Geoscience Reference
In-Depth Information
south-west (Braganza and Church,
2011
). As a result of these trends towards reduced
winter rainfall, the cities of Perth in the south-west and Adelaide in the south have
both invested in expensive desalinisation plants in order to secure their drinking water
supplies.
Concern over possible adverse impacts of future changes in climate has stimulated
research into historic trends in precipitation and evaporation (Suppiah et al.,
2006
;
Pittock,
2009
; Cleugh et al.,
2011
). The Australian Bureau of Meteorology has pro-
duced a series of maps showing changing regional patterns of mean annual precipita-
tion between 1900 and 2009 (Holper,
2010
). Summer rainfall in the tropical north-west
of the continent has increased progressively over the past hundred years and especially
over the past fifty years, while winter rainfall in the south-west, south and south-east of
the continent has decreased, especially since the mid-1970s. Although these changes
are more or less within the realm of natural rainfall variability within Australia, they
will soon lie outside those limits if current trends persist. We can say with confidence
that although the climatic changes evident across Australia do not necessarily
demand
anthropogenic forcing, they are certainly
consistent
with such forcing.
25.5 Possible impacts of future changes in climate on the desert world
Predicting the impacts of possible future changes in climate is based on three general
approaches, none of which is entirely satisfactory. One involves the use of global
atmospheric circulation models, which have become increasingly complex in the last
few decades. Another involves extrapolating from what we can reconstruct of past
climates. The third involves using current climatic trends.
A fundamental problem with all current climate models is their inability to make
credible predictions about changes in cloud cover and water vapour. Clouds are
highly dynamic and notoriously hard to model, and so are either omitted from climate
models or included in a highly schematic manner. Water vapour is the most potent of
the greenhouse gases, and is also very hard to model realistically. As a consequence,
current models are often more akin to sensitivity tests than otherwise when it comes to
predicting changes in regional and local precipitation, and do not really provide us with
watertight predictions that natural resource managers can use with confidence. It is
hard to see how this difficulty can be easily overcome. The unresolved cloud problem
also accounts for why existing models vary widely in predicting future changes in the
spatial and temporal distribution of precipitation but are all reasonably consistent in
predicting likely changes in temperature associated with various levels of greenhouse
gas emissions, since this is based on better-known aspects of atmospheric physics
(IPCC,
2007a
; Houghton,
2009
).
Extrapolating from inferred past climatic changes is equally fraught with uncer-
tainty, not least in terms of knowing whether present-day boundary conditions are the
same as they were in the past. These boundary conditions include the distribution of
