The 1942 June 24 Ms7.2, August 1 M87.0, and December 2 Ms6.0, Wairarapa, New Zealand, earthquakes: analysis of observational and instrumental data

Summary

In 1942, two large earthquakes, on June 24 (M_S7.2) and August 1 (M_S7.0) strongly shook the Wairarapa, Manawatu and Wellington districts, causing moderate to severe damage. A third earthquake, of magnitude M_S6.0, occurred in the same area on December 2. The June and August main shocks were each preceded within a few hours by a moderate magnitude "foreshock" (MS5.2 June; M_S5.3 August), previously assumed to originate from the same location. The seismological aspects of the earthquakes were not studied in detail at the time, although their locations were identified instrumentally, and by the distribution of intensity (Rossi-Forel scale), and by the discovery of a scarp that was assumed to be the surface rupture of a fault about 15 km east of Masterton. Although two Wellington-based engineers produced reports on aspects of damage in Wellington, no engineering report on the damage in the Wairarapa, closest to the epicentres, was published. The locations and effects of these earthquakes have now been studied in detail, by rereading and analysing seismograms from 1942 using modem techniques and by the collection and analysis of contemporary descriptive accounts and technical information from many archival sources. New isoseismal maps and new damage (casualties, building, lifelines, landslides and ground damage) summaries for the three main earthquakes are now available.

Except for the August "foreshock", all instrumentally locatable earthquakes in the sequence, including major aftershocks, cluster about Masterton. They are within about 35 km of each other and within 15-30 km west to southwest of the epicentres calculated in 1942. The August "foreshock" was shallow (fixed at 12 km) and located about 50km northeast of the June and August mainshocks. The June mainshock and foreshock and the December mainshock were also shallow (fixed at 5, 12 and 12 km resp.), while the August mainshock was deeper (fixed at 40 km). The distribution of the June earthquake aftershocks is compatible with an approximately NNE-SSW trending right-lateral strike-slip rupture suggested by the preliminary source mechanism solution of Doser & Webb. The aftershock S-P intervals at Wellington suggest a rupture length of 35-45 km. The maximum intensity in the M_S7.2 June 24 earthquake was MM9, which occurred at several locations within the MM8 isoseismal area. In the zone of strongest shaking (MM8), many brick buildings were badly damaged. However, all reinforced concrete (90+) and retrofitted brick (4) buildings performed well, and, except for chimney damage and a few buildings coming off their piles, most timber buildings also performed well. Underground services (water supply, drainage and gas supply) throughout the MM6 and higher intensity areas were almost undamaged. Roads, bridges and railways suffered moderate damage arising from small landslides and movement of fills or soft ground. Telephone and electricity services were disrupted for up to a few days in a few locations, partly due to overhead wires falling and partly due to damage to operational buildings.

Widespread landsliding and liquefaction-induced ground damage occurred throughout an area of about 11,500 km2. Most landslides were of small to moderate size (10³-10⁵m³), although a few large to very large landslides (10⁵-10⁶ m³ or greater) were also reported. The heaviest ground damage occurred within the MM8 isoseismal, the most intense landslide damage showing a close correlation with the zone of aftershocks and the epicentre. Within the MM7 isoseismal landslides were sparse on natural slopes, but many cut slopes and fills failed, while within the MM6 isoseismal there were mainly scattered minor rockfalls or fallen rocks. Two significant failures occurred at 90 and 110 km from the epicentre, however. Significant liquefaction effects were reported from the MM8 area and in the far field, at Shannon and Foxton, where liquefaction of highly susceptible swamp and alluvial deposits was probably a microzone effect. The earthquake apparently caused no surface rupture. The scarps observed by Ongley and others in 1942 are now identified as the heads of two incipient landslides, one continuous with a pre-existing large landslide.
Only one death and three moderate injuries were recorded, despite much falling masonry. The low casualty count can be attributed to the fact that the earthquake occurred at 11.16 pm (NZ time) when most people were inside timber houses. This was true also of the August 1 earthquake (at 00.34 am NZ time), in which only a few minor injuries were reported.

While the August 1 earthquake (M_S7.0,) was almost as large as the June 24 event and its epicentre was within a few kilometres, it was far less damaging overall because of its greater depth (40 km). The strongest isoseismal is MM7 compared to MM8 in the June event. The main damage in the W airarapa consisted of considerable further damage to unreinforced brick chimneys and to some of the brick buildings in the townships, particularly Masterton, Carterton and Eketahuna. Some timber houses in rural areas were also reported as being damaged, e.g. "badly wrenched" (the meaning of which is unclear). A notable feature of the August earthquake was the weakening effect of previous damage from the June earthquake, and the existence of many freshly rebuilt, and therefore weak, chimneys. This without doubt exaggerated the earthquake's effect, especially in Wellington where previously damaged parapets and bricks walls collapsed. Generally, lifelines (electricity, gas, telephone, telegraph services, roads, railways, water and drainage) were affected at fewer locations over a smaller area and for a shorter time than in the June earthquake. Nevertheless, in places there were significant delays and damage. The area in which earthquake-induced landsliding, ground cracking, subsidence and liquefaction effects occurred was extensive (approximately 5,600 km) although significantly smaller than that affected by the June earthquake.

Although Wellington lies within the MM6 isoseismal of both earthquakes, MM7 was experienced in parts of the city, particularly in the Central Business District (CBD). This has been previously interpreted as a microzone effect. A map (of unknown authorship) has been newly discovered that identifies the damaged non-Government buildings in the CBD, with the buildings classified by age of construction but not by the construction type or extent of damage. The map clearly shows the greater percentage damage to the pre-1914 buildings, and that much of the damage to these buildings occurs in areas of fill seaward of the pre-1855 shoreline. The map shows broad agreement between the 1992 shaking hazard zones and the percentage of damaged structures, with some discrepancies. Further, there is a strong correlation of damage with particular units of fill. These features might be resolved by an intensive study of newspaper accounts and the 420 building damage reports from 1942 that are held at the Wellington City Archives. The total damage costs for buildings in Wellington City as a result of both earthquakes was significant: non-domestic buildings £805,000; dwellings and chimneys £80,000 (respectively $48.3 million and $4.8 million in 1998 NZ values). The results of this study on the 1942 earthquakes have several significant implications for seismic hazard assessment of the east coast:

1. It is apparent that in this part of New Zealand, large strike-slip earthquakes can occur that do not result in a surface ground rupture, and thus they have a low, or at least reduced, likelihood of being identified in the geological record.
2. The June 1942 Wairarapa earthquake, along with the 1934 Pahiatua earthquake, indicate that a significant, but as yet un-quantified, amount of strike-slip displacement occurs parallel to the subduction margin well east of the main belt of strike-slip faults, defined to a large extent by the Wellington and Wairarapa Faults.
3. The closeness in time of the June foreshock and later mainshocks suggests that stress triggering may have occurred. Further, the closeness in time of the 1934 Pahiatua earthquake and the 1942 earthquakes, which are within 10-15 km and almost along strike of the assumed rupture zone of the 1934 earthquake, suggests that interaction between these events should also be considered. Both studies would require the source mechanisms of Doser & Webb, which are yet to be finalised.
4. The June and August 1942 earthquakes, one in the upper plate and the other almost immediately below it in the subducted plate, seem to be paired in a similar manner to the February and May 1990 Weber earthquakes, although in this latter case the first earthquake was the deeper event. This indicates that the 1990 Weber events were not unique and that seismic hazard assessments should take in account the possibility of similar couplings of two (or more) large earthquakes, and the possibility of cumulative structural damage.