Institute of Geological and Nuclear Sciences, Lower Hutt, New Zealand

Time-invariant, long-range, and short-range forecasting models were fitted to the earthquake catalogue of Greece for magnitudes 4.0 and greater to optimize their ability to forecast events of magnitude 6.0 and greater in the period 1966–1980. The models considered were stationary spatially uniform and spatially varying Poisson models, a long-range forecasting model based on the precursory scale increase phenomenon with every earthquake regarded as a precursor according to scale, and epidemic type short-range forecasting models with spatially uniform and spatially varying spontaneous seismicity. Each of the models was then applied to the catalogue for 1981–2002, and their forecasting performance was compared using the log likelihood statistic. The long-range forecasting model performed substantially better than the time-invariant models, and the short-range forecasting models performed substantially better again. The results show that the information value to be gained from modeling temporal and spatial variation of earthquake occurrence rate, at both long and short range, is much greater than can be gained from modeling spatial variation alone.

New Zealand’s tectonic setting, astride an obliquely convergent tectonic boundary, means that it has experienced many large earthquakes in its 200-year written historical records. The task of identifying and studying the largest early instrumental and pre-instrumental earthquakes, as well as identifying the smaller events, is being actively pursued in order to reduce gaps in knowledge and to ensure as complete and comprehensive a catalogue as is possible. The task of quantifying historical earthquake locations and magnitudes is made difficult by several factors. These include the range of possible earthquake focal depths, and the sparse, temporally- and spatially-variable historical population distribution which affects the availability of felt intensity information, and hence, the completeness levels of the catalogue. This paper overviews the procedures and tools used in the analysis, parameterisation, and recording of historical New Zealand earthquakes, with examples from recently studied historical events. In particular, the 1855 M 8+ Wairarapa earthquake is discussed, as well as its importance for the eminent 19th century British geologist, Sir Charles Lyell, and for future global understanding of the connection between large earthquakes and sudden uplift, tilting and faulting on a regional scale.

The Lake Edgar Fault in Western Tasmania, Australia is marked by a prominent fault scarp and is a recently reactivated fault initially of Cambrian age. The scarp has a northerly trend and passes through the western abutment of the Edgar Dam, a saddle dam on Lake Pedder. The active fault segment displaces geologically young river and glacial deposits. It is 29 ± 4 km long, and dips to the west. Movement on the fault has ruptured the ground surface at least twice within the Quaternary and possibly the last ca. 25 000 years; the most recent rupture has occurred since the last glaciation (within the last ca. 10000 years). This is the only known case of surface faulting in Australia with evidence for repeated ruptures in the Late Pleistocene. Along its central portion the two most recent surface-faulting earthquakes have resulted in about 2.5 m of vertical displacement each (western side up). The Lake Edgar Fault is considered capable of generating earthquakes in the order of magnitude 61/2-71/4. The Gell River Fault is another fault nearby that was apparently also active in the Late Pleistocene. It has yet to be studied in detail but the scarp appears to be more degraded and therefore older than the most recent movement on the Lake Edgar Fault.

The Conway Segment of the dextral-slip Hope Fault is one of the fastest slipping fault segments along New Zealand s plate boundary, but has not ruptured co-seismically in the historic period and little paleoseismic data exist to constrain its large earthquake record. Two paleoseismic trenches were opened adjacent to Greenburn Stream near Kaikoura for the 2001 ILP Paleoseismology Conference. Both trenches were excavated into deposits ponded against an uphill-facing shutter scarp. Trench 1, dug through a cobbly soil and surface deposit was dominated by a thick fan/fluvial sequence that was radiocarbon dated at 4409 ± 60 C14 years BP (4844-5288 cal years BP) at the base of the trench. This trench exhibited evidence of complex deformation from many paleoseismic events. The most recent earthquakes are difficult to constrain due to a lack of cover stratigraphy on the fan deposits. However, the modern soil appears to be faulted and is covered by cobbles with a weathering rind-derived age of 220 ± 60 years. Trench 2, dug ?? 50 m to the west has an expanded sequence of the younger cover deposits. Paleoseismic event horizons have been recognised from the combined evidence of upwardterminating faults, offset and mismatched units, a sandblow deposit, and abrupt landscape change shown by the burial of paleosol surfaces that form the event horizons. Two paleosols underlying the modern soil are clearly faulted by two separate rupture events. A dome of sand interpreted as a liquefaction sandblow deposit overlies the lower paleosol (event horizon). Both paleosols are overlain by metre-thick debris deposits, interpreted as earthquake-induced rock avalanches that cascaded off the hillslope following Mw 7 + events. Four radiocarbon dates place some constraints on the timing of the three recent surface-rupturing events. The youngest and lowest date is 548 ± 60 C14 years BP (504-656 cal years BP) and occurs below the lower paleosol. It constrains the maximum duration of time in which the last 2 earthquake events occurred to be 545 years (1295-1840 A.D.). This is consistent with the average Recurrence Interval (RI) of 180-310 years that we determine using two independent paths. The soil record indicates that each event is separated by a significant period of time, comparable to the calculated RI. The most recent event is constrained between ca. 1780 A.D. ± 60 years, taking into account the dates from these trenches, a weathering rind age, and from stratigraphic correlation at the site. Event III probably occurred before 1220 A.D. A maximum dextral slip rate of 23 ± 4 mm/yr is calculated from the minimum fan age and the offset/deflection of a stream channel along the shutter ridge. In concert with the estimate of single event displacement (5-6 m), these results show that the Conway Segment of the Hope Fault is fast-slipping and has ruptured regularly as a result of large earthquakes prior to the European colonisation of New Zealand.

A sudden increase in the scale of seismicity has occurred as a long-term precursor to twelve major earthquakes in California and Northern Mexico. These include all earthquakes along the San Andreas system during 1960-2000 with magnitude M 6.4. The full list is as follows: Colorado Delta, 1966, M 6.3; Borrego Mt., 1968, M 6.5; San Fernando, 1971, M 6.6; Brawley, 1979, M 6.4; Mexicali, 1980, M 6.1; Coalinga, 1983, M 6.7; Superstition Hills, 1987, M 6.6; Loma Prieta, 1989, M 7.0; Joshua Tree, 1992, M 6.1; Landers, 1992, M 7.3; Northridge, 1994, M 6.6; Hector Mine, 1999, M 7.1. Such a Precursory Scale Increase () was inferred from the modelling of long-term seismogenesis as a three-stage faulting process against a background of self-organised criticality. The location, onset-time and level of are predictive of the location, time and magnitude of the future earthquake. Precursory swarms, which occur widely in subduction regions, are a special form of ; the more general form is here shownto occur frequently in a region of continental transform. Other seismicity precursors, including quiescence and foreshocks, contribute to or modulate the increased seismicity that characterises . The area occupied by is small compared with those occupied by the seismicity precursors known as AMR, M8 and LURR. Further work is needed to formulate as a testable hypothesis, and to carry out the appropriate forecasting tests.

The Hellenic subduction region displays the same precursory swarm phenomenon as has been found in comparable regions of New Zealand and Japan. In the earthquake catalogue of the Aristotle University of Thessaloniki, 10 past sequences of precursory swarms and related major mainshock events have been identified. These correlate, in respect of location, magnitude and time, with the 9 sequences previously identified in New Zealand, and 9 in Japan, bringing the total of sequences to 28, and the totals of related events (allowing for clustering) to 56 precursory swarms and 42 mainshock events. The results add strength to the hypothesis that swarms are long-range predictors of mainshock events. A close similarity between the swarm and aftershock magnitudes in a given sequence is also confirmed in Greece, supporting the proposal that swarms are an integral part of the seismogenic process in subduction regions. Further, the modelling of swarms as part of an overall increase in seismicity, the onset of which marks the onset of seismogenesis, is well illustrated from past sequences in Greece. Formal tests are being carried out in Greece, in parallel with New Zealand and Japan, to ascertain the performance of the hypothesis as a basis for long-range synoptic forecasting.