Geoscience Reference
In-Depth Information
winds, they remain passive factors. Many authors have stressed the importance of
high wind velocities in mobilising dust from the ground surface (Maher et al.,
2010
),
but a more subtle interpretation involves wind gustiness (McGee et al.,
2010
).
McGee et al. (
2010
) noted that the glacial dust flux in the late Pleistocene was
roughly two to four times that of the interglacial dust flux in many parts of the semi-
arid world. They argued that enhanced glacial wind gustiness resulting from steeper
pole to equator thermal gradients was the dominant, or first-order, driver of such an
increase. They focussed their discussion on East Asia (north-western China), North
Africa (the Chad Basin) and southern South America. Using a threefold approach
(paleoenvironmental data; modern synoptic controls of dust-storms; global atmos-
pheric models), they considered and discounted a number of other possible causes of
enhanced dust flux, including changes in source area, sediment supply, plant cover,
aridity, atmospheric carbon dioxide concentration (pCO
2
) and sea level. However,
they did acknowledge some local influences involving sediment supply and vegetation
cover.
Modern dust-storms have thrown some light on this problem. Liu Tungsheng and
his colleagues (1985, pp. 149-157) have analysed the meteorological conditions under
which dust is mobilised from eastern Siberia and transported as far as eastern China,
Korea and Japan. Strong frontal winds, highwind velocities and powerful convectional
updrafts are associated with initial mobilisation and dust entrainment, while jet stream
activity seems implicated in long-distance transport.
A well-known modern dust-storm event in the Nile Valley is the
haboob
,ofwhich
there are three main types, each of which is associated with quite distinct synoptic
conditions (Griffiths and Soliman,
1972
, p. 93). Haboobs are associated with high
wind velocities (55 km/hour and more) and substantial turbulence, vindicating the
gustiness hypothesis, but their location is governed by a suitable supply of silt-sized
particles on the ground (Kendrew,
1961
, p. 71). Such dust-storms attain heights of at
least 1,500 m and advance along a sharp front some 25 km wide at a rate of about
55 km/hour.
The Australian dust record is considered in detail in
Section 9.7
, so it will be enough
to say here that peak dust flux coincided with the LGM in marine cores to the east
(Hesse,
1994
; Hesse et al.,
2004
) and south of Australia (Gingele and De Deckker,
2005
), as well as in the lunettes of the Willandra Lakes of arid western New South
Wa l es (Bowl er,
1998
). It also appears that the dust source areas have not remained
constant over time. For example, Hesse and McTainsh (
2003
) found that the northern
limit of the dust plume was 350 km, or 3
, north of the present limit during the interval
from 22 ka to 18 ka ago, indicating a major expansion in source area.
Nor should the role of aridity in preparing areas for deflation be minimised. The
Chad Basin was hyper-arid during the LGM and had a vastly expanded area available
for deflation, for which there is very strong local and regional evidence (Servant,
1973
; Servant and Servant-Vildary,
1980
; Hoelzmann et al.,
2004
). It seems very
°
