Bulletin of the Seismological Society of America, Vol. 102, No. 4, pp. 1514–1542, August 2012, doi: 10.1785/0120110170
Ⓔ
National Seismic Hazard Model for New Zealand: 2010 Update
by Mark Stirling, Graeme McVerry, Matthew Gerstenberger, Nicola Litchfield, Russ Van
Dissen, Kelvin Berryman, Philip Barnes, Laura Wallace, Pilar Villamor, Robert Langridge,
Geoffroy Lamarche, Scott Nodder, Martin Reyners, Brendon Bradley, David Rhoades,
Warwick Smith, Andy Nicol, Jarg Pettinga, Kate Clark, and Katrina Jacobs
Abstract
A team of earthquake geologists, seismologists, and engineering seismologists has collectively produced an update of the national probabilistic seismic
hazard (PSH) model for New Zealand (National Seismic Hazard Model, or NSHM).
The new NSHM supersedes the earlier NSHM published in 2002 and used as the hazard
basis for the New Zealand Loadings Standard and numerous other end-user applications. The new NSHM incorporates a fault source model that has been updated with
over 200 new onshore and offshore fault sources and utilizes new New Zealand-based
and international scaling relationships for the parameterization of the faults. The distributed seismicity model has also been updated to include post-1997 seismicity data,
a new seismicity regionalization, and improved methodology for calculation of the
seismicity parameters. Probabilistic seismic hazard maps produced from the new
NSHM show a similar pattern of hazard to the earlier model at the national scale, but
there are some significant reductions and increases in hazard at the regional scale. The
national-scale differences between the new and earlier NSHM appear less than those
seen between much earlier national models, indicating that some degree of consistency has been achieved in the national-scale pattern of hazard estimates, at least
for return periods of 475 years and greater.
Online Material: Table of fault source parameters for the 2010 national seismichazard model.
Introduction
Probabilistic seismic hazard models need to be updated
and improved with new data and methods whenever a significant improvement can be made. However, the scale and
complexity of the task of updating a national-scale model is
such that successive updates are often spaced by many years.
We hereby document a major update of the national probabilistic seismic hazard (PSH) model for New Zealand (National
Seismic Hazard Model, or NSHM). The new NSHM supersedes the earlier NSHM developed in 2000 (Stirling et al.,
2002) and used as the hazard basis for the New Zealand
Loadings Standard (Standard New Zealand, 2004) and
numerous other end-user applications. The update of the
NSHM is largely focussed on the source models, while the
choice of ground motion prediction equations and overall
PSH methodology used in the earlier NSHM have been preserved. The source model still comprises fault and distributed
seismicity components, but over 200 new onshore and
offshore fault sources (offshore faults were only minimally
considered in the earlier NSHM) and 11 additional years of
seismicity data have been included in the new NSHM. The
new NSHM is the product of a large team of earthquake
geologists, seismologists, and engineering seismologists reflecting the large multidisciplinary team-orientated approach
to seismic hazard modeling that has evolved in New Zealand
over the last decade.
Finally, it is important to note that this 2010 update
of the NSHM was completed prior to the occurrence of the
Mw 7.1, 4 September 2010 Darfield, and M w 6.2, 22
February 2011 Christchurch earthquakes. While the Greendale fault, source of the Darfield earthquake, has been added
to the fault source model, the large data set of earthquakes
and strong-motion records from the two events have not been
included in the model. Modeling of the aftershock-derived,
time-dependent hazard for the region of the two earthquakes
is the focus of considerable effort at the present time and will
represent a significant follow-up to this update of the NSHM.
General Approach
The probabilistic seismic hazard analysis (PSHA) methodology of Cornell (1968) forms the generic basis for
the NSHM. The steps undertaken for our PSHA were to
1514
National Seismic Hazard Model for New Zealand: 2010 Update
(1) use geologic data and the historical earthquake record to
define the locations of earthquake sources and the likely
magnitudes and frequencies of earthquakes that may be produced by each source, and (2) estimate the ground motions
that the sources will produce at a grid of sites that covers the
entire region.
Source Models
As is standard practice for PSH models, we use the
combination of a fault source model and distributed source
model as input to the NSHM. The fault source model uses the
dimensions and slip rates of mapped fault sources to develop
a single characteristic earthquake (magnitude and frequency)
for each fault source, while the spatial distribution of
historical seismicity is used to develop Gutenberg–Richter
magnitude–frequency estimates for the distributed seismicity
model Gutenberg–Richter (1944). The two source models
therefore combine to provide magnitude–frequency estimates from data representing a spectrum of time scales (prehistoric to historic) and magnitudes. The following Fault
Sources and Distributed Earthquake Sources sections describe the data and methodology used to develop the source
models. In construction of the fault and distributed seismicity
models, we limit our attention to providing mean or preferred
1515
parameters for the sources. In doing so we follow the same
approach taken in the earlier NSHM.
Fault Sources
The fault source model has been substantially updated
from that of the earlier NSHM (Fig. 1). New onshore geological mapping and interpretations, combined with a substantial contribution of offshore fault sources, have resulted in the
addition of over 200 new sources and the revision of many of
the existing sources.
The distribution of fault sources is shown in Figures 1b,
2, and 3, and the mean or preferred parameters and key
references for each source are listed and indexed in Ⓔ
Table S1 (available in the electronic supplement to this
paper). The fault sources shown on Figures 1b, 2, and 3 are
generalizations of mapped fault traces, which is appropriate
for regional-scale PSHA. All fault sources are modeled as discretized planes and assigned a general dip and dip direction
(Ⓔ see Table S1 in the supplement). The dips are either
obtained from seismic-reflection profiles (where available)
or inferred by taking into account surface exposures, surface
geology or geomorphology (e.g., fold geometries), and with
consideration of other fault geometries within similar tectonic settings. The model also includes nine blind faults
Figure 1. Comparison of the active fault sources in the earlier NSHM of (a) Stirling et al. (2002) model, and (b) those of the new model.
The fault sources are shown (black lines), along with the upper edges of subduction interface sources (blue lines), and relative plate motion
rate between the Australian and Pacific plates in the center of the country (black arrow).
1516
M. Stirling et al.
Figure 2. (a) Groupings of active fault sources into domains or regions they occupy in New Zealand. K–M F, Kapiti–Manawatu faults;
MFS, Marlborough fault system. (b) Sites where paleoearthquake information has been collected. Onshore (yellow) sites are trenches and
trenchlike natural exposures; offshore (orange) sites are high resolution seismic-reflection profiles. Only the upper plate (noninterface) active
faults are shown on these figures.
(upper edges inferred to be at 1 km depth), and 13 subduction
interface sources. The depth to the base of the faults is generally obtained from the depth distribution of seismicity (see
Distributed Earthquake Sources section). Most fault bases
are in the depth range 10–20 km, but notable variations include the westward deepening of faults interpreted to splay
from the Hikurangi subduction interface (faults in the Hikurangi subduction margin forearc, Barker et al., 2009; Wallace
et al., 2009; Mountjoy and Barnes, 2011), shallow depths
(∼8 km) within thin, hot crust of the Taupo rift (Bryan et al.,
1999), and the ∼35 km deep transition between the the Wairarapa fault and Hikurangi subduction interface (see source
index 345 in Ⓔ Table S1 in the supplement; Darby and
Beanland, 1992; Rodgers and Little, 2006).
Fault Source Parameterization. For simplicity of modeling
the large number of fault sources, we use a characteristic or
maximum magnitude model (e.g., Youngs and Coppersmith,
1985; Wesnousky, 1994; Stirling et al., 1996) to estimate the
likely maximum magnitude (M max ) and recurrence interval
of M max for each fault source. We therefore assume that
earthquakes smaller in size than the M max are modeled from
the distributed seismicity model, as was the case in the earlier
NSHM (Stirling et al., 2002). However, our methods of
developing these earthquake parameters have been substantially updated from the earlier approach developed by
Stirling et al. (1998) and again applied in Stirling et al.
(2002). M max is calculated from three regressions of moment
magnitude (M w ) on fault area (A, in km2 ). Equation (1) is for
global plate boundary strike-slip faults, and equations (2) and
(3) are regression equations for New Zealand earthquakes.
The equation of Hanks and Bakun (2002):
M w 3:07 4=3 log A;
(1)
is used for the Alpine fault (see Ⓔ type index 1 in Table S1 in
the supplement). The equation given in Villamor et al.
(2007):
Mw 3:39 1:33 log A;
(2)
is used for normal faults in volcanic and rift environments
(faults in and west of the Havre trough–Taupo rift) or inferred shallow normal faults (faults in the eastern Raukumara
Peninsula, Fig. 3a; see Ⓔ type index 2 in Table S1 in the
National Seismic Hazard Model for New Zealand: 2010 Update
1517
Figure 3. Active fault sources used as inputs for the probabilistic seismic hazard analysis (PSHA). The numbers beside each fault correspond to the number in Ⓔ Table S1 (see supplement). The county is split into five maps: (a), (b), (d), and (e) are at the same scale, but (c) is
at a larger scale to see the dense faults in the Havre trough–Taupo rift. Domain boundaries are shown in white, and white numbers refer to the
domain number. AR, Axial Ranges; BoP, Bay of Plenty; CH, Caswell high; CR, Chatham Rise; CS, Cook Strait; Fi, Fiordland; HT, Havre
trough; KI, Kapiti Island; KR, Kaikoura Ranges; RP, Raukumara Peninsula; SA, Southern Alps; SW, South Westland; TR, Taupo rift; WBa,
Wanganui basin; WBi, Wanganui Bight. (f) shows the detail of fault sources in the Wellington region, and (g) is an example area that shows
the relationship between active fault traces (red) and fault sources (black). The fault sources are numbered according to Ⓔ Table S1 (see
supplement).
(Continued)
1518
M. Stirling et al.
Figure 3.
supplement). Lastly, the equation given in Stirling et al.
(2008),
M w 4:18 2=3 log W 4=3 log L;
(3)
Continued.
where D is single-event displacement (mm), and SR is slip
rate (mm=yr). Displacement is calculated from seismic
moment M 0 by the equation of Aki and Richards (1980):
M0 μLWD;
(5)
is used for all remaining faults (Ⓔ see type index 3 in
Table S1 in the supplement); L is fault length in kilometers,
and W is fault width in kilometers). Equations (2) and (3)
were developed in unpublished, peer-reviewed studies (Berryman et al., 2001; Villamor et al., 2001) but were first published in the associated references shown previously.
Recurrence interval T in years is calculated from
where μ is rigidity modulus (assumed to be 3×
1011 dyn=cm2 ), L is fault length in centimeters, W is fault
width in centimeters (note different units from L and W
in equation 3), and D is single-event displacement in centimeters. M 0 is calculated from M w by the equation of Hanks
and Kanamori (1979):
T D=SR;
log M0 16:05 1:5M w :
(4)
(6)
National Seismic Hazard Model for New Zealand: 2010 Update
Figure 3.
Continued.
1519
1520
M. Stirling et al.
Figure 3.
Continued.
National Seismic Hazard Model for New Zealand: 2010 Update
The calculated (equations 1–6) values of D and T were
compared with those derived from paleoseismic studies. If
they differed significantly, the fault length (L) was adjusted
by shortening at segment boundaries, or lengthening by combining contiguous surface traces, until the mean calculated
values overlapped the paleoearthquake values. The previous
result was achieved through an expert panel review process,
in which experts with extensive local knowledge (most of
whom are coauthors) provided reality checks of the calculated values. This was initially achieved by four days of
round-table meetings, in which every fault source was
reviewed by the panel in series, and later by extensive review
of several draft versions of the fault source model. The
review process was completed in mid-2010.
All of the previously identified parameters apply to the
entire fault, and so embodied in the previous approach is
the assumption that the average coseismic displacement D
(derived as the average across the entire fault plane) is
equivalent to the surface coseismic displacement. While
surface coseismic D is often observed to be less than the
modeled subsurface D in well-instrumented earthquakes
(e.g., Wells and Coppersmith, 1994), we do not need to
make any subsurface-to-surface conversions for the NSHM.
This is because T is calculated from D and the slip rate
(equation 4) and both of these parameters are applicable
to the entire fault plane rather than just the surface trace.
Such conversions would instead be relevant in the case
of D being applied to fault displacement hazard analyses,
in which surface coseismic D is the parameter of fundamental importance.
Fault Source Domains. In this section we provide an overview of fault sources grouped according to the seismotectonic domains or regions they occupy in New Zealand
(Figs. 2a; 3), and as such, these domains bear a close resemblance to the seismotectonic zonation scheme used for the
distributed seismicity model (Fig. 4). Further details on the
individual faults can be found in the references cited in
Ⓔ Table S1 (see supplement). Figure 2b also shows sites
(typically trenches) where paleoearthquake information
has been collected. Domains 1 to 12 in Figures 2a and 3
are numbered sequentially in later parts of the text, along
with the Hikurangi and Fiordland subduction interfaces (domains 13 and 14). Fault numbers (e.g., 321) refer to the numbers in Ⓔ Table S1 (see supplement) and Figure 3.
1. Extensional Western North Island Faults: The western
North Island faults domain, located west of the active backarc Havre trough–Taupo rift (Figs. 2a; 3a,b), consists of two
sets of faults: (1) north-northwest-striking faults in the north
near Auckland, which formed early in the evolution of the
back-arc rift, and (2) northeast-striking faults south of Taranaki (westernmost land in Fig. 3a), which appear to form
the southern extension of the active back-arc rift and may
extend well to the east of where they are currently mapped
1521
(Fig. 3a,b; Giba et al., 2010). No faults in this domain have
ruptured in the historical period (post-1840 A.D.).
Slip rates are constrained for many faults (e.g., Pillans,
1990; Nodder, 1993; Townsend et al., 2010), but where
paleoseismic data are absent, rates were inferred from:
(1) geomorphic expression, (2) rates on contiguous alongstrike segments of the same fault, or (3) long-term (preQuaternary) rates.
The main differences between the new fault source model and the Stirling et al. (2002) model are the lengthening of a
number of the fault sources, especially the extension offshore
of faults in the southeast of the domain (National Institute
of Water and Atmospheric Research [NIWA], unpublished data).
The fault source model is likely to be incomplete in this
domain due to the difficulties in identifying active faults in
the soft, easily erodible Plio-Pleistocene sediments that underlie a significant portion of the domain (Edbrooke, 2005;
Townsend et al., 2008).
2. Extensional Havre Trough–Taupo Rift: This domain
represents the continental arc and back-arc of the Hikurangi
subduction margin, which widens northward as it transitions
from the Taupo rift to the southern Havre trough (Wright,
1993; Wysoczanski et al., 2010). There has been one major
historical surface rupturing earthquake on land and potentially one offshore. On land, the 1987 M w 6.5 Edgecumbe
earthquake ruptured the Edgecumbe fault (174), and several
other nearby faults (Beanland et al., 1989; Beanland et al.,
1990). The 1992 Mw 6.3 Bay of Plenty earthquake potentially ruptured a fault in the offshore Whakatane Graben
(Webb and Anderson, 1998).
Slip rates are well constrained for a number of on-land
faults (e.g., Berryman et al., 1998; Villamor and Berryman,
2001, 2006a; Villamor et al., 2007; Canora-Catalan et al.,
2008; McClymont et al., 2009). However, where paleoseismic data are absent, on-land slip rates were inferred: (1) from
faults along-strike, (2) as a proportion of the total extension
rate along transects across the rift, and (3) from geomorphic
expression. Offshore, rates are well constrained beneath the
continental shelf, from displacement of postglacial (< 20 ka)
markers on a dense grid of seismic-reflection profiles (Taylor
et al., 2004; Bull et al., 2006; Lamarche et al., 2006). In
water depths greater than 150 m, slip rates were estimated
on the basis of geomorphic expression and inferred proportion of the regional extension rate derived from kinematic
modeling by Wallace et al. (2004) (NIWA, unpublished data).
The most significant change from that of the earlier NSHM is the addition of ∼135 new faults in the offshore
Bay of Plenty. This is the result of a significant amount of
new mapping and reinterpretation of existing data (Lamarche
et al., 2006; Bull et al., 2006; NIWA unpublished data). As
well as the additional faults, some faults in the 2002 model
have been replaced with multiple, predominantly northeaststriking faults (e.g., Mayor Island 1, 2, 3, and 4 in the 2002
model; Fig. 1). On land, many of the fault traces have been
1522
Figure 4.
M. Stirling et al.
(a) Crustal seismicity of New Zealand. The map shows the distribution of M ≥ 4 earthquakes for the period 1964 to mid-2009,
for depths of 0–30 km (black circles) and for depths greater than 30 km (gray squares). The boundaries of zones defined for the new
regionalization are also shown. These are largely delineated from the Domains in Figure 2, along with additional considerations regarding
crustal seismicity patterns (black circles); (b) parameters associated with the regionalization scheme. The gray shades indicate zones of
similar slip type that are combined to provide more statistically stable seismicity parameters. The combinations of zones are distinguished
by the gray scales, and parameters TE (slip type), B (b-value of the Gutenberg–Richter relationship) and Mcutoff (maximum magnitude of
truncation for the Gutenberg–Richter relationship) are listed on the left of the map. See the text for further explanation; and (c) contour map
showing the N-values (number of events ≥ 4 for the period 1964–2009) per 0:1° × 0:1° cell for 0–30 km depth.
National Seismic Hazard Model for New Zealand: 2010 Update
mapped in more detail and a number of new faults added,
reflecting the large amount of new work on Taupo rift faults
(Nairn et al., 2005; Villamor and Berryman, 2006a,b;
Tronicke et al., 2006; Nicol et al., 2006; Villamor et al.,
2007; Berryman et al., 2008; Canora-Catalan et al., 2008;
Mouslopoulou et al., 2008; McClymont et al., 2008, 2009;
Nicol et al., 2009; Begg and Mouslopoulou, 2010; Nicol
et al., 2010; Villamor et al., 2011). The slip rates for many
faults have also been refined, the largest change being the
Rangipo fault (255, 256, 264, 268), which has reduced from
∼3 to ∼0:2 mm=yr (Villamor et al., 2007).
The fault source model is likely to be incomplete in the
areas of the Taupo and Okataina calderas. Thick vegetation
cover, thick volcanic deposits, and the presence of many
lakes are likely to have obscured some active faults and
produced the obvious gaps in fault sources in Figure 3b,c.
3. Contractional Kapiti–Manawatu Faults: The Kapiti–
Manawatu faults domain is located west of the southern
North Island axial ranges (Figs. 2a; 3a,b) and consists of reverse faults uplifting the eastern side of the Wanganui basin
(Fig. 3b). Many have associated hanging wall anticlines and
footwall synclines (Lamarche et al., 2005), and the onshore
faults are inferred to be blind reverse faults beneath anticlines
(Melhuish et al., 1996; Jackson et al., 1998). No faults in this
domain have ruptured the ground surface in the historical
period.
Slip rates are well constrained for only a few faults (e.g.,
Pillans, 1990; Nodder et al., 2007) and in the absence of
paleoseismic data were inferred for on-land faults by conversion of fold differential uplift rates to dip-slip rates. The offshore rates north of Kapiti Island (Fig. 3b) were determined
from displacement of a postglacial surface (∼10:8 ka) and
seafloor scarps assuming an inferred relict seafloor age of
6.5 ka. Rates south of Kapiti Island were inferred from consideration of bathymetric expression and regional geological
and GPS data.
The largest differences between this model and that of
the earlier NSHM are the addition of seven offshore faults,
which replace the earlier Wairau offshore fault source interpretation (Lamarche et al., 2005; Nodder et al., 2007; NIWA
unpublished data). Onshore the fault distribution has been
revised slightly, with many faults lengthened to include
adjoining structures in order to calculate displacements that
match observed field displacements.
4. North Island Dextral Fault Belt: The North Island
dextral fault belt or system (Beanland and Haines, 1998;
Mouslopoulou et al., 2007, 2008, 2009) is centered upon
the North Island axial ranges (Figs. 2a; 3a,b,c), and is a series
of range-parallel strike-slip faults that accommodate much of
the margin-parallel plate motion (Cashman et al., 1992; Van
Dissen and Berryman, 1996; Beanland and Haines, 1998;
Mouslopoulou et al., 2007; Nicol and Wallace, 2007). At
least two historical ground-surface rupturing earthquakes
have occurred in this domain: the 1855 M w 8.2 Wairarapa
1523
earthquake, which occurred on the Wairarapa (345) and Alfredton (321) faults (Grapes and Downes, 1997; Schermer
et al., 2004; Rodgers and Little, 2006), and the 1934
Mw 7.4 Pahiatua earthquake, which occurred on the Waipukaka and Saunders Road faults (326) (Schermer et al., 2004).
Slip rates have been constrained for many faults (e.g.,
Marden and Neall, 1990; Van Dissen and Berryman, 1996;
Begg and Van Dissen, 1998; Heron et al., 1998; Langridge,
Berryman, and Van Dissen, 2005; Mouslopoulou et al.,
2007; Little et al., 2009; Carne et al., 2011). Where paleoseismic data were absent, however, slip rates were assigned
from (1) rates on faults along-strike where possible, particularly in the central to northern area, (2) relative geomorphic
expression, and (3) inferred regional slip-rate budget (e.g.,
taking into account the decreasing margin-parallel component to the north). Offshore data in the Bay of Plenty only
constrain the dip-slip component of the displacement rate
(NIWA, unpublished data).
The main difference between the new model and that of
the earlier NSHM is the inferred continuity of faults alongstrike, especially in the central to northern area, and their extension offshore into the Bay of Plenty (Fig. 3a,c) in the north
(Lamarche et al., 2006: NIWA, unpublished data) and Cook
Strait (Fig. 3b) in the south (Pondard and Barnes, 2010).
5. Hikurangi Subduction Margin Forearc: This domain,
located east of the North Island axial ranges (Figs. 2a; 3a,b,
c), is the partly uplifted forearc (onshore–offshore) and accretionary wedge (offshore) portion of the Hikurangi subduction margin (Lewis and Pettinga, 1993; Barnes et al., 2010;
Paquet et al., 2011). Only structures interpreted to be rooted
in sufficiently strong rocks of the upper plate are included,
and so thrust faults from the Late Cenozoic frontal accretionary wedge are not included in this domain. The only fault to
have ruptured in the historical period is the onshore–offshore
Napier fault (242), which ruptured in the M w 7.8 1931
Napier earthquake (Hull, 1990; Kelsey et al., 1998).
Slip rates are well constrained for only a few faults (e.g.,
Kelsey et al., 1998; Barnes, Nicol, and Harrison, 2002; Langridge, Van Dissen, et al., 2005; Litchfield et al., 2007; Berryman et al., 2009; Mountjoy and Barnes, 2011), but where
paleoseismic data are absent were inferred by several methods. The majority of the Raukumara Peninsula (Fig. 3a)
normal faults were assigned low (< 1 mm=yr) slip rates in
accordance with the Repongaere fault (217, Berryman et al.,
2009). Slip rates for many of the faults beneath the continental shelf were constrained by postglacial (< 20 ka)
seismic-reflection markers. The rates for the remainder of the
offshore faults were inferred either from comparable structures along strike where possible, or by consideration of geomorphic and structural expression, available stratigraphic
constraints, and assumed total convergence rates (taking account of the increasing convergence rate to the north; Barnes
and Mercier de Lepinay, 1997; Barnes et al., 1998; Lewis
et al., 2004; Wallace et al., 2004; Barnes et al., 2010; Mountjoy and Barnes, 2011).
1524
The largest change from the Stirling et al. (2002) model
is the addition of about 30 offshore fault sources. This is
mainly the result of new data acquisition and interpretations
of active faulting (e.g., Barnes, Nicol, and Harrison, 2002,
Barnes, Lamarche, et al., 2010; Barnes and Nicol, 2004; Pedley et al., 2010; Paquet et al., 2011; Mountjoy and Barnes,
2011). Most of the faults in the 2002 model did not have sliprate estimates, and revisions have been made here to the few
that previously did. The interpreted sense of displacement of
some faults in the northwest of the Raukumara Peninsula has
changed from normal to reverse as a result of correlation with
seismically imaged offshore faults (NIWA, unpublished
data). A few faults in the 2002 model have been removed
because they are no longer considered to be active (e.g., the
Aorangi Anticline fault), are antithetic to other faults (e.g.,
the Kidnappers faults), or the calculated M max is < 6:0
(and these earthquakes are accommodated in the distributed
seismicity model) (e.g., some Raukumara Peninsula normal
faults).
Some active faults may be as yet unmapped in the central to southern part of the domain. This is attributed to complex Tertiary structure and presence of easily erodible
sediments in the area.
6. Extensional North Mernoo Fault Zone, Chatham
Rise: The entirely submarine North Mernoo fault zone is an
array of active continental extensional faults on the northwestern (Mernoo) slope of the Chatham Rise (Figs. 2a;
3d) (Barnes, 1994a, b). These reactivated faults lie adjacent
to the southern tip of the Hikurangi subduction zone and may
have formed by either lateral buckling of the crust (Anderson
et al., 1993) or flexural extension (Barnes, 1994a). The only
historical large magnitude earthquake in this domain is the
1965 Mw 6.1 Chatham Rise earthquake, which ruptured a
normal fault in the western part of the region (Anderson et al.,
1993).
Slip rates are not constrained for individual faults, but
low rates are assigned on the basis of an analysis of a
succession of Pliocene-Quaternary unconformities, which
demonstrated that extensional strain is unlikely to exceed
a few percent (Barnes, 1994a).
These fault sources replace the two previously modeled
in 2002 (faults 35 and 36 in the 2002 model), and have comparatively reduced slip rates.
7. Strike-Slip Marlborough Fault System: The Marlborough fault system domain lies east of the northern part of the
Alpine fault (387; Fig. 3d), in the transition zone from the
Hikurangi subduction margin in the north to the South Island
continental collision zone in the south (Fig. 2a). These faults
occur above the southern portion of the Hikurangi subduction interface, which at depths of < 30 km plays a minor role
in accommodating plate motions, relative to the Marlborough faults (Reyners et al., 1997; Reyners, 1998; Wallace
et al., 2012 ). Three faults have ruptured during the historical
period: (1) the Hope River segment of the Hope fault (409)
M. Stirling et al.
during the M w 7–7.3 1888 North Canterbury (Amuri) earthquake (McKay, 1888; Cowan, 1991); (2) the eastern segment
of the Awatere fault (357), during the M w ∼ 7:5 1848 Marlborough earthquake (Grapes et al., 1998; Mason and Little,
2006); and (3) the Poulter fault (422), during the Mw 7.0
1929 Arthur’s Pass earthquake (Berryman and Villamor, 2004).
Slip rates are well constrained on land from displaced
geomorphology (e.g., Freund, 1971; Knuepfer, 1992; Wood
et al., 1994; Benson et al., 2001; Nicol and Van Dissen,
2002; Langridge et al., 2003; Langridge and Berryman,
2005; Mason et al., 2006; Zachariasen et al., 2006; Van Dissen and Nicol, 2009). Where paleoseismic data are absent,
slip rates were inferred where possible from along-strike continuity of geomorphic expression. Offshore, only the Boo
Boo fault (383, 384) has a dextral rate inferred directly from
displaced submarine geomorphology (NIWA, unpublished
data). The remaining submarine dextral faults have rates assigned largely by extrapolation from contiguous onshore
traces, and consideration of the regional relative plate motion
budget (e.g., Stirling et al., 2008; Robinson et al., 2011).
The largest change to the Stirling et al. (2002) model is
the extension of the faults offshore into Cook Strait and eastern Marlborough (e.g., Barnes and Audru, 1999a, b; Barnes
and Pondard, 2010; Pondard and Barnes, 2010). One key
finding of these recent studies is confirmation that faults
do not directly link across Cook Strait with the North Island
dextral fault belt, that strike-slip faulting extends into the
southern Hikurangi margin, and that the offshore extent of
the Wairau fault is limited to about 40 km in Cook Strait
(cf. previous Wairau offshore source). Several scenarios
are accounted for in the new model for possible combinations of earthquake ruptures involving the Awatere, Vernon,
and Cloudy faults, and the Jordan, Kekerengu, Needles, and
Chancet faults. Onshore, a small number of faults have been
added, and some previous slip rates have been revised from
the 2002 model.
8. Contractional Northwestern South Island Faults: The
northwestern South Island faults domain, located north and
west of the central and northern Alpine fault, forms the western side of the South Island continental collision zone
(Fig. 2a). Many are former normal faults, now reactivated
as range-bounding reverse faults (Bishop and Buchanan,
1995; Ghisetti and Sibson, 2006). There have been two large
magnitude earthquakes in the historical period: (1) the
Mw 7.2 1968 Inangahua earthquake ruptured a blind fault
between two surface fault traces, while secondary ruptures
occurred on the Lyell (381) and Inangahua (386) faults (Anderson et al., 1994), and; (2) the White Creek fault (352)
ruptured in the M w 7.8 1929 Buller earthquake (Henderson,
1937; Berryman, 1980).
Slip rates are poorly constrained for these faults, in part
because of the heavily forested, mountainous terrain. Gentle
folding and minor faulting of river terraces, and other geomorphic expressions, however, indicate the slip rates are
National Seismic Hazard Model for New Zealand: 2010 Update
likely to be less than 1 mm=yr (Berryman, 1980; Suggate,
1987, 1989; Fraser et al., 2006), so low slip rates are inferred
for all of the faults.
The main change from the earlier NSHM is the extension
of faults along strike in order to match the calculated coseismic displacements with observed field displacements. There
have also been some minor revisions of slip rates.
9. Contractional North Canterbury Faults: The North
Canterbury faults domain lies in the southern part of the transition zone from the Hikurangi subduction margin to the
South Island continental collision zone (Figs. 2a; 3d). For
this study, the Marlborough slope thrust faults are included
in this domain, rather than domain 5, because they appear to
be associated with predominantly continental deformation
rather than active subduction (Barnes et al., 1998; Wallace
et al., 2009, Wallace et al., 2012). No faults in this domain
have ruptured the ground surface in the historical period.
Slip rates are generally poorly constrained (cf. Howard
et al., 2005), and where paleoseismic data are absent, slip
rates were largely inferred from their geomorphic expression
and presence of Pleistocene growth sequences. The slip rates
beneath the north Canterbury continental shelf were estimated from folding rates derived from Late Pleistocene unconformities (Barnes, 1996).
The largest change to the Stirling et al. (2002) model is
the addition of several offshore faults, particularly east of
Marlborough (e.g., Stirling et al., 2008). On land, many
faults have been reinterpreted, a few combined, and several
new faults are added in the southwest corner along the Canterbury foothills (GNS Science, unpublished data). The complex source model for the M w 6.7, 1994 Avoca earthquake
was treated as a fault source in the Stirling et al. (2002) model, but it has been removed from the present model. The exact
geometry of the source is ambiguous and is best accounted
for in the revised distributed seismicity model.
10. Contractional Southern South Island Faults: The
southern South Island faults domain, located east of the
central and southern Alpine fault, forms the southeastern side
of the South Island continental collision zone (Fig. 2a). Many
of the reverse faults bound ranges and basins and form two
predominant sets, trending northeast and northwest, which
probably reflect reactivation of basement normal faults (Norris and Turnbull, 1993; Ghisetti et al., 2007). There has been
one historical surface rupturing earthquake in this domain.
The Mw 7.1, 4 September 2010 Darfield, Canterbury, earthquake ruptured the Canterbury Plains in the northeastern corner of the domain (Quigley et al., 2010; Beavan, Samsonov,
Denys, et al., 2010).
Slip rates are well constrained for only a few faults (e.g.,
Beanland et al., 1986; Van Dissen et al., 1993; Litchfield and
Norris, 2000; Pace et al., 2005; Amos et al., 2007; Barrell
et al., 2009). Where paleoseismic data are absent, slip rates
were inferred from a combination of geomorphic expression
and long-term uplift rates (Bennett et al., 2005, 2006) and are
1525
within the bounds of ∼2 mm=yr total horizontal contraction
across the region (Norris, 2004; Wallace et al., 2007).
There have been two main changes from the faults
modeled by Stirling et al. (2002): (1) the lengthening of a
number of faults along-strike to match the calculated displacements with field observations, and (2) the addition of
faults, particularly in the east and offshore southwest Fiordland, as a result of new mapping (e.g., Pettinga et al., 2001;
Barnes et al., 2005; Sutherland, Berryman, and Norris,
2006). The most noteworthy recent addition is the Greendale
fault, the newly identified source of the M w 7.1, 4 September
2010 Darfield earthquake (Quigley et al., 2010). The earthquake also involved rupture of blind faults at the western and
central sections of the fault (Beavan, Samsonov, Motagh,
et al., 2010), not shown in our simple representation of
the Greendale source (Fig. 3d). A few previous faults have
also been removed (faults 6, 7, 19, 20, 31, 32, 36 of the 2002
model), because they are no longer considered active. Slip
rates for many faults have been refined, the largest change
being the reduction of slip rate on the Cape Balleny fault
(536, fault 21 in the 2002 model) from 25 to 5 mm=yr
(Lamarche and Lebrun, 2000; Lebrun et al., 2003), because
it is no longer considered to be associated with the southern
Alpine fault system (Sutherland, Berryman, and Norris,
2006; Sutherland, Barnes, and Uruski, 2006).
It is likely that additional, as yet unrecognized faults
occur in the rapidly uplifting and eroding Southern Alps,
where markers for mapping and measuring active deformation are not well preserved (e.g., Cox and Barrell, 2007).
There are also possibly other, as yet unidentified, fault
sources of low slip rate hidden beneath the thick sediments
of the Canterbury Plains (northeast corner of the domain),
which obscured the Greendale fault prior to the 4 September
2010 Darfield earthquake.
11. Alpine Fault: The Alpine fault, bounding the western
side of the Southern Alps and Fiordland (Figs. 2a; 3d,e), is
by far the longest fault in the South Island (Berryman et al.,
1992; Norris and Cooper, 2001; Barnes et al., 2005; Sutherland, Berryman, and Norris, 2006; Sutherland et al., 2007),
and is described as a separate domain. In this model the
Alpine fault is divided into three segments, the Resolution
(504), central (432), and northern (387) segments, but the
Wairau fault (376, domain 7) is also a contiguous segment
to the northeast (e.g., Berryman et al., 1992; Barnes and Pondard, 2010). The Alpine fault has not ruptured in the historical period.
Slip rates are well constrained, and decrease northward
from 31 mm=yr off central Fiordland (Barnes, 2009), to
23 mm=yr in southern Westland (Sutherland, Berryman, and
Norris, 2006), and to 14 mm=yr at the southern end of the
northern segment (Langridge et al., 2010).
Changes to the Alpine fault source modeled by the
earlier NSHM include revised structure, segmentation, and
slip rate of the southern, offshore portion of the fault, slight
1526
revisions to onshore slip rates, and the removal of overlapping sources north of the Hope fault.
12. Western Fiordland Margin and Caswell High: The
western Fiordland Margin and Caswell High domain is
located entirely offshore, west of Fiordland (Figs. 2a; 3e),
and includes two main groups: (1) westward-verging thrust
faults that accommodate northwest-southeast crustal shortening (Delteil et al., 1996; Lebrun et al., 2000; Wood et al.,
2000; Barnes, Davy, et al., 2002), and (2) normal faults along
the southeast edge of the Caswell High (Fig. 3e), which are
forming as a result of flexure of the Australian Plate beyond
the convergent deformation (Malservisi et al., 2003). While
there have been several recent large thrust earthquakes near
the Alpine fault in this region (see section 14, Fiordland subduction interface, and Van Dissen et al., 1994; Reyners and
Webb, 2002; Reyners et al., 2003; Robinson et al., 2003; Fry
et al., 2010; Beavan, Samsonov, Denys, et al., 2010), it is not
certain if any ruptured the seafloor in the Fiordland basin.
Individual fault slip rates are not well constrained.
Fiordland and Milford basin thrust slip rates are inferred
from displacement of late Pleistocene seismic stratigraphy
and regional shortening estimates (Barnes, Davy, et al.,
2002). Slip rates of the Caswell High normal faults are derived from an extension rate determined on a mid-Pliocene
(∼3 Ma) seismic marker.
This model represents a substantial revision of the earlier NSHM in this region. Together with the Fiordland subduction zone (518, 508), these faults replace five oblique-slip
reverse faults (faults 288 to 292 in the 2002 model).
13. Hikurangi Subduction Interface: One of the most important changes to the earlier NSHM is the treatment of the
Hikurangi subduction interface or megathrust, located on the
eastern side of the North Island (Fig. 3). When the earlier
NSHM was developed, less was known about the earthquake
potential of the Hikurangi subduction interface, and the parameter values assigned to it were uncertain. Since that time,
significant progress has been made in our understanding of
the seismogenic potential and tectonics of the megathrust
(see review in Wallace et al., 2009). In particular, geodetic
observations of contemporary interseismic coupling on the
megathrust and the occurrence of slow slip events at the
downdip termination of coupling beneath the North Island
suggest that the interface may produce earthquakes as large
as M w 9. A range of earthquake sizes may be produced by
the interface, given that the southern portion is strongly
coupled, and the northern portion currently shows aseismic
creep (Wallace et al., 2009).
To encompass the variety of opinion on the seismogenic
potential of the subduction thrust, an expert panel (encompassed by the coauthor list) was convened to develop and
weight a series of potential earthquake sources for the Hikurangi subduction interface. Based on the overall characteristics of the subduction margin, historical seismicity, and
distribution of interseismic coupling and slow slip events
M. Stirling et al.
(Wallace et al., 2009), the margin was divided into three different source lengths, each length with two coincident
sources defining end-member dimensions for the interface:
(1) southern Hikurangi margin (Wellington sources, 328 and
342), (2) central Hikurangi margin (Hawke’s Bay sources,
253 and 254), and (3) northern Hikurangi margin (Raukumara sources, 143 and 144). The sources are shown on
Figure 3 (as lines with teeth) and the parameters listed in
Ⓔ Table S1 (see supplement). The underpinning geodetic,
geological, and seismological data used to develop these
sources are provided in Wallace et al. (2009). A further
source based on a combination of the three larger sources
(Wellington, Hawke’s Bay, and Raukumara) is also defined
in the model (source 328b).
The southern Hikurangi source model is ∼220 km long
(extending along-strike from Cook Strait to Cape Turnagain)
and ∼60–140 km wide with a downdip rupture depth of 25 to
30 km and an updip rupture depth of 5–15 km. There was
healthy debate within the expert panel regarding the likely
updip limit of the interface sources. The 15-km depth limit
assumes that seismogenic rupture will not continue updip
beneath the Cretaceous Pahau Terrane (Reyners and
Eberhart-Phillips, 2009), while the shallower updip limit assumes that significant seismogenic rupture can propagate
updip until the unconsolidated, active portion of the accretionary wedge is reached (e.g., Wallace et al., 2009). Single-event displacements (D) derived with an equation of D as
a function of subduction source length (McCaffrey, 2008) for
the southern Hikurangi sources are in the range of 3–8 m,
while Ds of 9–22 m are defined for the combined source.
Long-term slip rates are taken to be 25 5 mm=yr and
interseismic coupling coefficients on the interface are 0:7 0:2 (Wallace et al., 2004, 2009). For the wider rupture model
(combining 5 km updip depth and 30 km downdip depth),
application of equations 5 and 6 give Mw 8:4, and with
the following equations:
T M 0 =M 0rate ;
(7)
M 0rate μLWSR;
(8)
and
(in which M0rate is the seismic moment rate, and SR is the
seismic slip rate; Brune, 1968; Wesnousky, 1986) a recurrence interval T of 1000 years is calculated. The narrower
rupture model (from 15 km to 25 km depth) produces
Mw 8.1 events and T of 550 years. Finally, the source combining all three of the larger sources has an M w 9.0 and T of
7050 years. All the previous estimates of T incorporate coupling coefficients estimated from the GPS modeling (Wallace
et al., 2009). Also, because the three coincident rupture models are intended to be used simultaneously in the NSHM, the
M0rate is assigned to each model according to a Gutenberg–
Richter-based weighting scheme. This balances the seismic
moment rate when used together in the fault source model
National Seismic Hazard Model for New Zealand: 2010 Update
and results in the lowest magnitude (M w 8.1) having the
shortest T, and the M w 9.0 having the longest T.
The methodology described for the southern Hikurangi
segment is also repeated for the northern and central Hikurangi segments, with the throughgoing M w 9.0 source (see the
previous discussion in Fault Sources) applied to these two
parts of the subduction zone as well. These segments are
each 200 km long and 68–101 km wide. In these cases, the
updip limit of the subduction earthquake rupture is 5 km and
the downdip limit varies from 15–20 km. The values of D
are in the range of 3–7 m based on McCaffrey (2008).
Convergence rates on the central Hikurangi margin segment
are 44 7 mm=yr and coupling coefficients are probably
low (0.1–0.3; Wallace et al., 2004, 2009), yielding an average T for events with M w 8.1–8.3 of 1100–1400 years.
Convergence rates at the northern Hikurangi margin are
54 6 mm=yr and coupling coefficients are also low
(0.1–0.3; Wallace et al., 2004). Thus, Mw 8.1–8.3 earthquakes are calculated with T of 900–1150 years. The lack
of historical or paleoseismic data constraining the recurrence
behavior of Hikurangi subduction interface earthquakes
makes the calibration of our Hikurangi source model difficult
at the present time.
Finally, the Kermedec trench is the continuation of the
Hikurangi subduction zone to the northeast of New Zealand,
and undoubtedly has considerable seismic potential. However, we assume that the contribution to on-land hazard in
New Zealand from Kermedec trench earthquakes will be
minor compared with the much closer Hikurangi subduction
zone sources.
14. Fiordland Subduction Interface: The Fiordland
subduction zone is located at the southwestern corner of
the South Island and has been the source of numerous large
earthquakes in the historical period. The most recent large
events are the M w 7.2, 10 August 2003 Secretary Island earthquake and the M w 7.6–7.8, 15 July 2009 Dusky Sound earthquake (Reyners et al., 2003; Fry et al., 2010; Beavan,
Samsonov, Motagh et al., 2010). Our characterization of
the Fiordland subduction interface sources is based on three
factors: (1) the extent of ruptures modeled for the 2003 and
2009 earthquakes (Fig. 3e, based on fig. 1 in Fry et al.,
2010); (2) a source linking the 2003 and 2009 ruptures,
and (3) a 550-km long source linking the 2009 source with
the Puysegur subduction interface. The methodology for
estimation of Mw and T for the four sources is essentially
the same as for the Hikurangi sources, using the orthogonal
component of plate motion minus the dip-slip component of
all upper-plate faults intersected along a roughly east–west
transect at the source latitudes. The resulting slip rate is
5–10 mm=yr (preferred over 8 mm=yr) at the latitude of the
2003 source, and 10–22 mm=yr (preferred over 16 mm=yr) at
the latitude of the 2009 source. Application of equations 7 and
8, the relationship of McCaffrey (2008), and a Gutenberg–
Richter-based weighting scheme to balance slip rate
across the overlapping sources produces M w 7.1 and T of
1527
200–350 years for the 2003 source, Mw 7.8 and T of 350–
700 years for the 2009 source, M w 7.9 and T of 778–
1550 years for the 2003 2009 source, and M w 8.6 and T
of 10,850–21,700 years for the Puysegur 2009 source.
We acknowledge that the long recurrence interval assigned
to the Puysegur source is only based on slip rate balancing
at the latitude of Fiordland. Shorter recurrence intervals
may apply well to the south of New Zealand where the subduction zone is the only active element of the plate boundary.
Distributed Earthquake Sources
In addition to defining the locations, magnitudes, and
frequencies of large (Mw 7–7.9) to great (Mw ≥ 8) earthquakes on the crustal faults and subduction zones, we also
allow for the occurrence of moderate-to-large earthquakes
(Mw 5 up to the maximum magnitude assumed for the
region) to occur on unmapped or unknown faults (e.g., the
Mw 7.1, 4 September 2010 Darfield earthquake). We refer to
this as the distributed seismicity source model. While our
overall approach is similar to that used in the earlier NSHM,
the input data and several important steps in the methodology
have been revised. The regionalization (zonation) scheme
used in the earlier NSHM has been substantially updated to
represent the current understanding of the distribution of
seismicity (Fig. 4a) and tectonics (Figs. 1–3) in the country
(Fig. 4b). The b-value of the Gutenberg–Richter distribution
is calculated on a regional basis (as for the earlier NSHM), but
using the maximum-likelihood method of Aki (1965) and
earthquake data from 1964 to early July 2009 (the earlier
NSHM used data to the end of 1997). The method used in
the older model was that of Weichert (1980), which allows
for combining multiple time periods with varying completeness levels. However, we now conclude that the larger uncertainties in the magnitudes, locations, and depths of the
older data do not justify use of the older data. To ensure
estimates of the b-value are based on a sufficient number
of earthquakes, the 16 crustal zones are combined into 5
zones of similar slip type (Fig 4b). This allows for larger data
samples than if the zones are used individually as done in the
earlier NSHM. Specifically, the b-value estimates are now
made on subcatalogs of between 658 and 1363 events.
Subcrustal zones are also defined to encompass the dipping
slabs of the Hikurangi and Fiordland subduction zones, and
normal-faulting mechanisms are assigned to these zones. The
a-values are calculated on a 0:1° × 0:1° grid, at five depth
levels (10, 30, 50, 70, and 90 km), and b-values are assigned
to the grid cells according to the b-value of the zone. In our
calculations we assume that the one M L magnitude scale
used extensively in the New Zealand catalog is equivalent to
Mw . While a New Zealand-based regression equation for
ML − Mw conversion is under development and therefore
unavailable at the present time, our methodology for obtaining distributed seismicity parameters from the cumulative
total of M ≥ 4 events since 1964 minimizes the impact of
ML − Mw differences in the PSH model. The impact of these
1528
differences on the hazard calculations would be negligible,
based on many past observations of how changes to source
models in high seismicity regions translate to only very small
changes in hazard estimates.
In the earlier NSHM, the gridded a-value was smoothed
using a 50-km Gaussian smoothing kernel. To optimize the
smoothing in the updated model, we compare smoothed seismicity models with three different smoothing kernels: (1) a
50-km Gaussian kernel; (2) a Stock and Smith (2002) adaptive Gaussian kernel; and (3) a density-based adaptive kernel
based on Werner et al. (2012). The models are retrospectively
tested using a learning period of 1840.0–1998.0 and a testing
period of 1998.0–1999.0. Using the information gain per
earthquake (Harte and Vere-Jones, 2005) to evaluate the
model forecasts, which is based on the likelihood of each
observed earthquake and the total number of events forecast
by the model, the 50-km Gaussian kernel model is significantly better for the period tested. We therefore use a 50-km
Gaussian to smooth the a-value in the updated model.
Additionally, we smooth the b-values across zone borders
using the same kernel in the same way as applied in the earlier NSHM.
The maximum magnitude, M cutoff , assigned to the
Gutenberg–Richter magnitude–frequency distribution of
each zone, and hence, each grid cell within each zone, is
revised. The revision is in consideration of the changes to
the fault source model, and our conclusions regarding incompleteness of the fault model. In the new model we apply a
uniform M cutoff of 7.2 for all zones, except the Taupo volcanic zone (TVZ, light gray zone in Fig. 4b), which is assigned
an Mcutoff of 6.5. A map showing the spatial distribution of
N-values (number of events ≥ M4 for the period 1964–
2009.5) per 0:1° × 0:1° grid cell is shown in Figure 4c.
An acknowledged problem with the current dataset is
the presence of restricted or assigned depths in the GeoNet
earthquake catalog for New Zealand (see Data and Resources). More than 75% of the events we use have assigned
depths of 0, 5, 12, or 33 km. The misfit of this assigned depth
to the true depth for each event is not well understood, despite significant past attempts to address the issue (McGinty,
2001). In the updated version of the model, we assign these
earthquakes equally to the 10-km and 30-km depth layers.
Lastly, we model the seismicity rates up to Mcutoff for
each grid cell without altering the source dimension for the
larger magnitudes. Although some hazard modeling methodologies assign fictitious faults to each grid cell by way of
Mw − L scaling relationships, experience has generally
shown these adjustments to have undetectable influence on
hazard in high seismicity regions.
Magnitude–Frequency Distributions
The magnitude–frequency distributions derived from the
source models are shown in Figure 5. We show the cumulative
annual rate of events (number of events per year ≥ M, in log
scale) as a function of magnitude, for the fault and distributed
M. Stirling et al.
Figure 5.
Magnitude–frequency distributions for the fault
model, distributed seismicity model, combined model, and the combined model of the earlier NSHM. See the text for further explanation (Magnitude–Frequency Distributions).
seismicity models and the combined model. We also show
for comparison the combined source model. For the fault
sources, the individual event rates and magnitudes are based
on the mean or preferred values given in Ⓔ Table S1 (see
supplement); the distributed seismicity rates are based on the
calculated a-values and b-values for all grid cells and depth
layers (Fig. 4b,c).
The magnitude–frequency distribution for the fault
source model shows a typical monotonic fall-off in cumulative frequency as a function of magnitude. The stepped shape
of the curve at M w greater than about 7.8 is due to the limited number of earthquakes of that magnitude range. At
magnitudes less than about M w 6 the slope of the fault distribution is subhorizontal, due to there being relatively few
fault-derived earthquakes at that magnitude range. In contrast, the distributed seismicity curve shows a sloping linear
trend from M w 5 to M w 7.2. This is a direct consequence of
the distributed seismicity events being calculated according
to the Gutenberg–Richter relationship. The combined curve
shows a generally linear slope throughout the magnitude
range of the graph, which is a typical observation globally
(e.g., Wesnousky et al., 1982; Stirling et al., 1996). The main
divergences from linearity are in the range of M w 7.2–8.0,
where the distribution shows downward steps in rates at
about Mw 7.2, and at about M w 8.0 where the rates show a
step to the right on the graph. This is in contrast with the
combined curve from the earlier NSHM, which shows an
slight upward bulge in rates relative to the new model at
about M w 7.3 and lower rates at M w greater than 8.0. The
shape of the combined distribution for the new model is controlled by: (1) the M cutoff for most of the distributed seismicity model being set to M w 7.2, therefore contributing
significantly to the combined distribution at M w less than
7.2; (2) the new modeling of crustal faults producing a slight
decrease in rates of Mw 7.2–7.5; and (3) the new Hikurangi
National Seismic Hazard Model for New Zealand: 2010 Update
subduction interface model producing an increase in rates for
Mw 8.0–9.0. The increase in maximum magnitude to M w 9.0
on the interface is clearly evident in Figure 5.
Ground-Motion Prediction Equations
The ground-motion prediction equations (GMPEs) used
in the NSHM for peak ground accelerations (PGAs) and 5%
damped acceleration response spectra in New Zealand crustal and subduction zone earthquakes are those developed by
McVerry et al. (2006). This is currently the only suite of
GMPEs available for New Zealand, and we have restricted
our analysis to this suite because the suitability of foreign
GMPEs for New Zealand conditions is presently unknown
(e.g., Next Generation Attenuation models, or NGAs). In
particular, concern surrounds application of the NGA site
condition metric (V S30 ) to New Zealand conditions, with resolution of this issue being the topic of considerable research
at the present time. Site classes are instead more descriptive
and geologically based (e.g., site class A, hard rock; B, rock;
C, shallow or stiff soil; D, deep or soft soil; McVerry et al.,
2006). Equations are given for both the larger of two
randomly oriented othogonal horizontal components of measured ground motion and the geometric mean of the components, and take account of the different tectonic types of
earthquakes in New Zealand. The equations also model
the faster attenuation of high-frequency earthquake ground
motions in the TVZv (domain 2) than elsewhere. Both the
crustal and subduction zone attenuation expressions have
been obtained by modifying overseas models for each of
these tectonic environments to better match New Zealand
data and to cover site classes that relate directly to those used
for seismic design in New Zealand codes.
The McVerry et al. (2006) model uses all available data
from the New Zealand strong-motion earthquake accelerograph network up to the end of 1995 that satisfy various
selection criteria, supplemented by selected data from digital
seismographs. A more up-to-date New Zealand model using
post-1995 data is not available at the present time. The seismographs provide additional records from rock sites, and
information on motions involving propagation paths through
the volcanic region, classes of data that are sparse in records
produced by the accelerograph network. The New Zealand
strong-motion dataset (prior to September 2010) lacks records in the near-source region, with only one record from
a distance of less than 10 km from the source and at magnitudes greater than M w 7.2. The New Zealand data used in the
regression analyses range in source distance from 6 to
400 km (the selected cutoff, in which the distance is the closest distance to the fault rupture plane) and in M w from 5.08
to 7.23 for PGA, with the maximum magnitude reducing to
Mw 7.09 for response spectra data. The required near-source
constraint is obtained by supplementing the New Zealand
dataset with overseas peak ground acceleration data (but not
response spectra) recorded at distances less than 10 km from
the source. Further near-source constraints are obtained from
1529
the overseas attenuation models, in terms of relationships
that are maintained between various coefficients that control
the estimated motions at short distances. Other coefficients
are fitted from regression analyses to better match the
New Zealand data.
The need for different treatment of crustal and subduction zone earthquakes is most apparent when the effects of
source mechanism are taken into account. For crustal earthquakes, reverse mechanism events produce the strongest
motions, followed by strike-slip and normal events. For subduction zone events, the reverse mechanism interface events
have the lowest motions, at least in the period range up to
about 1 s, while the slab events, usually with normal mechanisms, are generally strongest (McVerry et al., 2006).
The McVerry et al. (2006) GMPEs have been used in the
earlier NSHM (i.e., the GMPE was used prior to eventual publication) and in hundreds of hazard studies in New Zealand.
In particular, they have been used in the derivation of the
elastic site spectra in the new Loadings Standard for New
Zealand, NZS1170.5:2004 (Standards New Zealand, 2004).
Only one change has been implemented in the McVerry et al.
(2006) GMPE since being applied to the earlier NSHM, and
that is a correction to the site class term. Specifically, the
nonlinear site effect term has been been changed (McVerry
et al., 2006).
Hazard Analysis
Methodology
We use the locations, sizes, tectonic type or crustal
mechanism, and recurrence rates of earthquakes defined in
our source model to estimate the PSH for a grid of sites with
a spacing of 0.1 degrees in latitude and longitude. We use the
standard methodology of PSHA (Cornell, 1968) to construct
PSH maps. For a given site, we: (1) calculate the annual
frequencies of exceedance for a suite of ground motion
levels (i.e., develop a hazard curve) from the magnitude, recurrence rate, earthquake type, and source-to-site distance of
earthquakes predicted from the source model; and (2) estimate the maximum acceleration level that is expected to
be exceeded for a range of return periods. For each site, step
(1) is repeated for all sources in the source model, and step
(2) is calculated by summing the results of step (1) to give the
annual frequencies of exceedance for a suite of acceleration
levels and spectral periods at the site due to all sources. The
hazard calculations are made with the same FORTRAN code
used to develop the 2002 NSHM. This is our standard code
for PSHA in New Zealand and was the focus of considerable
in-house QA around the time of the 2002 NSHM.
We generally assume a Poisson model of earthquake
occurrence for the earthquake sources, in that earthquake
probabilities are based on the average time-independent rate
of earthquake occurrence on each fault. The three exceptions
to Poissonian modeling are the time-dependent probabilities
assigned to the Wellington (Wellington–Hutt Valley), Wair-
1530
arapa, and southern Ohariu fault sources as a result of the
“It’s Our Fault” research (Rhoades et al., 2010). New paleoseismic information on the activity of these faults accumulated over the last several years has resulted in revised
estimates for some, or all, of the following: (1) the timing
of the most recent rupture and several previous ruptures;
(2) the size of single-event displacements; and (3) the Holocene dextral slip rate. The conditional probability of rupture
of these faults has been evaluated in light of this new information, using a renewal process framework and four recurrence-time distributions (exponential, log-normal, Weibull,
and inverse Gaussian). Conditional probabilities for the next
100 years used in the new NSHM are: Wellington fault (Wellington–Hutt Valley segment), 11%; southern Wairarapa
fault, 1.3%–7.7%; and Ohariu fault, 3.9%–5.1% (Rhoades
et al., 2010).
Our calculation of ground motions follows the standard
practice of modern PSHA and accounts for the uncertainty in
estimates of ground motion from the McVerry et al. (2006)
GMPE in the calculation of PSH (up to 3 standard deviations
below and above the median). Only Mw 5.25 and greater are
included in the hazard analysis, consistent with the recommended usage of the McVerry et al. (2006) GMPE. In any
case, inclusion of earthquakes in the range of M w 5.0 to 5.25
does not have a perceptible impact on hazard in highseismcity New Zealand.
Because the McVerry et al. (2006) GMPE has separate
expressions for crustal earthquakes of different slip type (i.e.,
strike-slip, normal and reverse, and slip types intermediate
between these extremes), and for subduction interface, shallow subduction slab and deep slab earthquakes, we estimate
accelerations applicable to the slip type and tectonic environment of each earthquake source. Each fault source is
assigned a particular slip type, and the attenuation expression
for that slip type is used for the fault in the hazard calculations. In the case of the dipping subduction interface sources
we use the interface attenuation expression. For the distributed seismicity (point) sources, the slip type assigned to the
point source is the slip type of the enclosing seismotectonic
zone (Fig. 4b). For the deep zones we simply use the shallow
and deep slab expressions of the model, based on the observation that essentially all of the deep seismicity in the country is attributed to the dipping Hikurangi and Fiordland slabs.
Application of the volcanic path attenuation expression for
the Taupo volcanic zone (TVZ), which strongly reduces
accelerations with distance, is limited to faults and point
sources located in the TVZ (normal-volcanic zone in Fig. 4b).
We apply this to the whole path length for earthquakes in this
zone, rather than just the part of the path contained within the
TVZ zone.
Hazard Maps and Spectra
Hazard estimates are shown for the new NSHM in map
form (Fig. 6a–f), and we also compare these maps with
maps from the earlier NSHM, for the case of the 475-year
M. Stirling et al.
PGA hazard (Fig. 6g). Response spectra for the new NSHM
and earlier NSHM are also shown for the main centers
of Auckland, Wellington, Christchurch, and Dunedin
(Fig. 7a–d). All hazard maps and spectra are based on class
C (shallow soil) site conditions (see McVerry et al., 2006 for
definition of site classes). For each of the maps shown, we
calculate the hazard with the McVerry et al. (2006) GMPE
and with the GMPE uncertainty set to three standard deviations. Maps for PGA and spectral acceleration (0.5- and 1-s
periods) for 475-year and 2500-year return periods (approximately 10% and 2% probability of exceedance in 50 years)
are shown, with acceleration measured in units of g.
The 475-year PGA map shows a pattern of hazard that is
highest from the southwestern South Island to the eastern
North Island (the northeast-trending zone of PGA ≥ 0:5g).
Hazard in this northeast-trending zone reflects the greatest
concentration of active fault sources and seismicity in the
model. In particular, the Fiordland (southwestern-most South
Island), Alpine fault (western-central South Island), and
Hope fault (northern-central South Island) sources are most
responsible for this zone of high hazard in the South Island.
The crustal and Hikurangi subduction zone sources show a
strong influence on hazard in the eastern North Island.
Comparison of Figure 6a with the equivalent map from the
earlier NSHM (Stirling et al., 2002; Fig. 6g) shows that the
biggest change in hazard is in the southeast of the North
Island. The new Hikurangi subduction interface model
(see Fault sources) has obvious impact on the hazard in this
area, and highlights the importance of the new subduction
interface modeling. The new subduction interface modeling
more than counteracts the 50% reduction in the Wellington
fault (Wellington–Hutt Valley source) hazard from the
Poissonian-based estimates used in the earlier NSHM.
Otherwise, the pattern of hazard is similar at a national
scale, with hazard decreasing away from the axis of the country toward the far north and south. Similar patterns of hazard
are reflected by the 475-year spectral acceleration maps
(Fig. 6b,c ), and the 2500-year maps (Fig. 6e,f).
The graphs of site-specific response spectra for class C
(shallow soil) sites show hazard for 150-, 475-, 1000-, and
2500-year return periods, and the spectra are compared with
the corresponding spectra from the earlier NSHM (Fig. 7a–d).
For Auckland, the spectra have reduced significantly from
the earlier NSHM, reflecting the influence of the newly distributed seismicity model on the hazard estimates. The increased b-value in the new model (about 1.1 for the new
model versus less than 1 for the earlier NSHM) is the primary
reason for the reduced spectral accelerations in Auckland.
For Wellington, the only significant change to the spectra
from the earlier NSHM is a slight increase in hazard at spectral periods of 0.4 s and greater. This increase in longer period hazard is due to the changes to the Hikurangi subduction
interface sources. For Christchurch, hazard has increased for
periods less than about 0.6 s, reflecting the decreased b-value
in the new NSHM (less than 1 in the new NSHM and close to
1.2 in the older NSHM). Lastly, spectra for the city of
National Seismic Hazard Model for New Zealand: 2010 Update
1531
Figure 6. Seismic hazard maps for 475- and 2500-year return periods (10% and 2% probability of exceedance in 50 years) for class C
(shallow soil) site conditions: (a) peak ground acceleration (PGA) for 500-year return period; (b) spectral acceleration (SA) 0.5 s for 475-year
return period; (c) SA 1 s for 475-year return period; (d) PGA for 2500-year return period; (e) SA 0.5 s for 2500-year return period; (f) SA 1 s
(Continued)
for 2500-year return period; and (g) PGA for 475-year return period from the earlier NSHM for comparison.
1532
M. Stirling et al.
Figure 6.
Continued.
National Seismic Hazard Model for New Zealand: 2010 Update
1533
Figure 7. Response spectra for the cities of: (a) Auckland, (b) Wellington, (c) Christchurch, and (d) Dunedin. Spectra for 150, 475, 1000,
and 2500 years for class C (shallow soil) site conditions are shown in each case, along with the earlier NSHM spectra for comparison.
1534
Dunedin generally show a slight increase in hazard for the
new NSHM. This reflects a slightly decreased b-value in the
new NSHM (less than 1 in the new NSHM and greater than 1
in the earlier NSHM).
Deaggregation
Deaggregation graphs show the contributions to the
hazard estimates according to magnitude and source-to-site
distance; the graphs are typically used to develop design or
scenario earthquakes for specific sites. We show deaggregation graphs for Auckland, Wellington, Christchurch, and
Dunedin in Figure 8. Each graph shows the contributions
to the 475-year PGA, along with an additional graph showing
the 475-year 1-s spectral acceleration (SA) for Wellington.
The Auckland deaggregation plot (Fig. 8a) shows the
475-year PGA to be dominantly controlled by the distributed
seismicity sources at M w 5–6.8 and distances of less than
70 km. The minor contributions from the Wairoa north fault
(Mw 6.7 at 22 km; 4% contribution) and Kerepehi offshore
fault (M w 7.2 at 38 km; 2% contribution) can also be seen on
the graph.
Wellington’s 475-year PGA hazard is dominantly controlled by fault sources. Peaks representing the Wellington
fault (Wellington–Hutt Valley) (Fig. 8b; M w 7.5 at less than
1 km; 20% contribution), Ohariu south fault (M w 7.6 at 5 km;
20% contribution), and Wairarapa faults (Mw 8.1 at 17 km;
13% contribution) are most obvious on the plot. The
M. Stirling et al.
475-year SA 1-s graph for Wellington is also provided in
Figure 8c to show the contribution to hazard from two of the
southernmost Hikurangi subduction interface sources
(Mw 8.1, 8.4, and 9.0 at 23 km; combined contribution of
about 20%). The great earthquakes produced by these
sources contribute most significantly to long-period motions.
The 475-year PGA hazard for the city of Christchurch
(Fig. 8d) is dominated by the distributed seismicity model
(Mw 5–6.8 at distances of less than 50 km). Fault sources
at the margin of and beneath the Canterbury Plains also
contribute to the hazard. Specifically, the Springbank and
Pegasus 1 sources (453 and 449, respectively; Mw 7.0 at 27
and 22 km, respectively; 7% combined contribution), Ashley
fault (451; M w 7.2 at 30 km; 5% contribution), and Porters
Pass–Grey fault sources (446 and 447; Mw 7.5 at 44 km; 4%
contribution) dominate the hazard at the city. None of the
peaks on Figure 8d are due to the Greendale fault (450),
source of the M w 7.1, 4 September 2010 Darfield earthquake.
The preliminary recurrence interval assigned to that fault
source (minimum 16,000 years) is considerably longer than
the recurrence intervals for the previously cited fault sources.
The hazard for Dunedin is dominated by earthquakes
produced by the distributed seismicity model. The main fault
sources contributing to the model are the Akatore fault
(Fig. 8e; M w 7.4 at 13 km; contribution 14%), Taieri River
fault (Mw 7.1 at 47 km; contribution 6%), and Waipiata fault
(Mw 7.4 at 60 km; contribution 3%).
Figure 8. Deaggregation graphs for the cities of (a) Auckland, (b) and (c) Wellington, (d) Christchurch, and (e) Dunedin. In all cases
except for (c) the deaggregation is for 475 year PGA, for class C (shallow soil) site conditions. In the case of (c), the deaggregation is for
475 year 1 s SA to show the contribution of Hikurangi subduction interface sources at longer spectral periods.
National Seismic Hazard Model for New Zealand: 2010 Update
1535
Discussion
The new NSHM has involved a national-scale update of
the fault source and distributed source models, both in terms
of data and methods. The overall pattern of hazard remains
similar to that of the earlier NSHM, without the large differences seen between earlier NSHMs (compare Smith and
Berryman, 1986 with Stirling et al., 1998 and 2002). At the
national-scale the pattern of hazard therefore seems to have
achieved some degree of consistency, but with notable differences at the regional scale. Notable increases in hazard are
apparent in the eastern North Island due to the influence of
the Hikurangi subduction zone modeling and associated offshore upper plate fault sources; decreases are apparent in
Auckland due to the influence of the new distributed seismicity
model. Numerous combinations of decreases and increases in
hazard are apparent in many other parts of the country. The
continued use of the McVerry et al. (2006) GMPE means that
none of the differences in hazard are due to changes to the
choice of GMPE, except for the minor change to the site class
term (see Ground-Motion Prediction Equations).
The M w 7.1, 4 September 2010 Darfield, Canterbury,
earthquake occurred on a fault source not recognized prior
to the earthquake (Quigley et al., 2010). The Greendale fault
source has now been included in the fault source model (see
Ⓔ Table S1 in the supplement; source 450b). However, before this inclusion the earthquake was to some extent already
accounted for in the NSHM. The maximum magnitude
assumed in the distributed seismicity model for the area of
the earthquake is 7.2 (i.e., larger than the Darfield event;
Fig. 4b), and the combined annual rate of Darfield-size
events from the subset of distributed seismicity grid cells
located in the immediate vicinity of the earthquake (Fig. 9)
is around 1=11; 000. This is of similar order to the uncertain
field-based rate estimates for the Greendale fault (1=16; 000;
see Ⓔ Table S1 in the supplement; Quigley et al., 2010).
Lastly, the deaggregation for Christchurch (Fig. 8c) shows
that Mw 7–7.5 fault and distributed source-derived earthquakes at distances less than 40 km are important contributors to the hazard. Therefore, earthquakes similar to the
Darfield event feature prominently in the NSHM, and in fact
have already been identified as important scenario earthquakes for Christchurch in studies commissioned by the
territorial authority Environment Canterbury (Stirling et al.,
2001, 2008). Similarly, the size and location of the M w 6.2,
22 February 2011 Christchurch aftershock were also consistent with the distributed seismicity model for that area.
However, the ground motions produced by the earthquake
were considerably stronger in the near field (within 10 km)
than the motions expected for an earthquake of that size in
the McVerry et al. (2006) GMPE.
A significant issue limiting the applicability of the
NSHM to short return periods (certainly those less than the
475-year return period) is the availability of non-Poissonian
(time-dependent) earthquake probabilities. Non-Poisonnian
probabilities only exist for the major Wellington region faults
Figure 9. Simplified trace of the Greendale fault and the
distributed seismicity grid cells in the immediate vicinity. The accompanying magnitude–frequency distribution shows the combined earthquake rates for these cells (cumulative number of
events ≥ M), and the 1=16; 000 year maximum-rate estimate for
the Greendale fault from field data.
at the present time (Rhoades et al., 2010). As is the general
case for PSH models, the NSHM is not developed to provide
estimates of short return period hazard. This is because of the
following three reasons: (1) a lack of time-dependent earthquake probabilities for virtually all of the earthquake sources.
Obtaining time-dependent probabilities involves a considerable effort of data evaluation and processing, and is therefore
limited to faults that have been intensely studied, such as the
Wellington region faults for the “It’s Our Fault” project
(Rhoades et al., 2010); (2) a lack of statistical seismologybased time-dependent earthquake probabilities (e.g., aftershock probabilities), which is a topic of intense research at
the present time (e.g., Gerstenberger et al., 2007), and (3) the
inability to incorporate fault interaction physics and associated
probabilities into the NSHM with high confidence (e.g.,
Robinson and Benites, 2001). Items (2) and (3) are topics
of active research, testing, and evaluation, but are not yet ready
for implementation directly in the NSHM at the present time.
1536
Data quality and quantity with respect to the source
model have been greatly improved in the new NSHM, with
the inclusion of over 200 fault sources and 11 years of seismicity data. The largest contribution to the fault source model has been the inclusion of offshore fault data, mainly in the
offshore regions of domains 2, 3, 5, 6, 7, 9, and 12. Nevertheless, prominent gaps in the offshore coverage of fault
sources remain in some offshore areas of domains 1, 2 (far
northeast), 8, and 10. Future improvement to the onshore
component of the fault source model will benefit from studies that are able to identify active faults in areas of highly
erodible geological formations (e.g., domains 1 and 5), areas
of high sedimentation (e.g., domains 2 and 10), and areas of
thickly forested mountainous terrain (e.g., domain 8).
A major advance in the NSHM has been the use of geodetic data and models to develop a much-improved source
model for the Hikurangi subduction interface. However, we
acknowledge that this model is a simplified approximation of
the understanding of how plate motion is accommodated at
the subduction interface. The Hikurangi subduction interface
model will undoubtedly be improved in future. Future
geodetically orientated enhancements to the NSHM will also
arise from the inclusion of geodetic data in the calculation of
distributed seismicity rates, which are at present based solely
on the earthquake catalog.
A future update of the McVerry et al. (2006) GMPE is
justified by the lack of inclusion of any strong-motion data
post-1995 in the GMPE. In particular, inclusion of the strongmotion data from the Mw 7.1, 4 September 2010 Darfield,
Canterbury, earthquake and devastating M w 6.2, 22 February
2010 Christchurch aftershock will constitute significant
modifications to the database used for development of a new
GMPE in New Zealand. Near-field ground motions (less than
about 10 km distance) were unusually strong for the
Christchurch earthquake, relative to what was expected for
an earthquake of that size. Furthermore, relocation of the
earthquake catalog with the new nationwide 3D-seismic
velocity model (Eberhart-Phillips, 2010) is needed in order
to redefine the depths assigned to restricted events (see Distributed Earthquake Sources section).
Finally, we emphasize that the maps and spectra are based
on mean or preferred parameters for all of the sources and have
not included an analysis of the influence of parameter uncertainties on the hazard. Only the aleatory uncertainty in the
McVerry et al. (2006) GMPE has been incorporated into the
hazard analysis, as was the case for the earlier NSHM. Future
work will be to quantify the site-specific uncertainty in hazard
for the NSHM and enhance the model to include tools for
quantifiying short-term earthquake hazard (e.g., Gerstenberger et al., 2005; Rhoades and Evison, 2005). In particular,
focussed work on the aftershock sequence of the Mw 7.1,
4 September 2010 Darfield, Canterbury, earthquake, including the devastating M w 6.2, 22 February 2010 Christchurch
aftershock, is being given highest priority in this regard.
Occurrence of this major earthquake sequence in an area of
(formerly) low seismicity rates (northeastern area of domain
M. Stirling et al.
10), absence of known faults, and in the case of the Darfield
earthquake, featureless topography highlights the need for
better detection of low slip-rate faults that are buried beneath
thick sediments. The Canterbury earthquakes have clearly
shown that these low slip-rate faults can rupture with major
consequences. Finally, an even more challenging task is to
provide reliable short-term earthquake hazard estimates prior
to the initiation of a major earthquake sequence.
Conclusions
We have produced an update of the NSHM for New
Zealand, the first national-scale update since the NSHM of
Stirling et al. (2002). The model incorporates a fault source
model that has been updated with over 200 new onshore and
offshore fault sources and utilizes new New Zealand-based
and international scaling relationships for the parameterization of the faults. The distributed seismicity model has also
been updated to include post-1997 seismicity data, a new
seismicity regionalization, and improved methodology for
calculation of the seismicity parameters. Probabilistic seismic hazard maps produced from the new model show a
similar pattern of hazard to the older maps at the national
scale, but some significant reductions and increases in hazard
at the regional scale. The national-scale differences between
the new NSHM and earlier NSHM appear less than those seen
between earlier NSHMs, showing that some degree of consistency in the national-scale pattern of hazard estimates has
been achieved, at least for return periods of 475 years and
greater. While the new NSHM is by no means the final word
on seismic hazard in New Zealand, it represents a considerable advance in terms of data quality, quantity, methodology,
and evolution of a large multidisciplinary, multi-institutional
approach to PSH modeling in this country (New Zealand).
Data and Resources
Seismicity data are available on the GeoNet website
(www.geonet.org.nz; last accessed March 2012), and the
fault source and distributed seismicity models are available
in electronic form from Mark Stirling for use in research. All
maps and graphs are also available in electronic form from
Mark Stirling. Data on the active faulting in the Wairarapa
and Wellington regions are available from http://www.gw
.govt.nz/assets/council-publications/The_1855_Wairarapa
_Earthquake_Symposium_Proceedings_Volume_Web
_Version.pdf. Data on the all-day field trip to the Wairarapa
fault and 1855 rupture sites are avialable from http://
www.gw.govt.nz/assets/council-publications/1855%20Field
%20Trip%20guide.pdf (last accessed July 2010).
Acknowledgments
We wish to thank GeoNet (www.geonet.org.nz) for providing seismicity catalog data used in the distributed seismicity model, and numerous
people in GNS Science and NIWA who assisted with preparation of the active fault database used as input to the fault source model. Contributions to
National Seismic Hazard Model for New Zealand: 2010 Update
various expert panel meetings and additional information came from Ursula
Cochran, Roberto Basilli, and William Fry. The Foundation for Research
Science and Technology is thanked for financial support, and the manuscript
benefited from reviews by Mark Hemphill-Haley, an anonymous reviewer,
Annemarie Christophersen, and John Beavan.
References
Aki, K (1965). Maximum likelihood estimates of b in the formula log N a − b M and its confidence limits, Bull. Earthquake Research Institute 43, 237–239.
Aki, K., and P. G. Richards (1980). Quantitative Seismology: Theory and
Methods W. H. Freeman, San Francisco, California.
Amos, C. B., D. W. Burbank, D. C. Nobes, and S. A. L. Read (2007). Geomorphic constraints on listric thrust faulting: Implications for active
deformation in the Mackenzie Basin, South Island, New Zealand, J.
Geophys. Res. 112, B03S11, doi 10.1029/2006JB004291.
Anderson, H., S. Beanland, G. Blick, D. Darby, G. Downes, J. Haines,
J. Jackson, R. Robinson, and T. Webb (1994). The 1968 May 23
Inangahua, New Zealand, earthquake: An integrated geological,
geodetic, and seismological source model, New Zeal. J. Geol. Geophys. 37, 59–86.
Anderson, H., T. Webb, and J. Jackson (1993). Focal mechanisms of large
earthquakes in the South Island, New Zealand: Implications for the
accommodation of Pacific-Australia plate motion, Geophys. J. Int.
115, 1032–1054.
Barker, D. H. N., R. Sutherland, S. Henrys, and S. Bannister (2009).
Geometry of the Hikurangi subduction thrust and upper plate,
North Island, New Zealand, Geochemistry Geophysics Geosystems
10, Q02007, doi 10.1029/2008GC002153.
Barnes, P. M. (1994a). Continental extension of the Pacific Plate at the
southern termination of the Hikurangi subduction zone: The North
Mernoo Fault Zone, offshore New Zealand, Tectonics 13, 735–754.
Barnes, P. M. (1994b). Inherited structural control from repeated Cretaceous
to Recent extension in the North Mernoo Fault Zone, western
Chatham Rise, New Zealand, Tectonophysics 23, 27–46.
Barnes, P. M. (1996). Active folding of Pleistocene unconformities on the
edge of the Australian–Pacific plate boundary zone, offshore North
Canterbury, New Zealand, Tectonics 15, 623–640.
Barnes, P. M. (2009). Postglacial (after 20 ka) dextral strike slip of the offshore Alpine fault, New Zealand, Geology 37, 3–6.
Barnes, P. M., and J. -C. Audru (1999a). Quaternary faulting in the offshore
Flaxbourne and Wairarapa basins, southern Cook Strait, New
Zealand, New Zeal. J. Geol. Geophys. 42, 349–367.
Barnes, P. M., and J. -C. Audru (1999b). Recognition of active strike-slip
faulting from high-resolution marine seismic reflection profiles:
Eastern Marlborough fault system, New Zealand, Geol. Soc. Am.
Bull. 111, 538–559.
Barnes, P. M., and B. M. Mercier de Lepinay (1997). Rates and mechanics
of rapid frontal accretion along the very obliquely convergent
southern Hikurangi margin, New Zealand, J. Geophys. Res. 102,
24931–24952.
Barnes, P. M., and A. Nicol (2004). Formation of an active thrust triangle
zone associated with structural inversion in a subduction setting,
eastern New Zealand, Tectonics 23, TC1015, doi 10.1029/
2002TC001449.
Barnes, P. M., and N. Pondard (2010). Derivation of direct on-fault submarine paleoearthquake records from high-resolution seismic reflection
profiles: Wairau Fault, New Zealand, Geochemistry Geophysics
Geosystems 11, Q11013, doi 10.1029/2010GC003254.
Barnes, P. M., B. Davy, R. Sutherland, and J. Delteil (2002). Frontal
accretion and thrust wedge evolution under very oblique plate
convergence: Fiordland Basin, New Zealand, Basin Res. 14,
439–466.
Barnes, P. M., G. Lamarche, J. Bialas, S. Henrys, I. Pecher, G. L. Netzeband,
J. Greinert, J. J. Mountjoy, K. Pedley, and G. Crutchley (2010).
Tectonic and geological framework for gas hydrates and cold seeps
1537
on the Hikurangi subduction margin, New Zealand, Mar. Geol. 272,
26–48.
Barnes, P. M., B. M. Mercier de Lepinay, J.-Y. Collot, J. Delteil, and
J.-C. Audru (1998). Strain partitioning in the transition area between oblique subduction and continental collision, Hikurangi margin, New Zealand, Tectonics 17, 534–557.
Barnes, P. M., A. Nicol, and T. Harrison (2002). Late Cenozoic evolution
and earthquake potential of an active listric thrust complex above
the Hikurangi subduction zone, New Zealand, Geol. Soc. Am. Bull.
114, 1379–1405.
Barnes, P., R. Sutherland, and J. Delteil (2005). Strike-slip structure and
sedimentary basins of the southern Alpine Fault, Fiordland, New
Zealand, Geol. Soc. Am. Bull. 117, 411–435.
Barrell, D. J. A., S. A. L. Read, R. J. Van Dissen, D. F. Macfarlane, J. Walker,
and U. Rieser (2009). Aviemore: A dam of two halves, in Geological
Society of New Zealand & New Zealand Geophysical Society Joint
Annual Conference: Field trip guides I. M. Turnbull (Editor),
Geol. Soc. New Zeal. Misc. Publ. 128B, Oamaru, New Zealand,
23–27 November 2009, 30 pp.
Beanland, S., and J. Haines (1998). The kinematics of active deformation in
the North Island, New Zealand, determined from geological strain
rates, New Zeal. J. Geol. Geophys. 41, 311–323.
Beanland, S., K. R. Berryman, A. G. Hull, and P. R. Wood (1986). Late
Quaternary deformation at the Dunstan fault, central Otago, New
Zealand, Bull. Roy. Soc. New Zeal. 24, 293–306.
Beanland, S., K. R. Berryman, and G. H. Blick (1989). Geological
investigations of the 1987 Edgecumbe earthquake, New Zealand,
New Zeal. J. Geol. Geophys. 32, 73–91.
Beanland, S., G. H. Blick, and D. J. Darby (1990). Normal faulting in a
back-arc basin: Geological and geodetic characteristics of the
1987 Edgecumbe earthquake, New Zealand, J. Geophys. Res.
95, no. B4 4693–4707.
Beavan, J., S. Samsonov, P. Denys, R. Sutherland, N. Palmer, and
M. Denham (2010). Oblique slip on the Puysegur subduction
interface in the 2009 July M w 7 8 Dusky Sound earthquake from
GPS and InSAR observations: Implications for the tectonics of
southwestern New Zealand, Geophys. J. Int. 183, no. 3, 1265–1286.
Beavan, J., S. Samsonov, M. Motagh, L. Wallace, S. Ellis, and N. Palmer
(2010). The M w 7 1 Darfield (Canterbury) earthquake: Geodetic
observations and preliminary source model, Bull. New Zeal. Natl.
Soc. Earthquake Eng. 43, no. 4, 228–235.
Begg, J. G., and V. Mouslopoulou (2010). Analysis of late Holocene faulting
within an active rift using lidar, Taupo Rift, New Zealand, J. Volc.
Geotherm. Res. 190, 152–167.
Begg, J. G., and R. J. Van Dissen (1998). Whitemans Valley Fault:
A newly discovered active second order fault near Wellington,
New Zealand—Implications for regional seismic hazard, New Zeal.
J. Geol. Geophys. 41, 441–448.
Bennett, E. R., J. H. Youngson, J. A. Jackson, R. J. Norris, G. M. Raisbeck,
F. Yiou, and E. Fielding (2005). Growth of South Rough Ridge,
central Otago, New Zealand: Using in situ cosmogenic isotopes
and geomorphology to study an active, blind reverse fault, J. Geophys. Res. 110, B02404, doi 10.1029/2004JB003184.
Bennett, E., J. Youngson, J. Jackson, R. Norris, G. Raisbeck, and F. Yiou
(2006). Combining geomorphic observations with in situ cosmogenic isotope measurements to study anticline growth and fault propagation in central Otago, New Zealand, New Zeal. J. Geol.
Geophys. 49, 217–231.
Benson, A. M., T. A. Little, R. J. Van Dissen, N. Hill, and D. B. Townsend
(2001). Late Quaternary paleoseismic history and surface rupture
characteristics of the eastern Awatere strike-slip fault, New Zealand, Geol. Soc. Am. Bull. 113, 1079–1091.
Berryman, K. R. (1980). Late Quaternary movement of White Creek
fault, South Island, New Zealand, New Zeal. J. Geol. Geophys.
23, 93–101.
Berryman, K., and P. Villamor (2004). Surface rupture of the Poulter Fault in
the 1929 March 9 Arthur’s Pass earthquake, and redefinition of
1538
the Kakapo fault, New Zealand, New Zeal. J. Geol. Geophys. 47,
341–351.
Berryman, K. R., S. Beanland, A. F. Cooper, H. N. Cutten, R. J. Norris, and
P. R. Wood (1992). The Alpine Fault, New Zealand: Variation in
Quaternary structural style and geomorphic expression, Annales
Tectonicae VI, Special Issue: Supplement to Volume VI, 126–163.
Berryman, K., S. Beanland, and S. Wesnousky (1998). Paleoseismicity of
the Rotoitipakau fault zone, a complex normal fault in the Taupo volcanic zone, New Zealand, New Zeal. J. Geol. Geophys. 41, 449–465.
Berryman, K., M. Marden, A. Palmer, and N. Litchfield (2009). Holocene
rupture of the Repongaere Fault, Gisborne: Implications for
Raukumara Peninsula deformation and impact on the Waipaoa
Sedimentary System, New Zeal. J. Geol. Geophys. 52, 335–347.
Berryman, K., P. Villamor, I. Nairn, R. Van Dissen, J. Begg, and J. Lee (2008).
Late Pleistocene surface rupture history of the Paeroa fault,
Taupo Rift, New Zealand, New Zeal. J. Geol. Geophys. 51,
135–158.
Berryman, K. R., T. H. Webb, N. Hill, D. Rhoades, and M. Stirling (2001).
Seismic loads on dams—Waitaki system: Development of fault
scaling relationships and their application to fault source characterisation, Institute of Geological and Nuclear Sciences, GNS Science,
Lower Hutt, New Zealand, Client Report 2001/129.
Bishop, D. J., and P. G. Buchanan (1995). Development of structurally
inverted basins: A case study from the West Coast, South Island,
New Zealand, in Basin Inversion, J. G. Buchanan and P. G. Buchanan (Editors), Spec. Publ. Geol. Soc. Lond. 88, 549–585.
Brune, J. N. (1968). Seismic moment, seismicity and the rate of slip along
major fault zones, J. Geophys. Res. 73, no. 2, 777–784.
Bryan, C. J., S. Sherburn, H. M. Bibby, S. C. Bannister, and A. W. Hurst
(1999). Shallow seismicity of the central Taupo Volcanic Zone,
New Zealand: Its distribution and nature, New Zeal. J. Geol. Geophys. 42, 553–542.
Bull, J. M., P. M. Barnes, G. Lamarche, D. J. Sanderson, P. A. Cowie,
J. K. Dix, and S. Taylor (2006). High-resolution record of displacement accumulation on an active normal fault: Implications for models of slip accumulation during repeated earthquakes, J. Struct.
Geol. 28, 1146–1166.
Canora-Catalan, C., P. Villamor, K. Berryman, J. J. Martinez-Diaz, and
T. Raen (2008). Rupture history of the Whirinaki fault, an active
normal fault in the Taupo Rift, New Zealand, New Zeal. J. Geol.
Geophys. 51, 277–293.
Carne, R. C., T. A. Little, and U. Rieser (2011). Using displaced river
terraces to determine Late Quaternary slip rate for the central
Wairarapa fault at Waiohine River, New Zealand, New Zeal. J.
Geol. Geophys. 54, 217–236.
Cashman, S. M., H. M. Kelsey, C. F. Erdman, H. N. C. Cutten, and
K. R. Berryman (1992). Strain partitioning between structural
domains in the forearc of the Hikurangi subduction zone, New
Zealand, Tectonics 11, 242–257.
Cornell, C. A. (1968). Engineering seismic risk analysis, Bull. Seismol. Soc.
Am. 58, no. 5, 1583–1606.
Cowan, H. A. (1991). The North Canterbury earthquake of September 1
1888, J. Roy. Soc. New Zeal. 21, 1–12.
Cox, S. C., and D. J. A. Barrell (compilers) (2007). 1:250,000 Geological
Map 15, Geology of the Aoraki Area, Institute of Geological and
Nuclear Sciences, GNS Science, Lower Hutt, New Zealand, 71 pp.
+ 1 folded map.
Darby, D. J., and S. Beanland (1992). Possible source models for the 1855
Wairarapa earthquake, New Zealand, J. Geophys. Res. 97, no. B9
12375–12389.
Delteil, J., R. H. Herzer, M. Sosson, J. -F. Lebrun, J. -Y. Collot, and R. A.
Wood (1996). Influence of preexisting backstop structure on oblique tectonic accretion: The Fiordland margin (southwestern New
Zealand), Geology 24, 1045–1048.
Eberhart-Phillips, D. (2010). Establishing a versatile 3-D seismic
velocity model for New Zealand, Seismol. Res. Lett. 81, no. 6,
992–1000.
M. Stirling et al.
Edbrooke, S. W. (compiler) (2005). 1:250,0000 Geological Map 4, Geology
of the Waikato Area, Institute of Geological and Nuclear Sciences,
GNS Science, Lower Hutt, New Zealand, 68 pp. +1 folded map.
Fraser, J. G., A. Nicol, J. R. Pettinga, and M. R. Johnston (2006). Paleoearthquake investigation of the Waimea-Flaxmore fault system,
Nelson, New Zealand, in Earthquakes and urban development:
New Zealand Geotechnical Society 2006 Symposium, Nelson,
New Zealand, February 2006, 31, 59–67.
Freund, R. (1971). The Hope fault: A strike-slip fault in New Zealand, Bull.
New Zeal. Geol. Surv. 86,1–49.
Fry, B., S. C. Bannister, R. J. Beavan, L. Bland, B. A. Bradley, S. C. Cox,
W. J. Cousins, N. H. Gale, G. T. Hancox, C. Holden, R. Jongens,
W. L. Power, G. Prasetya, M. E. Reyners, J. Ristau, R. Robinson,
S. Samsonov, and K. J. Wilson and the GeoNet team (2010). The
Mw 7.6 Dusky Sound earthquake of 2009: Preliminary report, Bull.
New Zeal. Natl. Soc. Earthquake. Eng. 43, 24–40.
Gerstenberger, M. C., L. M. Jones, and S. Wiemer (2007). Short-term
aftershock probabilities: Case studies in California, Seismol. Res.
Lett. 78, no. 1, 66–77.
Gerstenberger, M. C., S. Wiemer, L. Jones, and P. Reasenberg (2005). Real
time forecasts of tomorrows earthquakes in California, Nature 435/
19, 328–331.
Ghisetti, F. C., and R. H. Sibson (2006). Accommodation of compressional
inversion in northwestern South Island (New Zealand): Old faults
versus new? J. Struct. Geol. 28, 1994–2010.
Ghisetti, F. C., A. R. Gorman, and R. H. Sibson (2007). Surface breakthrough of a basement fault by repeated seismic slip episodes: The
Ostler fault, South Island, New Zealand, Tectonics 26, TC6004, doi
10.1029/2007TC002146.
Giba, M., A. Nicol, and J. J. Walsh (2010). Evolution of faulting and
volcanism in a back-arc basin and its implications for subduction
processes, Tectonics 29, TC4020, doi 10.1029/2009TC002634.
Grapes, R. H., and G. Downes (1997). The 1855 Wairarapa, New Zealand,
earthquake: Analysis of historical data, Bull. New Zeal. Natl. Soc.
Earthquake. Eng. 30, 271–368.
Grapes, R., T. Little, and G. Downes (1998). Rupturing of the Awatere
Fault during the 1848 October 16 Marlborough earthquake,
New Zealand: Historical and present day evidence, New Zeal. J.
Geol. Geophys. 41, 387–389.
Gutenberg, B., and C. F. Richter (1944). Frequency of earthquakes in
California, Bull. Seismol. Soc. Am. 34, 185–188.
Hanks, T. C., and W. H. Bakun (2002). A bilinear source-scaling model for
M-log A observations of continental earthquakes, Bull. Seismol.
Soc. Am. 92, 1841–1846.
Hanks, T. C., and H. Kanamori (1979). A moment magnitude scale,
J. Geophys. Res. 84, 2348–2350.
Harte, D., and D. Vere-Jones (2005). The entropy score and its uses in earthquake forecasting, Pure Appl. Geophys. 162, 1229–1253, doi
10.1007/s00024-004-2667-2.
Henderson, J. (1937). The west Nelson earthquakes of 1929, New Zeal. J.
Sci. Tech. 19, 66–143.
Heron, D., R. Van Dissen, and M. Sawa (1998). Late Quaternary movement
on the Ohariu fault, Tongue Point to MacKays Crossing, North
Island, New Zealand, New Zeal. J. Geol. Geophys. 41, 419–439.
Howard, M., A. Nicol, J. Campbell, and J. R. Pettinga (2005). Holocene
paleoearthquakes on the strike-slip Porters Pass Fault, Canterbury,
New Zealand, New Zeal. J. Geol. Geophys. 48, 59–74, doi 10.1080/
00288306.2005.9515098.
Hull, A. G. (1990). Tectonics of the 1931 Hawke’s Bay earthquake,
New Zeal. J. Geol. Geophys. 33, 309–320.
Jackson, J., R. Van Dissen, and K. Berryman (1998). Tilting of active folds
and faults in the Manawatu region, New Zealand: Evidence from
surface drainage patterns, New Zeal. J. Geol. Geophys. 41, 377–385.
Kelsey, H. M., A. G. Hull, S. M. Cashman, K. R. Berryman, P. H. Cashman,
J. H. Trexler, and J. C. Begg (1998). Paleoseismology of an active
reverse fault in a forearc setting: The Poukawa fault zone, Hikurangi forearc, New Zealand, Geol. Soc. Am. Bull. 110, 1123–1148.
National Seismic Hazard Model for New Zealand: 2010 Update
Knuepfer, P. L. K. (1992). Temporal variations in latest Quaternary slip
across the Australian-Pacific plate boundary, north-eastern South
Island, New Zealand, Tectonics 11, 449–464.
Lamarche, G., and J. -F. Lebrun (2000). Transition from strike-slip faulting
to oblique subduction; Active tectonics at the Puysegur Margin,
South New Zealand, Tectonophysics 316, 67–89.
Lamarche, G., P. M. Barnes, and J. M. Bull (2006). Faulting and extension
rate over the last 20,000 years in the offshore Whakatane Graben,
New Zealand continental shelf, Tectonics 25, TC4005, doi 10.1029/
2005TC001886.
Lamarche, G., J. -N. Proust, and S. D. Nodder (2005). Long-term slip rates
and fault interactions under low contractional strain, Wanganui
Basin, New Zealand, Tectonics 24, TC4004, doi 10.1029/
2004TC001699.
Langridge, R. M., and K. R. Berryman (2005). Morphology and slip rate of
the Hurunui section of the Hope fault, South Island, New Zealand,
New Zeal. J. Geol. Geophys. 48, 43–57.
Langridge, R. M., K. R. Berryman, and R. J. Van Dissen (2005). Geometric
segmentation and Holocene slip rate of the Wellington fault New
Zealand: The Pahiatua Section, New Zeal. J. Geol. Geophys. 48,
591–607.
Langridge, R. M., J. Campbell, N. L. Hill, V. Pere, J. Pope, J. Pettinga, B.
Estrada, and K. R. Berryman (2003). Paleoseismology and slip rate
of the Conway segment of the Hope fault at Greenburn Stream,
South Island, New Zealand, Ann. Geophys. 46, 1119–1139.
Langridge, R. M., R. J. Van Dissen, U. A. Cochran, N. J. Litchfield,
K. R. Berryman, J. G. Begg, P. Villamor, D. Heron, A. Nicol,
and D. Townsend (2005). Active faulting and paleoearthquakes
in the Wairarapa and Wellington regions, in Proceedings, The 1855
Wairarapa Earthquake Symposium, 150 years of thinking about
magnitude 8+ earthquakes and seismic hazard in New Zealand,
8–10 September, 2005, J. Townend, R. Langridge, and A. Jones
(Editors), 237 pp.
Langridge, R. M., P. Villamor, R. Basili, P. Almond, J. J. Martinez-Diaz, and
C. Canora (2010). Revised slip rates for the Alpine fault at Inchbonnie: Implications for plate boundary kinematics of South Island,
New Zealand, Lithosphere 2, 139–152.
Lebrun, J. -F., G. Lamarche, J. -Y. Collot, and J. Delteil (2000). Abrupt
strike-slip fault to subduction transition: The Alpine Fault–
Puysegur trench connection, New Zealand, Tectonics 19, 688–706.
Lebrun, J. -F., G. Lamarche, and J.-Y. Collot (2003). Subduction initiation at
a strike-slip plate boundary: The Cenozoic Pacific–Australian plate
boundary, south of NZ, J. Geophys. Res. 108, no. B9, 2453, doi
10.1029/2002JB002041.
Lewis, K. B., and J. R. Pettinga (1993). The emerging, imbricate frontal
wedge of the Hikurangi Margin, in South Pacific Sedimentary
Basins P. B. Balance (Editor), Elsevier, New York, Sedimentary
Basins of the World 2, 225–250.
Lewis, K. B., S. E. Lallemand, and L. Carter (2004). Collapse in a Quaternary shelf basin off East Cape, New Zealand: Evidence for passage
of a subducted seamount inboard of the Ruatoria giant avalanche,
New Zeal. J. Geol. Geophys. 47, 415–429.
Litchfield, N. J., and R. J. Norris (2000). Holocene motion on the Akatore
fault, south Otago Coast, New Zealand, New Zeal. J. Geol.
Geophys. 43, 405–418.
Litchfield, N., R. Van Dissen, and A. Nicol (2007). Re-assessment of slip
rate and implications for surface rupture hazard of the Martinborough fault, South Wairarapa, New Zealand, New Zeal. J. Geol.
Geophys. 50, 239–243.
Little, T. A., R. Van Dissen, E. Schermer, and R. Carne (2009). Late
Holocene surface ruptures on the southern Wairarapa fault,
New Zealand: Link between earthquakes and the uplifting of beach
ridges on a rocky coast, Lithosphere 1, 4–28.
Malservisi, R., K. P. Furlong, and H. Anderson (2003). Dynamic uplift
in a transpressional regime: Numerical model of the subduction
area of Fiordland, New Zealand, Earth Planet. Sci. Lett. 206,
349–364.
1539
Marden, M., and V. E. Neall (1990). Dated Ohakean terraces offset by the
Wellington fault, near Woodville, New Zealand, New Zeal. J. Geol.
Geophys. 33, 449–453.
Mason, D. P. M., and T. A. Little (2006). Refined slip distribution and
moment magnitude of the 1848 Marlborough earthquake,
Awatere fault, New Zealand, New Zeal. J. Geol. Geophys. 49,
375–382.
Mason, D. P. M., T. A. Little, and R. J. Van Dissen (2006a). Refinements to
the paleoseismic chronology of the eastern Awatere fault from
trenches near Upcot Saddle, Marlborough, New Zealand, New Zeal.
J. Geol. Geophys. 49, 383–397.
McCaffrey, R. (2008). Global frequency of magnitude 9 earthquakes,
Geology 36, 263–266.
McClymont, A. F., A. G. Green, P. Villamor, H. Horstmeyer, C. Grass, and
D. C. Nobes (2008). Characterization of the shallow structures of
active fault zones using 3-D ground-penetrating radar data,
J. Geophys. Res. 113, B10315, doi 10.1029/2007JB005402.
McClymont, A. F., P. Villamor, and A. G. Green (2009). Fault displacement
accumulation and slip rate variability within the Taupo Rift
(New Zealand) based on trench and 3-D ground-penetrating radar
data, Tectonics 28, doi 10.1029/2008TC002334.
McGinty, P. (2001). Preparation of the New Zealand Seismicity Catalog for a
probabilistic seismic hazard analysis of New Zealand, Bull. N. Z.
Soc. Earthq. Eng. 34, 60–67.
McKay, A. (1888). Remarks on the earthquakes in the Amuri, Trans.
New Zeal. Inst. 21, 269–293.
McVerry, G. H., J. X. Zhao, N. A. Abrahamson, and P. G. Somerville (2006).
Response spectral attenuation relations for crustal and subduction
zone earthquakes, Bull. New Zeal. Soc. Earthquake Eng. 39, 1–58.
Melhuish, A., R. Van Dissen, and K. Berryman (1996). Mount StewartHalcombe Anticline: A look inside a growing fold in the Manawatu
region, New Zealand, New Zeal. J. Geol. Geophys. 39, 123–133.
Mountjoy, J. J., and P. M. Barnes (2011). Active upper plate thrust faulting in
regions of low plate interface coupling, repeated slow slip events,
and coastal uplift: Example from the Hikurangi Margin, New Zealand, Geochem. Geophys. Geosyst. 12, Q01005, doi 10.1029/
2010GC003326.
Mouslopoulou, V., A. Nicol, T. A. Little, and J. J. Walsh (2007). Termination
of large strike-slip faults: An alternative model from New Zealand,
in Tectonics of Strike-Slip Restraining and Releasing Bends, W. D.
Cunningham and P. Mann (Editors), Spec. Publ. Geol. Soc.
Lond.290, 387–415.
Mouslopoulou, V., A. Nicol, J. J. Walsh, D. Beetham, and V. Stagpoole
(2008). Quaternary temporal stability of a regional strike-slip
and rift fault intersection, J. Struct. Geol. 30, 451–463.
Mouslopoulou, V., A. Nicol, T. A. Little, and J. G. Begg (2009). Palaeoearthquake surface rupture in a transition zone from strike-slip to
oblique-normal slip and its implications to seismic hazard, North
Island Fault System, New Zealand, in Palaeoseismology: Historical and Prehistorical Records of Earthquake Ground Effects for
Seismic Hazard Assessment L. Reicherter, A. M. Michetti, and
P. G. Silva (Editors), Spec. Publ. Geol. Soc. Lond. 316, 269–292.
Nairn, I. A., J. W. Hedenquist, P. Villamor, K. R. Berryman, and P. A. Shane
(2005). The ∼AD1315 Tarawera and Waiotapu eruptions,
New Zealand: Contemporaneous rhyolite and hydrothermal eruptions driven by an arrested basalt dike system?, Bull. Volcanol.
67, 186–193.
Nicol, A., and R. Van Dissen (2002). Up-dip partitioning of displacement
components on the oblique-slip Clarence fault, New Zealand, J.
Struct. Geol. 24, 1521–1535.
Nicol, A., and L. M. Wallace (2007). Temporal stability of deformation rates:
Comparison of geological and geodetic observations, Hikurangi
subduction margin, New Zealand, Earth Planet. Sci. Lett. doi
10.1016/j.epsl.2007.03.039.
Nicol, A., J. Walsh, K. Berryman, and P. Villamor (2006). Interdependence
of fault displacement rates and paleoearthquakes in an active rift,
Geology 34, 865–868.
1540
Nicol, A., J. Walsh, V. Mouslopoulou, and P. Villamor (2009). Earthquake
histories and Holocene acceleration of fault displacement rates,
Geology 37, 911–914.
Nicol, A., J. J. Walsh, P. Villamor, H. Seebeck, and K. R. Berryman (2010).
Normal fault interactions, paleoearthquakes and growth in an active
rift, J. Struct. Geol. 32, 1101–1113.
Nodder, S. (1993). Neotectonics of the offshore Cape Egmont fault zone,
Taranaki Basin, New Zealand, New Zeal. J. Geol. Geophys. 36,
167–184.
Nodder, S. D., G. Lamarche, J.-N. Proust, and M. Stirling (2007).
Characterizing earthquake recurrence parameters for offshore faults
in the low-strain, compressional Kapiti–Manawatu fault system,
New Zealand, J. Geophys. Res. 112, B12102, doi 10.1029/
2007JB005019.
Norris, R. (2004). Strain localisation within ductile shear zones beneath
active faults: The Alpine fault contrasted with the adjacent Otago
fault system, New Zealand. Earth Planets Space 56, 1095–1101.
Norris, R. J., and A. F. Cooper (2001). Late Quaternary slip rates and slip
partitioning on the Alpine Fault, New Zealand, J. Struct. Geol. 23,
507–520.
Norris, R. J., and I. M. Turnbull (1993). Cenozoic basins adjacent to an
evolving transform plate boundary, southwest New Zealand, in
P. F. Balance (Editor), South Pacific Sedimentary Basins, Sedimentary Basins of the World 2, 251–270.
Pace, B., M. W. Stirling, N. J. Litchfield, and U. Rieser (2005). New
active fault data and seismic hazard estimates for west Otago,
New Zealand, New Zeal. J. Geol. Geophys. 48, 75–83.
Paquet, F., J.-N. Proust, P. M. Barnes, and J. R. Pettinga (2011). Controls on
active forearc basin stratigraphy and sediment fluxes: The
Pleistocene of Hawke Bay, New Zealand, Geol. Soc. Am. Bull.
123, 1074–1096.
Pedley, K. L., P. M. Barnes, J. R. Pettinga, and K. B. Lewis (2010). Seafloor
structural geomorphic evolution of the accretionary frontal wedge
in response to seamount subduction, Poverty Indentation,
New Zealand, Mar. Geol. 270, 119–138.
Pettinga, J. R., M. D. Yetton, R. J. Van Dissen, and G. Downes (2001).
Earthquake source identification and characterisation for the Canterbury region, South Island, New Zealand, Bull. New Zeal. Natl.
Soc. Earthquake Eng. 34, 282–317.
Pillans, B. (1990). Vertical displacement rates on Quaternary faults,
Wanganui Basin, New Zeal. J. Geol. Geophys. 33, 271–275.
Pondard, N., and P. M. Barnes (2010). Structure and paleoearthquake
records of active submarine faults, Cook Strait, New Zealand: Implications for fault interactions, stress loading, and seismic hazard,
J. Geophys. Res. 115, B12320, doi 10.1029/2010JB007781.
Quigley, M., R. Van Dissen, P. Villamor, N. Litchfield, D. Barrell, K. Furlong, T. Stahl, B. Duffy, E. Bilderback, D. Noble, D. Townsend, J.
Begg, R. Jongens, W. Ries, J. Claridge, A. Klahn, H. Mackenzie, A.
Smith, S. Hornblow, R. Nicol, S. Cox, R. Langridge, and K. Pedley
(2010). Surface rupture of the Greendale fault during the M w 7 1
Darfield (Canterbury) earthquake, New Zealand: Initial findings,
Bull. New Zeal. Natl. Soc. Earthquake Eng. 43, no. 4, 236–242.
Reyners, M. (1998). Plate coupling and the hazard of large subduction thrust
earthquakes at the Hikurangi subduction zone, New Zealand, New
Zeal. J. Geol. Geophys. 41, 343–354.
Reyners, M., and D. Eberhart-Phillips (2009). Small earthquakes provide
insight into plate coupling and fluid distribution in the Hikurangi
subduction zone, New Zealand, Earth Planet. Sci. Lett. 282, doi
10.1016/j.epsl.2009.03.034.
Reyners, M., and T. Webb (2002). Large earthquakes near Doubtful Sound,
New Zealand, 1989–93, New Zeal. J. Geol. Geophys. 45, 109–120.
Reyners, M., P. McGinty, S. Cox, I. Turnbull, T. O’Neill, K. Gledhill, G.
Hancox, J. Beavan, D. Matheson, G. McVerry, J. Cousins, J. Zhao,
H. Cowan, G. Caldwell, and C. BennieThe GeoNet Team (2003).
The Mw 7.2 Fiordland earthquake of August 21, 2003: Background
and preliminary results, Bull. New Zeal. Natl. Soc. Earthquake Eng.
36, 233–248.
M. Stirling et al.
Reyners, M., R. Robinson, and P. McGinty (1997). Plate coupling in the
northern South Island and southernmost North Island, New
Zealand, as illuminated by earthquake focal mechanisms, J. Geophys. Res. 102, 15197–15210.
Rhoades, D. A., and F. F. Evison (2005). Test of the EEPAS forecasting
model on the Japan earthquake catalogue, Pure Appl. Geophys.
162, 1271–1290.
Rhoades, D. A., R. J. Van Dissen, R. M. Langridge, T. A. Little, D. Ninis,
E. G. C. Smith, and R. Robinson (2010). Re-evaluation of the
conditional probability of rupture of the Wellington–Hutt Valley
segmentof the Wellington fault, Bull. New Zeal. Natl. Soc. Earthquake Eng. 44, 77–86.
Robinson, R., and R. Benites (2001). Upgrading a synthetic seismicity
model for more realistic fault ruptures, Geophys. Res. Lett. 28,
1843–1846.
Robinson, R., R. Van Dissen, and N. Litchfield (2011). Using synthetic
seismicity to evaulate seismic hazard in the Wellington region,
New Zealand, Geophys. J. Int. 187, 510–528, doi 10.1111/
j.1365-246X.2011.05161.x.
Robinson, R., T. Webb, P. McGinty, J. Cousins, and D. Eberhart-Phillips
(2003). The 2000 Thompson Sound earthquake, New Zealand,
New Zeal. J. Geol. Geophys. 46, 331–341.
Rodgers, D. W., and T. A. Little (2006). World’s largest coseismic strike-slip
offset: The 1855 rupture of the Wairarapa fault, New Zealand,
and implications for displacement/length scaling of continental
earthquakes, J. Geophys. Res. 111, B12408, doi 10.1029/
2005JB004065.
Schermer, E. R., R. J. Van Dissen, K. R. Berryman, H. M. Kelsey, and
S. M. Cashman (2004). Active faults, paleoseismology and
historical fault ruptures in northern Wairarapa, North Island,
New Zealand, New Zeal. J. Geol. Geophys. 47, 101–122.
Smith, W. D., and K. R. Berryman (1986). Earthquake hazard in New
Zealand: Inferences from seismology and geology, in Recent crustal movements of the Pacific region, W. I. Reilly and B. E. Harford
(Editors), Bull. Roy. Soc. New Zeal. 24, 223–243.
Standards New Zealand (2004). Structural design actions–Part 5: Earthquake actions—New Zealand, New Zealand Standard NZS 1170.5,
Dept. Building and Housing, Wellington, New Zealand.
Stirling, M., M. Gerstenberger, N. Litchfield, G. McVerry, W. Smith,
J. Pettinga, and P. Barnes (2008). Seismic hazard of the Canterbury
region, New Zealand: New earthquake source model and methodology, Bull. New Zeal. Natl. Soc. Earthquake Eng. 41, 51–65.
Stirling, M. W., G. H. McVerry, and K. R. Berryman (2002). A new
seismic hazard model of New Zealand, Bull. Seismol. Soc. Am.
92, 1878–1903.
Stirling, M., J. Pettinga, K. R. Berryman, and M. D. Yetton (2001).
Probabilistic seismic hazard assessment of the Canterbury region,
New Zealand, Bull. New Zeal. Natl. Soc. Earthquake Eng. 34,
318–334.
Stirling, M. W., S. G. Wesnousky, and K. Shimazaki (1996). Fault trace complexity, cumulative slip, and the shape of the magnitude-frequency
distribution for strike-slip faults: A global survey, Geophys. J. Int.
124, 833–868.
Stirling, M. W., S. G. Wesnousky, and K. R. Berryman (1998). Probabilistic
seismic hazard analysis of New Zealand, New Zeal. J. Geol. Geophys. 41, 355–375.
Stock, C., and E. G. C. Smith (2002). Adaptive kernel estimation and continuous probability representation of historical earthquake catalogs,
Bull. Seismol. Soc. Am. 92, no. 3 904–912.
Suggate, R. P. (1987). Active folding in North Westland, New Zealand, New
Zeal. J. Geol. Geophys. 30, 169–174.
Suggate, R. P. (1989). The postglacial shorelines and tectonic development
of the Barrytown coastal lowland, North Westland, New Zealand,
New Zeal. J. Geol. Geophys. 32, 443–450.
Sutherland, R., P. M. Barnes, and C. Uruski (2006). Miocene–Recent deformation, surface elevation, and volcanic intrusion of the overriding
plate during subduction initiation, offshore southern Fiordland,
National Seismic Hazard Model for New Zealand: 2010 Update
Puysegur margin, southwest New Zealand, New Zeal. J. Geol. Geophys. 49, 131–149.
Sutherland, R., K. R. Berryman, and R. J. Norris (2006). Displacement rate
on the Alpine fault determined from offset glacial landforms between Milford Sound and the Cascade Valley: Implications for
the kinematics and seismic hazard of southern New Zealand, Geol.
Soc. Am. Bull. 118, 464–474.
Sutherland, R., D. Eberhart-Phillips, R. A. Harris, T. Stern, J. Beavan, S.
Ellis, S. Henrys, S. Cox, R. J. Norris, K. R. Berryman, J. Townend,
S. Bannister, J. Pettinga, B. Leitner, L. Wallace, T. A. Little, A. F.
Cooper, M. Yetton, and M. Stirling (2007). Do great earthquakes
occur on the Alpine fault in central South Island, New Zealand?
in A Continental Plate Boundary: Tectonics at South Island,
New Zealand D. A. Okaya, T. A. Stern, and F. J. Davey (Editors),
American Geophysical Monograph 175, 235–251, doi 10.1029/
175GM12.
Taylor, S., J. Bull, G. Lamarche, and P. Barnes (2004). Normal fault growth
and linkage in the Whakatane Graben, New Zealand, during the last
1.3 Myr, J. Geophys. Res. 109, B02408, doi 10.1029/2003JB002412.
Townsend, D., A. Nicol, V. Mouslopoulou, J. G. Begg, R. J. Beetham, D.
Clark, M. Giba, D. Heron, B. Lukovic, A. McPherson, H. Seebeck,
and J. J. Walsh (2010). Paleoearthquake histories across a normal
fault system in the southwestern Taranaki Peninsula, New Zealand,
New Zeal. J. Geol. Geophys 53, 375–394, doi 10.1080/
00288306.2010.526547.
Townsend, D., A. Vonk, and P. J. J. Kamp (compilers) (2008). 1:250,0000
Geological Map 7, Geology of the Taranaki Area, Institute of
Geological and Nuclear Sciences, GNS Science, Lower Hutt,
New Zealand, 77 pp + 1 folded map.
Tronicke, J., P. Villamor, and A. G. Green (2006). Detailed shallow geometry and vertical displacement estimates of the Maleme fault zone,
New Zealand, using 2-D and 3-D georadar, Near Surf. Geophys. 4,
155–161.
Van Dissen, R. J., and K. R. Berryman (1996). Surface rupture earthquakes
over the last ∼1000 years in the Wellington region, New Zealand,
and implications for ground shaking hazard, J. Geophys. Res. 101,
no. B3, 5999–6019.
Van Dissen, R., and A. Nicol (2009). Mid-late Holocene paleoseismicity of
the eastern Clarence fault, Marlborough, New Zeal. J. Geol. Geophys. 52, 195–208.
Van Dissen, R., J. Cousins, R. Robinson, and M. Reyners (1994). The
Fiordland earthquake of 10 August, 1993: A reconnaissance report
covering tectonic setting, peak ground acceleration, and landslide
damage, Bull. New Zeal. Natl. Soc. Earthquake Eng. 27, 147–154.
Van Dissen, R. J., A. G. Hull, and S. A. L. Read (1993). Timing of some
large Holocene earthquakes on the Ostler fault, New Zealand. Proc.
of the CRCM ’93, Kobe, Japan, December 6–11, 381–386.
Villamor, P., and K. Berryman (2001). A late Quaternary extension rate in
the Taupo volcanic zone, New Zealand, derived from fault slip data,
New Zeal. J. Geol. Geophys. 44, 243–269.
Villamor, P., and K. Berryman (2006a). Late Quaternary geometry and kinematics of faults at the southern termination of the Taupo volcanic
zone, New Zealand, New Zeal. J. Geol. Geophys. 49, 1–21.
Villamor, P., and K. R. Berryman (2006b). Evolution of the southern termination of the Taupo Rift, New Zealand, New Zeal. J. Geol. Geophys. 49, 23–37.
Villamor, P., K. R. Berryman, T. Webb, M. Stirling, P. McGinty, G. Downes,
J. Harris, and N. Litchfield (2001). Waikato seismic loads—Task
2.1. Revision of seismic source characterisation,Institute of
Geological and Nuclear Sciences, GNS Science, Lower Hutt,
New Zealand, Client Report 2001/59, 109 pp.
Villamor, P., K. R. Berryman, I. A. Nairn, K. Wilson, N. Litchfield, and
W. Ries (2011). Associations between volcanic eruptions from
Okataina Volcanic Center and surface rupture of nearby active
faults, Taupo rift, New Zealand: Insights into the nature of
volcano-tectonic interactions, Geol. Soc. Am. Bull. 123, 1383–
1405.
1541
Villamor, P., R. Van Dissen, B. V. Alloway, A. S. Palmer, and N. Litchfield
(2007). The Rangipo fault, Taupo Rift, New Zealand: An example
of temporal slip-rate and single-event displacement variability in a
volcanic environment, Geol. Soc. Am. Bull. 119, 529–547.
Wallace, L. M., P. Barnes, J. Beavan, R. Van Dissen, N. Litchfield, J. Mountjoy, R. Langridge, G. Lamarche, and N. Pondard (2012). The kinematics of a transition from subduction to strike-slip: An example
from the central New Zealand plate boundary. J. Geophys. Res 117,
doi 10.1029/2011JB008640.
Wallace, L., J. Beavan, R. McCaffrey, and D. J. Darby (2004). Subduction
zone coupling and tectonic block rotations in the North Island,
New Zealand, J. Geophys. Res. 109, B12406, doi 10.1029/
2004JB003241.
Wallace, L. M., R. J. Beavan, R. McCaffrey, K. Berryman, and P. Denys
(2007). Balancing the plate motion budget in the South Island,
New Zealand using GPS, geological and seismological data,
Geophys. J. Int. 168, 332–352. doi 10.1111/j.1365246X.2006.03183.x.
Wallace, L., M. Reyners, U. Cochran, S. Bannister, P. Barnes, K. Berryman,
G. Downes, D. Eberhart-Philips, A. Fagereng, S. Ellis, A. Nicol, R.
McCaffrey, R. J. Beavan, S. Henrys, R. Sutherland, D. H. N. Barker, N. Litchfield, J. Townend, R. Robinson, R. Bell, K. Wilson,
and W. Power (2009). Characterizing the seismogenic zone of a
major plate boundary subduction thrust: Hikurangi Margin, New
Zealand, Geochemistry Geophysics Geosystems 10, Q10006, doi
10.1029/2009GC002610.
Webb, T. H., and H. J. Anderson (1998). Focal mechanisms of large earthquakes in the North Island of New Zealand: Slip partitioning at an
oblique active margin, Geophys. J. Int. 134, 40–86.
Weichert, D. H. (1980). Estimation of the earthquake recurrence parameters
for unequal observation periods for different magnitudes, Bull.
Seismol. Soc. Am. 70, 1337–1346.
Wells, D. L., and K. L. Coppersmith (1994). New empirical relationships
among magnitude, rupture width, rupture area, and surface displacement, Bull. Seismol. Soc. Am. 84, 974–1002.
Werner, M. J., A. Helmstetter, D. Jackson, and Y. Kagan (2012). High resolution long- and short-term earthquake forecasts for California,
Bull. Seismol. Soc. Am. 101, 1630–1648, doi 10.1785/0120090340.
Wesnousky, S. G. (1986). Earthquakes, Quaternary faults, and seismic
hazards in southern California, J. Geophys. Res. 91, no. B12
12587–12631.
Wesnousky, S. G. (1994). The Gutenberg–Richter or characteristic earthquake distribution, which is it? Bull. Seismol. Soc. Am. 84,
1940–1959.
Wesnousky, S. G., C. H. Scholz, and K. Shimazaki (1982). Deformation of
an island arc: Rates of moment release and crustal shortening in
intraplate Japan determined from seismicity and quaternary fault
data, J. Geophys. Res. 87, no. B8 6829–6852.
Wood, R., R. Herzer, R. Sutherland, and A. Melhuish (2000). CretaceousTertiary tectonic history of the Fiordland Margin, New Zealand,
New Zeal. J. Geol. Geophys. 43, 289–302.
Wood, R. A., J. R. Pettinga, S. Bannister, G. Lamarche, and T. J. McMorran
(1994). Structure of the Hanmer strike-slip basin, Hope fault, New
Zealand, Geol. Soc. Am. Bull. 106, 1459–1473.
Wright, I. C. (1993). Southern Havre Trough Bay of Plenty (New Zealand):
Structure and seismic stratigraphy of an active backarc basin
complex, in South Pacific Sedimentary Basins, P. F. Ballance
(Editor), Sedimentary Basins of the World 2, Elsevier, Amsterdam,
196–211.
Wysoczanski, R. J., E. Todd, I. C. Wright, M. I. Leybourne, J. M. Hergt,
C. Adam, and K. Mackay (2010). Backarc rifting, constructional
volcanism and nascent disorganised spreading in the southern Havre
Trough backarc rifts (SW Pacific), J. Volcanol. Geoth. Res. 190,
39–57.
Youngs, R. R., and K. J. Coppersmith (1985). Implications of fault slip rate
and earthquakes recurrence models to probabilistic seismic hazard
estimates, Bull. Seismol. Soc. Am. 58, 939–964.
1542
Zachariasen, J., K. Berryman, R. Langridge, C. Prentice, M. Rymer,
M. Stirling, and P. Villamor (2006). Timing of late Holocene
surface rupture of the Wairau fault, Marlborough, New Zealand,
New Zeal. J. Geol. Geophys. 49, 159–174.
M. Stirling et al.
Department of Geological Sciences
University of Canterbury
P.O. Box 4800
Christchurch, New Zealand
(J.P.)
GNS Science
P.O. Box 30368
Lower Hutt, New Zealand
(M.S., G.M., M.G., N.L., R.V., K.B., M.R., D.R., W.S., A.N., K.C.)
National Institute of Water and Atmospheric Research
Private Bag 14901
Wellington, New Zealand
(P.B., L.W., P.V., R.L., G.L., S.N.)
Department of Civil and Natural Resources Engineering
University of Canterbury
P.O. Box 4800
Christchurch, New Zealand
(B.B.)
Victoria University of Wellington
P.O. Box 600
Wellington, New Zealand
(K.J.)
Manuscript received 2 June 2011