The MW7.8 14 November 2016 Kaikoura earthquake generated more than 10000 landslides over a total area of about 10000 km2, with the majority concentrated in a smaller area of about 3600 km2. The largest landslide triggered by the earthquake had an approximate volume of 20 (±2) M m3, with a runout distance of about 2.7 km, forming a dam on the Hapuku River. In this paper, we present version 1.0 of the landslide inventory we have created for this event. We use the inventory presented in this paper to identify and discuss some of the controls on the spatial distribution of landslides triggered by the Kaikoura earthquake. Our main findings are: 1) the number of medium to large landslides (source area 10000 m2) triggered by the Kaikoura earthquake is smaller than for similar sized landslides triggered by similar magnitude earthquakes in New Zealand; 2) seven of the largest eight landslides (from 5 to 20 x 106 m3) occurred on faults that ruptured to the surface during the earthquake; 3) the average landslide density within 200 m of a mapped surface fault rupture is three times that at a distance of 2500 m or more from a mapped surface fault rupture ; 4) the "distance to fault" predictor variable, when used as a proxy for ground-motion intensity, and when combined with slope angle, geology and elevation variables, has more power in predicting landslide probability than the PGA or PGV variables typically adopted for modelling; and 5) for the same slope angles, the coastal slopes have landslide point densities that are an order of magnitude greater than those in similar materials on the inland slopes, but their source areas are significantly smaller.
Recent earthquakes involving multiple fault ruptures highlight the need to evaluate complex coastal deformation mechanisms, which are important for understanding plate boundary kinematics and seismic and tsunami hazards. We compare ages and uplift of the youngest Holocene marine terraces at Puatai Beach and Pakarae River mouth (∼10 km apart) in the northern Hikurangi subduction margin to examine whether uplift is the result of subduction earthquakes or upper-plate fault earthquakes. From stepped platform-cliff morphology, we infer uplift during 2–3 earthquakes and calculate an average uplift-per-event of 2.9±0.5 m at Puatai Beach and 2.0±0.5 m at Pakarae River mouth. Radiocarbon ages from the youngest beach deposit shells on each terrace and a tephra coverbed on one terrace constrain the timing of earthquakes to 1770–1710, 1100–910, and 420–250 cal. B.P. at Puatai Beach, and 1490–1290 and 660–530 cal. B.P. at Pakarae River mouth. The ages differ at each site indicating uplift is neither the result of subduction earthquakes nor single upper-plate fault earthquakes. A reinterpretation of new and existing bathymetry and seismic reflection data, combined with dislocation modeling, indicates that near-shore fault segmentation is more complex than previously thought and ruptures likely involve multiple upper-plate faults. Future updates of the New Zealand National Seismic Hazard Model should revise the northern Hikurangi subduction seismic sources so that rupture does not uplift Puatai Beach and Pakarae River mouth and include new near-shore upper-plate faults as multifault sources.
On rocky tectonic coasts, data from Holocene marine terraces may constrain the timing of coseismic uplift and help identify the causative faults. Challenges in marine terrace investigations include: (1) identifying the uplift datums; (2) obtaining ages that tightly constrain the timing of uplift; (3) distinguishing tsunami deposits from beach deposits on terraces; and (4) identifying missing terraces and hence earthquakes. We address some of these challenges through comparing modern beach sediments and radiocarbon ages with those from a trench excavated across three terraces at Aramoana, central Hikurangi Subduction Margin, New Zealand. Sedimentary analyses identified beach and dune deposits on terraces but could not differentiate specific environments within them. Modern beach shells yielded modern radiocarbon ages, regardless of position or species, showing age inheritance and habitat is likely not an issue when dating shells on these terraces. By integrating terrace mapping, stratigraphy, morphology, and radiocarbon ages we develop a conceptual model of coastal uplift and terrace formation following at least two, possibly three, earthquakes at 5490-5070, 2620-2180, and 950-650 cal. yr BP. A high step and time gap between the upper two terraces raises the possibility that at least one intervening terrace is completely eroded. The trench exposure also showed that terrace stratigraphy may differ from that inferred from surface geomorphology, with apparent beach ridges being of composite origin and draping of younger beach deposits on the outer edge of a previous terrace. Dislocation modelling and comparison of marine terrace and earthquake ages from $4 km south and ≤ 73 km north confirms that the most likely earthquake source is the nearshore, landward-dipping, Kairakau Fault. Alternative sources, such as multi-fault ruptures of the Kairakau-Waim arama faults or Hikurangi subduction earthquakes, and/or a combination of the two are also possible and should be examined in future studies.
We use a mapped landslide inventory coupled with a 2-m resolution vertical difference model covering an area of 6,875 km 2 to accurately constrain landslide volume-area relationships. We use the difference model to calculate the source volumes for landslides triggered by the M W 7.8 Kaikōura, New Zealand, earthquake of 14 November 2016. Of the 29,519 mapped landslides in the inventory, 28,394 are within the analysis area, and of these, we have calculated the volume of 17,256 source areas that are ≥90% free of debris. Of the 28,394 landslides, about 80% are classified as soil or rock avalanches and the remainder as mainly translational slides. Our results show that both the soil avalanches and the rock avalanches, ignoring their source geology, have area to volume power-law scaling exponents (γ) of 0.921 to 1.060 and 1.040 to 1.138, respectively. These are lower than the γ values of 1.1-1.3 (for soil) and 1.3-1.6 (for rock) reported in the literature for undifferentiated landslide types. They are, however, similar to those γ values estimated from other coseismic landslide inventories. In contrast, for 50 selected rotational, translational (planar slide surfaces), or compound slides, where much of the debris remains in the source area, we found γ values range between 1.46 and 1.47, indicating that their slide surfaces were considerably deeper than those landslides classified as avalanches. This study, like previous studies on coseismic landslides, shows that soil and rock avalanches (disrupted landslides) are the dominant landslide type triggered by earthquakes and that they tend to be shallow.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.