Paleo-quakes illuminate earthquake gates in New Zealand

A new study of New Zealand’s Alpine Fault reveals details about past earthquakes and the likelihood of a large future rupture.

By Fionna M. D. Samuels, Ph.D. Student, Temblor Earthquake News extern

Citation: Samuels, F. M. D., 2021, Paleo-quakes illuminate earthquake gates in New Zealand, Temblor,

New Zealand’s Alpine Fault cuts 530 miles (850 kilometers) along the western edge of Te Waipounamu — the South Island — and marks the boundary between the Australian and Pacific plates. A Few Māori oral histories and tree rings document the most recent rupture along this fault — likely a magnitude-8.1 in 1717. This fault releases a steady beat of large earthquakes through time, and scientists know that the area is due for another event. “Using a credible scenario for the Alpine Fault is the best way we can imagine what this event might be like — it gives us the power of the hypothetical,” says Caroline Orchiston, the science lead for Alpine Fault Magnitude 8 (AF8), a national program that links scientists with the public to prepare for the next Alpine Fault earthquake.

Satellite image of island partly covered with snow
The trace of the Alpine Fault made visible from space by snow from a blizzard. Credit: Jacques Descloitres, NASA (Public Domain)


A recently published study in Nature Geosciences by an international team of geoscientists combines detailed paleoseismic information, fault geometry and physics-based computer modeling to refine possible timing and magnitude of the next Alpine Fault earthquake. They found the fault has a 75% chance of a surface-rupturing earthquake on the order of magnitude-7.0 or greater in the next 50 years, well above the previously calculated 29% (Biasi et. al., 2015). They also found that if it does, in fact, rupture, this looming earthquake has an 82% chance of registering as a magnitude-8.0 event or larger.

Paleosiesmic records in lake-floor cores

The increased chance of an earthquake on the Alpine Fault results from more detailed paleoseismic data, says Nicolas C. Barth, a geologist at the University of California, Riverside, and a coauthor on the study. The new details come from drilling sediment cores from four lakes that lie along the central section of the Alpine Fault, Barth elaborates.

Photo of lake in foreground with reflection of mountains present in background
Lake Kaniere, one of the lakes from which sediment cores were drilled to determine past ruptures on the Alpine Fault. Credit: Aaron Rees (CC BY-SA 4.0)


Lead author Jamie D. Howarth, at Victoria University of Wellington, developed the protocol presented in the paper for using lake sediment cores to accurately date past earthquakes, says Barth. Specifically, they identified turbidites, which are distinct underwater deposits formed by slope failure, sometimes triggered by earthquakes. Howarth and his coauthors dated fossilized leaves found in the sediment layers above the turbidites that fell into the lake during or shortly after an earthquake, constraining the timing of deposition of the turbidite, and by inference, the earthquake.

From the cores, Howarth and his colleagues determined that the fault ruptures every 249 years, give or take 58 years. Considering the time since the last earthquake in 1717, they calculate a three-quarters chance of a magnitude-7.0 or greater rupture in the next 50 years. Moreover, the likelihood will rise each year until an earthquake strikes.

Combining this new paleosiemic lake dataset with previously published data from swampy areas on the southern section of the fault, Howarth and his coauthors created one of the most complete earthquake records of its kind. The researchers found evidence of 20 earthquakes over 4,000 years and almost 190 miles (300 kilometers) of the fault. Additionally, each core sampling site acts like a rupture length mile marker; if core samples from different sites experienced the same ancient quake, the fault must have ruptured at least the distance between them. This minimum rupture length establishes a minimum magnitude estimate.

Photo of rectangular sections of sediment on a laboratory table
Lake sediment cores used to date ancient earthquakes. Credit: Jamie Howarth


Earthquake gates

“There is this funny kink in the fault plane at the boundary between the central and southern sections,” says Barth, explaining that the angle of the fault changes from shallow in the north to much steeper in the south. “It has been at the back of my mind for years — do the earthquakes care about this spot or rip right through?”

The newly established paleoseismic record shows that about half of the ancient earthquakes terminated near the kink, with rupture lengths indicating these were likely magnitude-7.0 events. The remaining earthquakes tore through the kink as magnitude-8.0 or higher events, releasing more than 30 times as much energy. Further, the pattern is not random but appears periodic, says Barth, explaining that this kink acts as an earthquake gate — a location on a fault that either halts the quake or lets it pass. After a certain number of magnitude-8 or higher earthquakes make it through the gate, it closes and limits the maximum rupture length so that the next earthquakes are more likely to be magnitude-7 events.

“I like to think of it as someone directing traffic at a construction zone,” explains Barth. “Sometimes you get a green go sign along with the cars behind you, [and] other times you get a red stop sign until conditions change.” Understanding the status of the gate may hold the key to estimating the magnitude of the next quake.

Photo of person standing next rock near a shoreline
Barth working along the Alpine Fault. Credit: Andre Mere


Using modeling to determine the status of the gate

Howarth and his team turned to physics-based earthquake simulations on the Alpine Fault to produce 100,000 years of synthetic earthquakes. This gave the researchers a nuanced picture of possible paleo-quakes to compare with the considerably shorter amount of paleosiesmic data they collected. Only the model with the most realistic Alpine Fault geometry successfully produced synthetic earthquakes that mimicked actual paleoseismic patterns.

Because the code simplified the otherwise computationally intensive physical equations governing earthquake rupture, Barth says they could model very long records of earthquakes. After finding a model that matched observations from the past paleoseismic record, Howarth and colleagues then extrapolated their model into the future, finding an 82% chance the next earthquake will be at least a magnitude-8.0 event. “We think our study shows that this modeling approach can produce realistic earthquake records and provide new insights into what future earthquakes we can expect,” says Barth.

New predictions bring preparedness into the public focus

New Zealand’s population is familiar with earthquakes, with the country experiencing thousands of them every year. However, only a few events are strong enough to be felt, and even fewer cause damage.

“The new science has certainly stimulated interest in the risk presented by the Alpine Fault, as tends to happen when the fault is in the public eye,” Orchiston says. With the higher risk presented by this fault, she says, “[it’s] even more relevant and important for our society to get prepared.”

Panorama of small town street with cars and buildings
Franz Josef township, New Zealand, one of the towns built along the Alpine Fault. Credit: Greg O’Beirne (CC BY-2.5)



Biasi, G. P.; Langridge, R. M.; Berryman, K. R.; Clark, K. J.; Cochran, U. A. Maximum‐Likelihood Recurrence Parameters and Conditional Probability of a Ground‐Rupturing Earthquake on the Southern Alpine Fault, South Island, New ZealandML Recurrence Parameters and Conditional Probability of a Ground‐Rupturing Earthquake. Bulletin of the Seismological Society of America 2015, 105 (1), 94–106.

Howarth, J. D.; Barth, N. C.; Fitzsimons, S. J.; Richards-Dinger, K.; Clark, K. J.; Biasi, G. P.; Cochran, U. A.; Langridge, R. M.; Berryman, K. R.; Sutherland, R. Spatiotemporal Clustering of Great Earthquakes on a Transform Fault Controlled by Geometry. Nature Geoscience 2021, 14 (5), 314–320.

Upton, P.; Cochran, U.; Orchiston, C.; Howarth, J.; Pettinga, J.; Townend, J. Tercentenary of the 1717 AD Alpine Fault Earthquake: Advances in Science and Understanding Hazards. New Zealand Journal of Geology and Geophysics 2018, 61 (3), 247–250.