Upper plate deformation and its relationship to the underlying Hikurangi subduction interface, southern North Island, New Zealand

Abstract

What do ancient shore platforms on the south coast tell us about megathrust earthquakes in the lower North Island of New Zealand?

With the Pacific and Australian plates “locked” under New Zealand’s lower North Island as part of the Hikurangi subduction zone, strain may be building up that could be released in a devastating megathrust earthquake and tsunami.

This research looked at land movements across the lower North Island over the past 200,000 years to see what had most likely been created by historical megathrust earthquakes on the plate boundary, and what had likely been created by shallower crustal earthquakes on upper plate faults such as the Wellington, Ohariu and Wairarapa faults.  The results will add to what we know about the activity of upper plate faults, and the likelihood and likely severity of an earthquake on the plate boundary under the lower North Island.

Findings

The height of ancient beaches that have been uplifted to form terraces on the south coast, between Tongue Point, west of Wellington, and Cape Palliser to the east, vary depending on their distance from the plate boundary. This elevation pattern indicates that they were created at least partly by rupture on the underlying subduction interface.  Evidence also suggests that terraces near crustal faults above the plate boundary have been moved by rupture on these shallower faults. 

Research on terraces along the Hutt River showed that the Wellington fault was more active between about 8,000 and 10,000 years ago, and then relatively quiet until it became active again about 4,500 years ago.

Use

The results of this research, when collated with other research, will create a combined picture to help understand how large and how often shaking by the local faults and plate boundary movements is likely to be.  Combined results will ultimately feed into the National Seismic Hazard Model and building standards via updates to the Building Code.

---

Technical Abstract

At the southern Hikurangi Margin, the subduction interface between the Australian and Pacific plates, beneath the southern North Island of New Zealand, is ‘locked’. It has previously been estimated that sudden slip on this locked portion of the interface could result in a subduction megathrust earthquake of Mw 8.0-8.5 or larger. Historically, however, no significant (>Mw 7.2) subduction earthquake has occurred at the southern Hikurangi Margin, and the hazard from subduction earthquakes to this region, which includes New Zealand’s capital city of Wellington, remains largely unknown.

Patterns of tectonic deformation at subduction margins can provide insight into subduction processes, including megathrust earthquakes. With the objectives to i) contribute to the understanding of partitioning of margin-parallel plate motion on to upper plate faults, and ii) provide insight into the relationship of permanent vertical deformation to subduction processes at the southern end of the Hikurangi margin, I investigate the flights of late Pleistocene fluvial and marine terraces preserved across the lower North Island. Such geomorphic features, when constrained by absolute dating, provide a valuable set of data with which to quantify tectonic deformation - be they locally offset by a fault, or collectively uplifted across the margin.

Fault-offset fluvial terraces along the Hutt River, near Wellington, record dextral slip for the southern part of the Wellington Fault – the Wellington Hutt Valley (WHV) segment. From re-evaluated fault displacement measurements and new Optically Stimulated Luminescence (OSL) age data, I estimate an average slip rate of 6.3 ±   mm/yr during the last ~100 ka. However, slip on this segment of the Wellington Fault has not been steady throughout this time. During the Holocene, there was a phase of heightened ground rupture activity between ~8 and 10 ka, a period of relative quiescence between ~4.5 and 8 ka, and another period of heightened activity during the last ≤ 4.5 ka. These results agree with independent paleoseismological evidence from other sites along the Wellington Fault for the timing of ground rupture events. The time-varying activity observed on the Wellington Fault may be regulated by stress interactions with other nearby upper plate active faults.

Net tectonic uplift across the southern Hikurangi Margin is recorded by ancient emergent shore platforms preserved along the south coast of the North Island. I provide a new evaluation of the age and distribution of the Pleistocene marine terraces. OSL analyses provide absolute ages for these terraces, many for the first time. The data suggest that the most extensive Pleistocene terraces formed during sea level highstands of Marine Isotope Stages (MIS) 5a (~82 ka), 5c (~96 ka), 5e (~123 ka) and 7a (~196 ka). Shore platform elevations are accurately measured for the first time using Global Navigational Satellite Systems (GNSS) surveying. From these data I have determine the shore platform attitudes where they are preserved along the coast. Because the ancient shorelines are now obscured by coverbed deposits, I use the calculated shore platform attitudes to reconstruct strandline elevations. These strandline elevation, corrected for sea level during their formative highstands, have been used to quantify rates of uplift across the southern Hikurangi Margin.

In the forearc region of the Hikurangi Margin, uplift observed on the marine terraces monotonically decreases away from the Hikurangi Trough. The highest uplift rate of ~1.7 mm/yr is observed at the easternmost preserved terrace, near Cape Palliser, ~40 km from the trough. The lowest rate of uplift, <0.2 mm/yr, is observed at Wharekauhau, ~ 70 km from the trough. Overall, the terraces are tilted towards the west, away from the trough, with older terraces exhibiting the most tilting. This ~30 km long pattern of uplift suggests that, in this forearc region of the margin, deep-seated processes are the main contributors to permanent vertical deformation, the most likely contenders being subduction of the buoyant Hikurangi Plateau and permanent uplift resulting from repeated megathrust earthquakes.

At a distance of >70 km from the Hikurangi Trough, vertical offsets on the marine terraces are evident across upper plate faults, most notably the Wairarapa and Ohariu faults. The uplift rate at Baring Head, west and on the upthrown side of the Wairarapa Fault, is as much as ~1.6 mm/yr. At Tongue Point, where the Ohariu Fault offsets the marine terraces preserved there, uplift calculated from the western, upthrown side of the fault is ~0.6 mm/yr, whereas uplift calculated from the downthrown side is <0.2 mm/yr. These vertical offsets suggest that at this distance, in the Axial Ranges of the Hikurangi Margin, the major active upper plate faults contribute to enhanced uplift rates on the upthrown sides of the faults. However, because this entire region is elevated overall, uplift here also likely has a deeper source, and is possibly related to the sediment underplating previously identified beneath this region.