Geoscience Reference
In-Depth Information
cells was accompanied by their expansion and intensi-
fication, because of the enhanced Hadley cell circula-
tion caused by steeper temperature gradients (Flohn and
Nicholson, 1980). A logical consequence of this would
have been the southern extension of the Sahara arid zone
as indicated by early studies, which suggested low lake
levels in this region (Street and Grove, 1979; Servant
and Servant-Vildary, 1980). More recent evidence is less
equivocal, with some studies suggesting higher lakes in
parts of this region during the last glacial (Scholz
et al.
,
2007) and other studies suggesting enhanced dune activ-
ity (Lancaster
et al.
, 2002). This is a further example of a
growth in data leading to the emergence of a more complex
record of change. Offshore ocean-core records containing
dust layers (Sarnthein, 1978; Tjallingii
et al.
, 2008), and
in-blown terrestrial phytoliths (Abrantes, 2003), have also
yielded important evidence that correlate cold stages and
interstadials as times of arid conditions in the southern
Sahara and Sahel regions. Other records suggest that even
beyond the expanded arid zone, conditions were drier and
cooler in parts of tropical Africa (e.g. Livingstone, 1975;
Kadomura and Hori, 1990; Barker and Gasse, 2003).
At the last glacial maximum, tropical land areas were
about 2-6
◦
C cooler than today (e.g. Williams
et al.
, 2009).
The steeper thermal gradient led to higher wind speeds
(see Petit, Briat and Royer, 1981); therefore the Late
Pleistocene low-latitude deserts were probably windier
as well as colder than their modern counterparts (Hesse
and McTainsh, 1999). Over the oceans, this could have in-
creased evaporation, counteracting the effects of reduced
sea-surface temperatures on atmospheric moisture levels.
It has been suggested that stronger trade winds caused by
steeper pressure gradients also led to enhanced upwelling
of cold ocean waters in subtropical locations, further en-
hancing the aridity of coastal deserts such as the Namib
and Atacama (Williams
et al.
, 1993).
Other authors suggest several factors, however, that
could have mityigated seem to against this. First, the
stronger Hadley cell circulation resulted in a greater up-
welling of cold water in equatorial oceans (Hays, Imbrie
and Shackleton, 1976; Molina-Cruz, 1977; Prell
et al.
,
1980), so that tropical waters were up to 8
◦
C cooler than
today in the main areas of upwelling, potentially enhanc-
ing aridity in west coast deserts such as the Namib and
Atacama (Williams
et al.
, 1993). Second, the equator-
ward compression of the atmospheric circulation and the
expansion of subtropical highs resulted in the latter be-
coming even more persistent features than they are to-
day (see Brookfield, 1970; Kolla and Biscaye, 1977).
The penetration of northern hemisphere southwest trade
winds into west Africa and northwestern India would con-
eas. Third, lower glacial-age sea levels increased the con-
tinentality of areas with broad continental shelves, further
limiting the penetration of precipitation. In some areas,
e.g. northeastern Australia, this was a major contributory
factor in the spread of aridity (see Chappell, 1978). How-
ever, a further complication is added, at least in the south-
ern African context, by marine-core studies that suggest
through pollen analysis that the Namib Desert may have
been wetter during the late glacial due to the equator-
ward displacement of westerly trade winds (Shi
et al.
,
2001). Whether this enhanced impact of winter westerlies
occurred, or even penetrated into the southern African
interior (Chase and Meadows, 2007), is hotly debated
(Thomas and Burrough, 2011). Similar debates exist for
the location of the impact of westerlies on western South
America (Ammann
et al.
, 2001).
3.4.2
Drivers of late glacial tropical aridity
Late glacial aridity in north Africa and Australia is now
well documented and relatively well dated (see Gasse
et al.
, 2008; Williams
et al.
, 2009). In north Africa in par-
ticular, the onset of the Holocene brings with it an abrupt
step-change towards more humid conditions and the de-
velopment and expansion of lake bodies and increased flu-
vial activity (Street and Grove, 1979; Geyh and Thiedig,
2008; Drake
et al.
, 2008). Pokras and Mix (1985) pro-
posed a climatic model that equates north African trop-
ical aridity with phases of high latitude ice growth and
maximum humidity with deglaciated conditions at high
latitudes. More recently, attention has turned to the role
that insolation changes in the tropics may have played in
affecting rainfall changes. Thus in Africa a more common
hypothesis for humid/arid changes is via the influence of
the monsoons, which are in turn governed by precessional
variations in summer insolation (Anderson
et al.
, 1988;
Kutzbach and Liu, 1997; DeMenocal
et al.
, 2000).
Over shorter timescales, major wet-dry phases at low
latitudes either side of the equator in Africa, show a strong
coincidence with abrupt warm-cold events in high north-
ern latitudes, presumably related to ITCZ migrations. For
example, the younger Dryas cooling event at 11-10 000
BP, recognised in the climate records of the mid latitudes,
has now also been identified as causing a period of se-
vere low-latitude aridity, as evidenced from East African
lake-level records (Gasse, 2001).
On the Australian continent, environmental change dur-
ing the last glacial period was regionally variable but can
be broadly characterised by the expansion of the arid zone
in response to weakened monsoon rains in the northern
