Late Holocene paleoseismicity of the Pahiatua section of the Wellington fault, New Zealand

Six trenches and additional exposures have been investigated along a 23 km length of the Pahiatua section of the Wellington Fault. The sites show a consistent fault rupture record for the last four surface-faulting events along the Pahiatua section. This multi-site record of events, supported by 28 radiocarbon ages that span the last c. 4500 yr, is the most complete paleoseismic record for the Wellington Fault. From southwest to northeast, the trenches are called Death-1 and -2, Hughes-2 and -1, and Ebbett-2 and -1, named after local farmers. Additional data come from an exposure at the Army Depot northeast of Ebbett-1 and a stream cutting near the Death-1 trench. Earthquake events are recognised on the basis of upward terminations of faults and stratigraphic evidence (unconformities; scarp-derived colluvial deposits; “co-seismically” generated, organic, poorly sorted units; and subsequent scarp-ponded units).On this basis, evidence for the most recent surface-faulting event is recognised at all sites (i.e., between Death-1 and the Army Depot site) and has a timing of AD 1670-1800 (150-280 cal. yr BP). The penultimate faulting event has been recognised at all six trenches and occurred between AD 690 and 1170 (780-1260 cal. yr BP). Clear evidence, and datable material associated with events III and IV, is observed only in the older records of the Ebbett-1 and Hughes trenches. Event III is bracketed by faulting and two ages on co-seismic deposits to the interval AD 65 to 400 BC (1885-2350 cal. yr BP). Event IV is bracketed by three key ages in the Hughes-1 trench to the interval 1600-2140 BC (3550-4090 cal. yr BP). The trench-derived recurrence interval based on these four events is c. 1200 ±110 yr, which is consistent with, though longer than, an estimate of the recurrence interval based on slip rate (5.1-6.2 mm/yr) and single-event displacement (4.5 ± 1 m), which gave a range of 820 ± 260 yr.When the current paleoseismic data for the Wellington Fault are considered together, they suggest that: (1) the last four large earthquakes along the Pahiatua section have involved rupture of the entire section; (2) these events broadly correlate with other dated paleoearthquake events along the Wellington-Hutt Valley segment and Tararua section of the fault, and the southern end of the Mohaka Fault; (3) the most recent faulting event along the Pahiatua section is somewhat younger than for the Wellington-Hutt Valley segment or Tararua section; and (4) either single or dual segment/section rupture scenarios existed along these four fault sections.