The Te Puninga Fault, Hauraki Plains: a new seismic source in the low seismicity northern region of New Zealand

Figures & data

Figure 1. Tectonic setting of the Te Puninga Fault. A, The Hauraki Rift (dashed rectangular zone), regional active faults (red lines, from the New Zealand Active Faults Database; KF, Kerepehi Fault; TPF, Te Puninga Fault), and recent seismicity (M_W >4 earthquakes from 1952 to 2022, from Geonet, https://quakesearch.geonet.org.nz/). B, Detailed map of the Te Puninga Fault (blue, orange and black lines), showing the locations of the two trenches investigated in this project, Ryland and Arnold, and the two swathes where slip rate was calculated (yellow bands).

Table 1. Radiocarbon ages obtained from the Arnold and Ryland paleoseismic trenches on the Te Puninga Fault, and from the Hinuera Formation (Hinuera C) near Matamata.

Waikato Radiocarbon Laboratory number	Material dated and pre-treatment^a	Conventional radiocarbon age (^14C yr BP ± 1 sd)	Calibrated ^14C age^b (cal yr BP, 95.4% probability range)		δ^13C^c (‰)	Location^d
Wk52561	Twigs and plant material	10,071 ± 20	11,690 ± 30 11,485 ± 165	11.5% 84.0%	*	Ryland Trench^e
Wk52552	Peat	19,773 ± 46	23,795 ± 75 23,485 ± 55	88.3% 7.1%	*	Arnold Trench^e
Wk52554	Peat	19,963 ± 47	23,970 ± 170	95.4%	*	Arnold Trench^e
Wk52555	Leaf in peat	20,705 ± 50	24,900 ± 220	95.4%	*	Arnold Trench^e
Wk216	Peat	19,400 ± 200	23,781 ± 200	95.4%	−28.2	Quarry near Matamata c. 2.5 m depth^f
Wk217	Peat	18,400 ± 200	22,863 ± 200	95.4%	−26.8	Cutting near Matamata c. 6.2–7.0 m depth^f
Wk218	Peat	23,900 ± 400	28,904 ± 200	95.4%	−25.9	Bore 37 near Matamata c. 28.1 m depth^f

^a Pretreatment: Visible contaminants removed, sample washed in hot HCl, rinsed and treated with multiple hot NaOH washes. The NaOH insoluble fraction was treated with hot HCl, filtered, rinsed, and dried. Dated via AMS. Surfaces scraped clean. Wood was chopped into small splinters and milled. Washed in demineralised water and dried. Treated with sodium chlorite to leave holocellulose, rinsed with distilled water and dried. Dated via liquid scintillation spectrometry.

^b Calibrated using OxCal v.4.4 (updated 2023; based on Bronk Ramsey 2001) and SHCal20 (atmospheric curve) (Hogg et al. 2020) (see Supplementary material S2).

^c The δ¹³C value was measured on prepared graphite using an AMS spectrometer. The radiocarbon date has therefore been corrected for isotopic fractionation. However, the AMS-measured δ¹³C value can differ from the δ¹³C of the original material and it is therefore not shown.

^d See Figures 4 and 5 showing the sampling positions of the Arnold and Ryland trench walls, respectively.

^e Data from this study.

^f Data from Hogg et al. (1987) and Houghton and Cuthbertson (1989).

Figure 2. Views of the different geomorphic characteristics of the Te Puninga Fault. A, Area of the Te Puninga Fault around the Waiharakeke West Stream that illustrates the interplay between fault scarps and fluvial terrace risers in locations where the fluvial drainage is parallel with the fault strike. B, Fault mapping interpretation for panel A. C, Area of the Te Puninga Fault in generally very flat topography where the fault scarps are very subtle in the LiDAR hillshade model. D, Fault mapping interpretation of panel C, showing how overlaying a dynamic-range-adjusted digital elevation model over the hillshade model enhances the fault scarps in this area of low topography. The pale-grey lines depict roads.

Figure 3. Example of topographic profiling undertaken across the Te Puninga Fault. Arnold (A) and Ryland (B) profiling swaths are indicated by the yellow boxes. White diamonds are location of trenches. Red lines are mapped traces of the Te Puninga Fault and blue dots are topographic profile locations. The white dashed line in (A) delineates the boundary between the two broad groups of surfaces discussed in the text: to the west are older alluvial fan deposits, while to the east, are the low lying surfaces of the Hauraki Plains and the youngest alluvial fan deposit. (C) and (D) are example topographic profiles from the Arnold and Ryland swathes, respectively, where red points are the elevations manually selected to define the slope of the upthrown and downthrown side of the fault resulting in the orange and blue best-fitting lines, respectively.

Figure 4. Trench log of the south wall of the Arnold trench. Offset of trench Unit A is indicated by the dashed lines on the left side of the trench log. Units A-C correspond to Hinuera C (Manville and Wilson 2004) of Hinuera Formation and unit 1 corresponds to Waitoa Formation (of Houghton and Cuthbertson 1989; Brathwaite and Christie 1996).

Figure 5. Trench log of the north wall of the Ryland trench. Units A-C correspond to Hinuera C (Manville and Wilson 2004) of Hinuera Formation and unit 1 corresponds to Waitoa Formation (of Houghton and Cuthbertson 1989; Brathwaite and Christie 1996).

Figure 6. Schematic diagram of the general stratigraphy and chronology of the Arnold (A) and Ryland (B) trenches as east-west cross sections spanning the Te Puninga Fault, showing two or three main stratigraphic units, respectively, and associated pedological soils at the land surface. The surface soil classifications are based on Hewitt (2010) and Hewitt et al. (2021a, 2021b); the soil series names are from Wilson (1980). Age on Hinuera C deposits based on this study (Table 1); age on base of tephra mantle (in Arnold trench) is based on likely presence of Okareka tephra (c. 23,500 cal yr BP), identified by Hughes (2023); age on Waitoa Formation is from this study (Table 1). In Ryland trench, the asterisks in the soil profile represent redox features. Both sketches mimic the trench walls but are not to scale.

Figure 7. The age model used to constrain the age of the Unit A deposits (95.4% probability range). Median age is 23,385 cal yrs BP. Age model derived from OxCal (v. 4.4; Bronk Ramsey 2001, 2009) using the SHCal20 calibration curve (Hogg et al. 2020).

Figure 8. Profiles of cumulative net offset across the Arnold (A) and Ryland (B) profiling swaths. Down to the southwest is positive and down to the northeast is negative. The bold black line is the mean offset and orange bars reflect the 95% confidence interval. Grey lines are different simulations in the Monte Carlo used to propagate uncertainty.

Figure 9. Interpolated along-fault net slip rate profile for the Te Puninga Fault. The red line is the average slip rate whereas the grey dashed lines reflect the two-sigma uncertainties on the net slip rate. Open circles are the net slip rates measured along the Ryland (south part of the fault) and Arnold (middle part of the fault) profiling swaths. The fault is assumed to have 0 mm/yr slip rate at its southern and northern ends.

Table 2. Potential Peak Ground Acceleration (PGA) and Modified Mercalli Intensity (MMI) that could be expected in selected townships following rupture of the Te Puninga Fault.

Township	MMI* for Mw 6.9(average; this study) ^+	MMI* for Mw 7.25 (geology peak value from Gerstenberger et al. 2022)	PGA** for Mw 6.9 (average; this study)^++	PGA** for Mw 7.25 (geology peak value from Gerstenberger et al. 2022)	MMI *** for Mw 6.7 (from Persaud et al. 2016)
Morrinsville	9.7 ± 1.02	10.0 ± 1.02	0.35 ± 1.35	0.38 ± 1.7
Tauranga	6.6 ± 1.04	7.2 ± 1.03	0.04 ± 1.45	0.05 ± 1.6	6
Te Aroha	8.9 ± 1.04	9.5 ± 1.03	0.1 ± 1.38	0.17 ± 1.62	8
Hamilton	7.9 ± 1.03	8.4 ± 1.03	0.09 ± 1.40	0.11 ± 1.58	7

* Median MMI and 2σ.

** Median PGA and 2σ. PGA of ‘1’ is 9.81 m/s².

*** Median MMI.

⁺ Scaling relations and GMPEs from Gerstenberger et al. (2022).

⁺⁺ Scaling relations from Gerstenberger et al. (2022); IPE from Dowrick and Rhoades (2005).

Figure 10. A, Fault sources around the Te Puninga Fault from the NZ Community Fault Model (NZ CFM, Seebeck et al. 2023). In blue are active faults that could rupture with the Te Puninga (TP) Fault based on the grand inversion in the 2022 NZ NSHM (Gerstenberger et al. 2022), namely: segments of the Kerepehi Fault (KO, Okoroire; KTP, Te Poi; KW, Waitoa; KE, Elstow; KA, Awaiti; KOf, Offshore) and the Mangatangi (Ma), Pokeno (Po) and Aka Aka (AA) faults. In red are other fault sources in the NZ CFM. B, Single-event ruptures of multiple faults that include the Te Puninga Fault (dashed box) are depicted by the filled polygons. The larger magnitude events are superimposed by the smaller magnitude ones. Open polygons denote fault segments (of the Community Fault Model; Seebeck et al. 2023) that do not participate in any of the ruptures involving the Te Puninga Fault. C, Participation magnitude frequency distribution of the events associated with the Te Puninga Fault. The event rates include those given by inversion of slip rates and distributed (or background) seismicity. The former accounts for magnitudes larger than Mw 6.8. The event rates derived from the inversion of the geologic slip rates (red lines) show a characteristic event with the mode at Mw ∼7.25, while those derived from the inversion of geodetic slip rates (blue lines) show the mode at Mw ∼7.45. The dataset shown in (B) and (C) was extracted from the 2022 NZ NSHM (Gerstenberger et al. 2022).

Bronk Ramsey C. 2001. Development of the radiocarbon calibration program. Radiocarbon. 43:355–363. doi:10.1017/S0033822200038212.

 Web of Science ®Google Scholar

Hogg AG, Heaton TJ, Hua Q, Palmer JG, Turney CSM, Southon J, Bayliss A, Blackwell PG, Boswijk G, Bronk Ramsey C, et al. 2020. SHCal20 Southern Hemisphere calibration, 0–55,000 years cal BP. Radiocarbon. 62(4):759–778. doi:10.1017/RDC.2020.59.

 Web of Science ®Google Scholar

Hogg AG, Lowe DJ, Hendy CH. 1987. University of Waikato radiocarbon dates I. Radiocarbon. 29:263–301. doi:10.1017/S0033822200056976.

 Web of Science ®Google Scholar

Houghton BF, Cuthbertson AS. 1989. Sheet T14 BD: Kaimai. Geological map of New Zealand 1:50,000. Wellington (NZ): New Zealand Geological Survey, Department of Scientific and Industrial Research. 1 sheet + 35p.

 Google Scholar

Manville V, Wilson CJN. 2004. The 26.5 ka Oruanui eruption, New Zealand: a review of the roles of volcanism and climate in the post-eruptive sedimentary response. New Zeal J Geol Geop. 47:525–547. doi:10.1080/00288306.2004.9515074.

 Web of Science ®Google Scholar

Brathwaite RL, Christie AB. 1996. Geology of the Waihi area 1:50,000: part sheets T13 and U13. Lower Hutt (NZ): Institute of Geological & Nuclear Sciences. 1 sheet + 64p.

 Google Scholar

Hewitt AE. 2010. New Zealand Soil Classification. Lincoln (NZ): Manaaki Whenua Press. (Landcare Research science series; 1). doi:10.7931/DL1-LRSS-1-2010.

 Google Scholar

Hewitt AE, Balks MR, Lowe DJ. 2021a. Allophanic soils. In: The soils of Aotearoa New Zealand. Cham: Springer; p. 21–39.

 Google Scholar

Hewitt AE, Balks MR, Lowe DJ. 2021b. Gley Soils. In: The soils of Aotearoa New Zealand. Cham: Springer; p. 73–85.

 Google Scholar

Wilson AN. 1980. Soils of the Piako County. Wellington (NZ): Department of Scientific and Industrial Research. New Zealand Soil Survey Report 39.

 Google Scholar

Hughes JW. 2023. Late Quaternary paleoseismic history of the Te Puninga Fault, Hauraki Plains, northern New Zealand [unpublished MSc thesis]. Hamilton (NZ): University of Waikato.

 Google Scholar

Bronk Ramsey C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon. 51:337–360. doi:10.1017/S0033822200033865.

 Web of Science ®Google Scholar

Gerstenberger MC, Bora S, Bradley BA, DiCaprio C, Van Dissen RJ, Atkinson GM, Chamberlain C, Christophersen A, Clark KJ, Coffey GL, et al. 2022. New Zealand National Seismic Hazard Model 2022 revision: model, hazard and process overview. Lower Hutt (NZ): GNS Science. GNS Science report: 2022/57. doi:10.21420/TB83-7X19.

 Google Scholar

Persaud M, Villamor P, Berryman KR, Ries W, Cousins WJ, Litchfield N, Alloway BV. 2016. The Kerepehi Fault, Hauraki Rift, North Island, New Zealand: active fault characterisation and hazard. New Zeal J Geol Geop. 59:117–135. doi:10.1080/00288306.2015.1127826.

 Web of Science ®Google Scholar

Dowrick DJ, Rhoades DA. 2005. Revised models for attenuation of modified Mercalli intensity in New Zealand earthquakes. Bull N Z Soc Earthq Eng. 38(4):185–214. doi:10.5459/bnzsee.38.4.185-214.

 Google Scholar

Seebeck H, Van Dissen RJ, Litchfield NJ, Barnes PM, Nicol A, Langridge RM, Barrell DJA, Villamor P, Ellis SM, Rattenbury MS, et al. 2023. The New Zealand Community Fault Model–version 1.0: an improved geological foundation for seismic hazard modelling. New Zeal J Geol Geop. doi:10.1080/00288306.2023.2181362.

 Web of Science ®Google Scholar

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials [https://doi.org/10.21420/bh9t-fr52]. Active fault traces are currently in the processed of being incorporated into the publicly available NZ Active Fault Database [https://www.gns.cri.nz/data-and-resources/new-zealand-active-faults-database/].