Geoscience Reference
In-Depth Information
used an expanded database and produced the North American Drought Atlas. A
similar approach by Cook et al. (
2010
) resulted in the Monsoon Asia Drought Atlas,
which was based on a network of 327 tree-ring chronologies. The results showed four
intervals of severe and widespread drought during the last 1,000 years, which they
named the 'Ming Dynasty drought' of 1638-1641, the 'Strange Parallels drought'
of 1756-1768, the 'East India drought' of 1790-1796 and the 'late Victorian great
drought' of 1876-1878. The PDSI is very useful for comparing the relative severity
of different historic droughts for which there is a detailed tree-ring chronology; it is
less useful for evaluating recent trends.
The PDSI calculates potential evaporation (E
pot
) from temperature data using the
empirical Thornthwaite equation (Thornthwaite,
1948
). However, a number of factors
besides temperature control E
pot
, including wind speed, humidity and near-surface
radiation. Despite early claims that historical droughts in the United States were on
the increase, detailed analyses in the 1990s showed no evidence of an increase in
drought frequency, intensity, magnitude or duration (Karl and Heim, 1990; Idso and
Balling,
1992
; Soule,
1993
). A more recent study by Sheffield et al. (
2012
) compared
the PDSI based on the Thornthwaite E
pot
equation (PDSI _Th) and the PDSI in which
E
pot
is estimated from the physically based Penman-Monteith equation (PDSI_PM).
The results show that PDSI _Th overestimates the global increase in drought, whereas
the PDSI_PM shows an increase in some areas and a decrease in others. These results
cast doubt on the IPCC (
2007a
) conclusion that 'more intense and longer droughts
have been observed over wider areas since the 1970s, particularly in the tropics and
subtropics. Increased drying linked with higher temperatures and decreased precipit-
ation has contributed to changes in drought'. One reason why the IPCC conclusion
may be badly flawed concerns the use of temperature to assess E
pot
using the PDSI
_Th index. If global temperatures are indeed increasing, then use of a drought formula
based on temperature will inevitably show an increase in drought frequency. Overall,
Sheffield et al. (
2012
) found little evidence of change in global drought over the past
sixty years, and they noted that severe droughts in the 1950s and 1960s took place well
before the rapid increase in global warming of the last few decades. The 2012 IPCC
report on extreme events is more circumspect in predicting drought trends, observing
that there has been undue reliance on the PDSI and possible overestimation of the
increase in regional and global droughts (Seneviratne et al.,
2012
).
23.11 Extreme events as partial analogues of early Holocene environments
Modern extreme events can sometimes provide insights into the nature of early Holo-
cene prehistoric environments in ways that are simply not possible using global atmo-
spheric circulation models. Let one example suffice. The year 1999 was unusually
wet in many regions, with severe floods in Australia, China and India and very high
Nile discharge. In central Sudan, the rainfall in July 1999 was unusually heavy and
