Geoscience Reference
In-Depth Information
coincides with, say, the 4,000 m contour and with a temperature of 10
Cinthe
warmest month, and if the former lower limit of periglacial solifluction deposits was
at 3,000 m, then the difference between the temperature at that elevation today and
10
°
C would give the amount of temperature increase since then. However, long-
term temperature measurements are often lacking in desert uplands, so an alternative
method involves estimating temperature lowering using plausible lapse rate values.
In the semi-arid Semien Highlands of Ethiopia, frost shattering ('gelifraction')
of the bedrock is only active today at elevations above 4,250-4,300 m (Hastenrath,
1974
; Williams et al.,
1978
). Frost-shattered angular rubble has been identified in these
mountains at elevations between 3,100 and 3,750 m (Hastenrath,
1974
; Williams et al.,
1978
). Because freezing temperatures would have been needed to shatter bedrock at
these lower elevations, the amount of temperature lowering requiredmay be calculated
using the present-day mean lapse rate of 0.6
°
C/100 m measured for the East African
highlands, including Ethiopia (Williams et al.,
1978
). The temperature lowering of
between 4
°
C estimated by this method is consistent with estimates of late
Pleistocene temperature lowering elsewhere in East Africa that are based on last glacial
snow-lines (Flint,
1959b
) and on pollen data (van Zinderen Bakker and Coetzee,
1972
).
In using periglacial deposits to reconstruct past changes in temperature, it is essen-
tial not to confuse such deposits with debris flows and angular colluvium, as noted
by Hurni (
1982
). Periglacial solifluction deposits lack sorting, contain angular clasts
with their long axes oriented downslope, have a porous and often fine-textured matrix
and are locally derived. They also tend to be coarser towards the surface and finer at
depth. In addition, such deposits become thicker and more widespread with increasing
elevation (Embleton and King,
1968
, p. 513; Davies,
1969
, pp. 32-35; Flint,
1971
,
pp. 275-277; Washburn,
1973
, pp. 189-193). In practice, most workers use a com-
bination of both glacial features (erosional and depositional) and periglacial features
to reconstruct past changes in the equilibrium snow-line and the changes in the upper
tree-line or limit of tussocky grassland, both of which are temperature controlled
(Hastenrath,
1972
; Hastenrath,
1973
; Hastenrath and Wilkinson,
1973
; Hastenrath,
1974
).
°
Cand8
°
13.7 The glacial record from Africa and the Near East
Messerli et al. (
1980
) summarised many years of pioneering work mapping glacial
landforms in the mountains of North and East Africa. They were able to determine
the approximate present-day equilibrium snow-line and the snow-line for the most
recent putative maximum glacial advance in the Atlas Mountains of north-west Africa,
Tibesti and the Hoggar Massif in the Central Sahara, the Semien and Bale Mountains
in Ethiopia, and Mount Kenya, Mount Kilimanjaro and the Ruwenzori Mountains
in East Africa. Messerli and Winiger (
1992
) revisited their earlier work and stressed
