Tropical cyclones - Extreme Events: A Physical Reconstruction and Risk Assessment - page 106

is little doubt about the veracity of the statistical methods and numerical mod-
els used in these types of analyses. However, when the observed cyclone fre-
quency and characteristics are extended over several centuries by incorporating
theprehistoric record of these events, there is a considerable difference in the
estimated return intervals of severe magnitude cyclones (storm tide inundation)
for the same region (Nott and Hayne, 2001;Nott, 2003).
Analysis of the prehistoric record near Cairns, Queensland show that during
theperiod AD 1800--1870, the latter being the date of first European settlement,
three cyclone events occurred producing storm tides between 2.52 and 4.51 m
AHD. The highest historic storm tide was 2.5 m. Because there have been no
inundations of this magnitude (>2.5 m) recorded or observed in this region
since AD 1870, the prehistoric evidence suggests that the incidence of severe
cyclones during the 19th Century was much higher than the 20th Century.
When these data are combined with the historic record of inundations, the
magnitude of the 100 year return interval storm tide (including wave set-up)
increases by approximately 1 m (Nott, 2003).
The last few centuries have seen a regime shift in the occurrence of tropical
cyclones crossing the coast in north Queensland. The causes behind this regime
shift are not yet known, but they are likely to be climatic. Even in the absence
of an understanding of the cause of the regime shift, it is still important for
this shift to be considered in risk assessments. Yet risk assessments of tropi-
cal cyclone impacts to date in this region have been entirely based upon the
historic record, and hence the present regime (Granger et al ., 1999). The impli-
cations of this approach for assessing building exposure to storm tide, amongst
other important factors, are potentially serious. Based upon the variability of
cyclones within the present regime, and at 1999 building patterns and levels,
approximately 4% of housing and accommodation buildings would be inundated
by storm tide during the 1 in 100 year event. Under the same scenario 17% of
business and industry buildings would be inundated. If a return to the previous
regime of cyclone variability were to occur, however, the percentage number
of housing and accommodation buildings and business and industry buildings
inundated by the 1 in 100 year event increases by greater than six times and four
times, respectively. Given that these cyclone regimes appear to be century scale
in length, it is entirely possible that a shift in regime may occur in the near
future.
Future developments in palaeotempestology
Sediment deposited by wind during tropical cyclones is difficult to dif-
ferentiate from normal aeolian deposits, or those deposited by winds generated