Geoscience Reference
In-Depth Information
developed by Peveril Meigs (1953), excluded regions too
cold for plant growth (and therefore the polar deserts,
notably Antarctica) and utilised Thornthwaite's (1948)
indices of moisture availability (
Im
):
that term proposed by Aubreville (1949). The delimita-
tion of the different types of dryland environments by
AI
values are dry-subhumid (
AI
=
0.50
−
<
0.65), semi-arid
(
AI
=
0.20
−
<
0.50), arid (
AI
=
0.05
−
<
0.20) and
hyper-arid (
AI
0.05). According to this scheme, these
four environments cover about 47 % of the global land
area (Figure 1.2). This is significantly more than the areas
considered in other schemes (Table 1.1), with this dif-
ference principally due to the inclusion of dry-subhumid
areas.
In this volume the global
arid zone
is considered to
include all elements of this fourfold classification. There
are several reasons for this:
=
<
Im
=
(100
S
−
60
D
)
/
PET
where
PET
is potential evapotranspiration, calculated
from meteorological data, and
S
and
D
are, respectively,
the moisture surplus and moisture deficit, aggregated on
an annual basis from monthly data and taking stored soil
moisture into account. Meigs (1953) identified three types
of arid environments, delimited by different
Im
index val-
ues: semi-arid, arid and hyper-arid. Grove (1977) sub-
sequently attached mean annual precipitation values to
the first two categories (200-500 mm and 25-200 mm,
respectively), though these are only approximate. Hyper-
arid areas with no consecutive months with precipitation
have been recorded (Meigs, 1953).
The UN (1977) delimitation of drylands is used as the
spatial framework in another desert geomorphology vol-
ume (Abrahams and Parsons, 1994). UN (1977) also pro-
vided the climatic input to the UNESCO (1979) survey of
arid lands utilised in the Cooke, Warren and Goudie (1993)
text. The UN (1977) approach defines aridity zones using a
P/PET
index.
PET
is calculated using Penman's formula,
which requires a large body of directly measured meteo-
rological data for its calculation and which in practice is
not consistently available at the global scale required for
dryland delimitation (Hulme and Marsh, 1990). A new
assessment of the extent of drylands, based on an arid-
ity index (
AI
), where
AI
1. The divisions between the elements of the classifica-
tions are somewhat arbitary. For example, in UN (1977)
the boundary between arid and hyper-arid areas was
taken as
P/PET
=
0.03, while in UNEP (1992) it is
0.05.
2. In studies prior to UNEP (1992) the climatic data input
to the calculation of indices was from nontimebounded
and therefore temporally variable data sets.
3. In drylands, annual precipitation frequently varies sub-
stantially from year to year, so that in dry-subhumid,
semi-arid, arid and hyper-arid areas, the only safe as-
sumption is that any year could be extremely arid
(Shantz, 1956).
4. Distinct geomorphic thresholds in terms of processes
and landforms have not been identified between the
four elements of the scheme.
P/PET
and
PET
is calculated
using the simpler Thornthwaite method, has been con-
ducted by Hulme and Marsh (1990) on behalf of UNEP.
This new assessment differs from earlier ones by using
meteorological data from a fixed time period (a 'time-
bounded' study), rather than simply from mean data from
the full length of records available, to calculate index val-
ues. This is significant given that climate variability can
cause mean data to differ depending on the period under
consideration (Hulme, 1992). The data used in this new
scheme, adopted and utilised in dryland studies such as
The World Atlas of Desertification
(UNEP, 1992), Thomas
(1993) and Thomas and Middleton (1994), cover the pe-
riod 1951-1980 and are based on records from over 2000
meteorological stations worldwide.
The UNEP (1992) classification of drylands also dif-
fers from previous estimates by including dry-subhumid
areas. This was done because these areas experience many
of the climatic characteristics of semi-arid areas (Thomas
and Middleton, 1994) and, with UNEP (1992) concerned
=
5. Climatic fluctuations and anthropegenic activities in
the twentieth century have caused the expansion of arid
surface conditions, especially a decrease in vegetation
cover, into some semi-arid environments.
6. Semi-arid areas are often called 'deserts' by their in-
habitants.
1.4
The age of aridity on Earth
Desert dune and evaporite sediments preserved in the
solid-rock record indicate that aridity has occurred on
Earth since Precambrian times (Glennie, 1987), with per-
haps the earliest recorded aridity being represented by the
c.1800-million-year-old dune sediments in the Hornby
Bay Group in the Canadian Northwest Territories (Ross,
