Amy Heath, Dan Clarke, Gary Gibson, Kevin McCue and Russ Van Dissen
The Cadell Fault zone is at the northern end of the Silurian to Devonian aged Heathcote Line that trends 200 km north-south across Victoria to the NSW border. Significant rejuvenation of the Cadell Fault in the late Quaternary caused major drainage modifications and realignment of the Murray River. Large future earthquakes on the Cadell Fault Zone put nearby urban centres including Echuca, Deniliquin, and Bendigo at risk. In addition, the drainage patterns of both the Goulburn and Murray Rivers could again change, significantly affecting productive farming land. The Echuca South Fault scarp is a small secondary fault within a bend in the main fault structure. It is about 10 km west of the main fault zone, and is 12 km long with up to a few metres of vertical offset. Following an e-m survey across the Echuca South scarp during April 2002 we excavated a trench across its southern end to determine its late Quaternary palaeoseismic history. The relief at the trench site was less than two metres, and the fault plane was not unequivocally and clearly exposed in the trench. However, ages of trench wall deposits place important constraints on the timing of faulting. Further investigation will help us select trench sites along the main segment of the Cadell scarp north of Echuca for excavation in early 2005.
Paper P02: [Read]
Peter Stafford, John Berrill and Jarg Pettinga
The Buller region has been one of the most active seismic regions of New Zealand over historical time. In order to conduct Probabilistic Seismic Hazard Assessments for the region, definition and quantification of all seismic sources in the region capable of generating potentially damaging earthquakes must be carried out. Bayesian Inference is used to associate events in the historical record to identified faulting sources in the region. Magnitude-Frequency relations for these seismic sources are derived. Subsequent activity rates are compared to plate motion estimates of Seismic Moment release rate for the region in an attempt to quantify any departure from long-term activity rates. Future occurrence rates of events are estimated using both Poisson and Time-dependent probability models. Resulting Magnitude-Frequency relations for these sources are readily applicable to Probabilistic Seismic Hazard Analyses for the region.
Paper P03: [Read]
Characterisation of late Quaternary displacements on the Waitangi Fault at Aviemore Dam, New Zealand
David Barrell, Russ Van Dissen, Kelvin Berryman and Stuart Read
Aviemore Dam was constructed across the Waitaki River valley in the mid-1960s. The dam straddles the steeply WSW-dipping Waitangi Fault. At the time of construction, no evidence of late Quaternary movement was documented on the fault. However, in the mid 1990s, dam safety review investigations unearthed evidence of late Quaternary deformation, and this led to the detailed paleoseismological investigations that are summarized here. The investigations documented two, and possibly three, surface rupture fault movements on the Waitangi Fault in the last c. 23,000 years, with the most recent movement between 13,100 and 14,100 years ago. These ruptures were located up to 10 m west of the bedrock fault that juxtaposes Mesozoic- and Tertiary-age rocks (respectively east and west of the fault), and had single-event vertical separations in the range of 0.5 m to 2.0 m. Slickenside-lineations, and other slip indicators, show oblique right-lateral – reverse movement along the fault in the late Quaternary with horizontal (H) to vertical (V) ratios of displacement in the range of 1H:3V to 1H:1V, suggesting net single-event surface-rupture displacements in the range of 0.53 m to 2.9 m. In addition, a zone of “small-scale” late Quaternary faults and folds extends up to at least 150 m west of the Waitangi Fault. The fault displacement characterisations documented here were subsequently used to derive earthquake performance assessments for the dam, and to evaluate dam safety under direct fault rupture loadings.
Paper P04: [Read]
Euan Smith and Annemarie Christophersen
We have developed a new type of model for the recurrence time between earthquakes greater than some (large) reference magnitude e.g. M 7. The model is the sum of two parts, representing 'aftershocks' and 'background' earthquakes. The model allows ready computation of quantities of interest in seismic hazard. In particular it allows a calculation of the distribution of time of occurrence to the next (large) earthquake in a region given any elapsed time since the last. The model is characterised by four parameters, of which two are critical: the relative weights of the two parts and a time constant for the 'background'. It appears that 30 - 45% of M ≥ 7 earthquakes can be regarded as aftershocks, the balance being 'new' earthquakes. The model fits quite well the time between earthquakes of M ≥ 7 in New Zealand since 1840. However, comparison with similar global models suggests that some M ≥ 7 aftershocks are missing from the NZ catalogue. Conditional probability curves show that, in New Zealand, the probability of an M 7 earthquake following an M 7 earthquake is several times greater than predicted by current Poisson models for intervals up to about three years following the first earthquake.
Paper P05: [Read]
David Rhoades and Frank Evison
Studies of patterns of earthquake occurrence, revealed by high-quality catalogues, show that most major crustal earthquakes are preceded in the long term by an increase in the rate of occurrence of minor earthquakes. This is called the precursory scale increase (Ψ) phenomenon. The EEPAS (“Every Earthquake a Precursor According to Scale”) forecasting model is a simple point-process model of earthquake occurrence which exploits the Ψ-phenomenon. It adopts predictive scaling relations derived from many examples of the Ψ-phenomenon, and applies them to all earthquakes, thus setting aside the problem of identifying those earthquakes that are actually precursory. It was originally fitted to New Zealand earthquakes with M > 5.75, where it explains the data much better than a quasi-static baseline model. Now it has been shown to be much more informative than the baseline model when tested on catalogues for California and Japan.
Paper P06: [Read]
Graeme McVerry, Jim Cousins, Des Bull and David Brunsdon
The first issue in Urban Search and Rescue (USAR) is the safety of the rescuers. In the post-earthquake situation, a key consideration is the likelihood of aftershocks, how large these are likely to be and how strong the shaking will be at the site. Important features of aftershocks are that their expected frequency decreases with time, and that their maximum magnitudes are generally related to the main shock magnitude. Also, as time evolves, more information becomes available about the particular aftershock sequence, and its decay. Such specific information may be used in preference to the generic properties of aftershock sequences to guide the timing of various rescue activities. A NO-GO assessment immediately after the main shock may be reassessed as a viable operation with the passage of time, when all factors are taken into account.
Paper P07: [Read]