Landslides - Extreme Events: A Physical Reconstruction and Risk Assessment - page 187

125

Gaussian
probability
density plot

100

75

50

25

1835

1837 1841

0
1825 1830 1835 1840 1845

Calendric age AD

Figure 7.3. Probability density plot of times of regional rock fall events (from
47 lichenometry sites), New Zealand. Note that there are three distinct events
here all separated by only 6 yr (from Bull and Brandon, 1998 ).

measurement is displayed most efficiently by using a probability density plot
(Fig. 7. 3 ).
Prominent peaks in the FALL distribution are viewed as a record of short-
lived events that caused the reworking of parts of the deposit. Narrow peaks in
the FALLdistribution are typically associated with multi-event rock fall deposits
with smaller standard deviations. Single event deposits have larger standard
deviations and produce broader peaks. Multiple events can also be characterised
by overlapping peaks. Probability densities are normalised to the same units,
being percent per millimetre, to facilitate comparisons between density plots.
Bull and Brandon identified three closely spaced events on New Zealand's South
Island (AD 1835, 1837 and 1841). These landslides were related to earthquakes
so the technique is also very useful for determining the long-term history of
major earthquake events. The advantages of using lichenometry to date prehis-
toric earthquakes over more traditional methods, such as measuring fault move-
ments, are that blind thrust faults and offshore subduction zones are very diffi-
cult to use for stratigraphic studies that show displaced rock and sediment units.
Earthquakes associated with fault movements in these settings will, because of
ground shaking, produce rock falls in mountainous areas.
It should be noted that most lichenometrical analyses are applied to an area
without thorough knowledge of variations in lichen growth with local micro-
climate, altitude, temperature and precipitation, duration of snow cover and