Extreme events over time - Extreme Events: A Physical Reconstruction and Risk Assessment - page 254

precise to determine the exact years in which these events occurred; however,
thefact that they definitely occurred within a seventy year period, and probably
within a shorter time interval, suggests that this may have been a period of
enhanced cyclogenesis which has not occurred since. The implication is that if
long-term tropical cyclone records of higher resolution could be obtained for this
region, particularly in terms of covering the entire range of cyclone magnitudes,
then randomness and stationarity may not be apparent. Indeed, the fact that
these extreme cyclones did occur just prior to the start of the historical record
provides a test of the historical record and shows that the assumptions about
thefrequency distribution of tropical cyclones in this region are incorrect.
Long-term records of terrestrial floods certainly show that these events clus-
terinto phases of heightened and lesser activity. The luminescence line record
within corals of the Great Barrier Reef shows that terrestrial floods were con-
siderably larger between AD 1640 and 1760 and AD 1870 and 1910 with periods
of relative quiescence between AD 1830 and 1870 and again for much of the
1900s. The gauged flood record for this region starts around AD 1916, and, there-
fore, does not capture the true variability of flooding. At longer time scales, the
plunge pool record from the Northern Territory, Australia also shows variability
with heightened flood activity between 8000 and 4000 years BP and lower flood
activity (indeed flood magnitudes are about 20% lower on average) over the past
4000 years. While this is not critical for most planning timescales, it is critical
in this region because the catchment from which some of these records are
derived is mined for uranium, and tailings from this process are expected to
be stored in a stable environment for at least 1000 years. Nott and Price (1999)
showed that the enhanced phase of flooding between 8000 and 4000 years BP
corresponded to a different climatic regime, and one which is thought to be
similar to that expected under an enhanced greenhouse climate. It is now clear
that small shifts in average climatic conditions can result in substantial shifts
in flood magnitude and a return to a phase of greater flooding is entirely possi-
ble within the next few decades or century in the Northern Territory. So, even
though the timescale of prehistoric flood regime shifts may be relatively large,
they are certainly within the time frame, given the prospects of future climate
change, to be taken seriously when considering the safe storage of uranium
tailings.
Ely et al .(1993)andKnox (1993, 2003)havelikewisedemonstrated a clear
relationship between shifts in climate state and the magnitude of floods. Ely
et al .(1993)showed that floods in the southwest USA were relatively frequent
between about 5800 and 4200 years BP and after 2400 years BP, except for the
period between 800 and 600 years BP. An absence of, or rarely occurring, floods
dominated the period between 4200 and 2400 years BP. Knox (2003) concluded