Geoscience Reference
In-Depth Information
despite continual improvements, models have key limita-
tions. Four key sources of uncertainty are:
It is not possible to make predictions of
future atmo-
spheric greenhouse gas concentrations
. As a result,
there will always be an envelope of climate projections
corresponding to the range of emission scenarios even
assuming perfect climate models.
Climate model limitations
impose uncertainty on the
climate projections. Projected precipitation is particu-
larly uncertain as this variable results from microscale
processes, which, at the resolution of GCMs, is repre-
sented indirectly from parameters that are resolved by
the models.
Table 24.1
Global mean surface temperature warming in
◦
C
for the multimodel ensemble mean from IPCC AR4 for (four time
periods relative to 1980-1999 and for three emission scenarios)
(after Meehl
et al.
, 2007).
SRES
2011-2030
2046-2065
2080-2099
2180-2199
A2
0.64
1.65
3.13
A1B
0.69
1.75
2.65
3.36
B1
0.66
1.29
1.79
2.10
Based on an ensemble mean of 14 leading IPCC AR4
GCMs, temperatures under the A2 scenario for the last
two decades of the twenty-first century are projected to
rise by more than 5.5 degrees in southern Africa (eastern
Botswana, southern Angola and northern Namibia) dur-
ing the September to November months. Over the Saha-
ran Empty Quarter of western Algeria, northern Mali and
eastern Mauritania, the warming is even greater, exceed-
ing 5 degrees between March and December and peaking
above 6 degrees between June and September inclusive,
nearly double the global mean increase. These arid to
hyper-arid regions are already the hottest on the continent
and among the hottest regions in the world.
The interior of Australia (between approximately
22-26 degrees south and 118-130 degrees east) is pro-
jected to warm by 4.6 or more degrees by the last two
decades of the twenty-first century. Peak warming of
5.2 degrees is projected for October. Similar, though
slightly higher, projections (peak 5.5 degrees between
June and August) are made for the interior of Saudi Arabia
while eastern Iraq and western Iran are projected to warm
by more than 5 degrees in between April and October,
with a peak of 6.7 degrees in July and an increase above
6 degrees for May to September inclusive.
At the global scale, the hydrological response to global
warming is reasonably simple (Figure 24.1, middle).
There is an increase in precipitation in the tropics, a cor-
responding increase in subsidence, decrease in cloudi-
ness (Figure 24.2) and therefore aridity in the subtropics.
The subtropical anticyclones also expand poleward (Fig-
ure 24.1, right). Most of the reductions in precipitation
in forthcoming decades results from increases in subsid-
ing air into the subtropics. The precise nature of regional
precipitation changes, while of clear importance to es-
tablishing future arid zone processes, are much harder
to pinpoint because of the uncertainty with which climate
models simulate precipitation and the resultant intermodel
differences in the projections. Figure 24.3 provides a more
Regional climate change
is less well modelled and un-
derstood than global change. One of the limitations of
current climate models is their relative inability to pre-
scribe detail on the regional scale as compared to their
abilities at the global scale; the IPCC has recognised
this as a prime area for continuing research. A key fea-
ture of the Fifth Assessment will be a dedicated regional
modelling exercise with resolutions at or better than 50
km. However, improvements in model resolution do not
necessarily equate to improvements in the fidelity of the
simulated climate.
Natural climate variability
will continue to exert an in-
fluence on future climate and will be relatively strong
compared with greenhouse gas forcing until about 2025.
Multiyear trends of warming or cooling, wetting or dry-
ing, have been observed in the past, and will continue
to be a feature in the future.
24.3
Overview of global climate change
projections in the context of arid zones
Projected changes in global mean surface temperature for
the twenty-first century (Table 24.1) are about 0.6
◦
Cfor
the period 2011-2030 (during these decades the change
is largely independent of the emission scenario) to 3.1
◦
C
for the last two decades for the high scenario. Changes are
large over land areas (compared with the ocean) and larger
still over some of the arid subtropics, particularly southern
Africa and the Sahara (northern Mali, southern Algeria
and eastern Mauritania) (Figure 24.1). Trends in observed
temperature extremes (e.g. increases in the occurrence of
warm nights) over the period 1951-2003 are also higher
in these parts of Africa than almost anywhere else in the
