pressure gradients that drive NLLJs
satisfactorily, they often struggle to capture the distinct diurnal cycle in near-surface
winds. Several studies have shown model weaknesses with temperature inversions
over arid areas being underestimated, leading to too little decoupling and therefore
too much vertical dispersion during the night, even in analysis data (Fiedler et al.
2013
). Simulations of NLLJs over the Sahara are sensitive to vertical resolution,
PBL scheme and roughness length (Todd et al.
2008b
), but less so to horizontal
Search WWH ::
Custom Search