New Zealand sees exotic earthquake sequence

A mystifying series of earthquakes that struck north of New Zealand last week may have resulted from a unique form of seismic triggering.

By Hector Gonzalez-Huizar, Ph.D., Ensenada Center for Scientific Research and Higher Education (CICESE) and Shinji Toda, Ph.D., IRIDeS, Tohoku University

Citation: Gonzalez-Huizar, H., Toda, S., 2021, New Zealand sees exotic earthquake sequence, Temblor,

On March 4, 2021, a series of three large earthquakes struck within six hours of one another in the South Pacific. The earthquakes struck along 560 miles (900 kilometers) of the Kermadec trench, where the Pacific and Australian tectonic plates converge. The first shock — a magnitude-7.3 — struck at 2:27 am local time in the southern part of the trench, just off the northeastern coast of New Zealand’s North Island (Te Ika-a-Māui). The resulting pattern of seismic waves suggests that it was a complex rupture along multiple adjacent faults.

Large earthquakes usually trigger aftershocks on nearby faults and in general, the largest aftershock is one unit of magnitude smaller than its mainshock — a magnitude-6.3 in this case. However, this earthquake was instead followed by both a magnitude-7.4 and a magnitude-8.1 earthquake along the same trench, just four and six hours later, respectively.

This progressive increase in magnitude with time is rare for an individual sequence, particularly for earthquakes this large. When this occurs, the events are simply renamed and we call the largest earthquake in the sequence the mainshock (in this case the magnitude-8.1), and those preceding it are named foreshocks.

Map of South Pacific
Red stars in the map represent the epicenters of the magnitude-7.3, 7.4 and 8.1 earthquakes that occurred on Mach 4, 2021. The blue arrow represents the direction of motion of the Pacific Plate relative to the Australian Plate. The inset figure shows a seismogram for that day, recorded by station ARMA.AU in Australia, highlighting the timing of the three earthquakes.


The short time between the three earthquakes suggests that they were part of a same foreshock-mainshock-aftershock sequence; however, the large distance between the first and the other two earthquakes — more than 560 miles (900 kilometers) — makes it difficult to establish a clear connection.

Triggering of earthquakes

When a fault slips, or ruptures, during an earthquake, rock masses on either side of the fault are displaced. This shifting of mass results in a redistribution of stress within the crust. Ruptures occur because of stress applied to the fault surface and when stress changes, new earthquakes are sometimes triggered on nearby faults. The potential for this so-called “static triggering” following a large earthquake can be quantified using a parameter called Coulomb stress change. In general, a stress change greater than 0.1 bars imposed on a particular fault suggests a high probability that an earthquake will occur within a relatively short period of time (Hill, 2008).

We estimate that the stress change imparted by the first of the earthquakes (the magnitude-7.3) on the rupture surface of the subsequent earthquakes (magnitude-7.4 and magnitude-8.2) was less than 0.01 bars. Therefore, it is very unlikely that the first earthquake caused the other two by static triggering. However, there are other mechanisms that might explain how the first earthquake could have indirectly triggered the other two, regardless of the large distance between them. One of these mechanisms is known as dynamic triggering.

Dynamic triggering of the magnitude-7.4 earthquake

When a fault slips, seismic waves radiate outward from the ruptured area. You feel these waves passing when the ground shakes during an earthquake. Not only can they damage buildings, but these waves can temporally increase the stress on other faults and trigger more earthquakes. Unlike static stress changes, these “dynamic” stress changes are transient, but they can be much larger at great distances.

It is possible that the passing of the seismic waves generated by the magnitude-7.3 caused temporary changes on faults several hundred miles to the north, resulting in their triggering. Given their depth and location relative to known geologic features, the magnitude-8.1 likely occurred on the interface between the two tectonic plates (the so-called “megathrust surface”) and the magnitude-7.4 likely struck on a tear fault in the descending Pacific Plate. We estimate the dynamic stress change imparted by the seismic waves from the magnitude-7.3 on the tear of the rupture area of the magnitude-7.4.

According to our calculations, stress change temporarily alternated between -0.1 to +0.1 bars as the seismic “wavetrain” from the magnitude-7.3 passed over the tear fault. This suggests that these passing seismic waves had a high potential to trigger the magnitude-7.4.

Computer model of subduction zone, colored by stress change
Model shows the temporal (dynamic) Coulomb stress change caused by the passing of the Rayleigh surface seismic waves, from the magnitude-7.3 earthquake, on a tear in the descending slab where the magnitude-7.4 earthquake occurred four hours later. Stress was obtained as in Hill (2008) and Gonzalez-Huizar and Velasco (2011).


The seismic surface waves — a specific type of seismic wave that travels only along earth’s surface — generated by the magnitude-7.3 took only around four minutes to arrive at the location of the magnitude-7.4. Yet the magnitude-7.4 occurred around four hours later. In general, instances of dynamic triggering are difficult to prove, especially when the triggered earthquake does not occur instantaneously with the arrival of the seismic waves. However, cases of “delayed” dynamic triggering are well documented. Seismologists think that in these instances, the stress changes caused by the passing of seismic waves cause permanent damage to fault contacts, a slow slip event or the intrusion of fluids into the faults, resulting in a slowly progressing process that ends with the earthquake triggering (Parsons, 2005; Shelly et al., 2011; Castro et al., 2015).

Interestingly, we found that this area has experienced several instances of delayed dynamic triggering in the past. We found that at least four other remote, large (greater than magnitude-8.0), recent earthquakes potentially triggered moderate (greater than magnitude-5.0) earthquakes in this area. At least one of these moderate earthquakes occurred within 15 hours after the seismic waves from the triggering earthquakes passed through the area. By comparison, there were no moderate magnitude earthquakes there in the previous three days.

Our preliminary results suggest that earthquakes at the Kermadec trench can be triggered by the small stress fluctuations, like those generated by the passing of seismic waves. Previous studies show that even the small stress changes generated by Earth tides are capable of controlling seismicity along the trench (Hirose et al., 2019), suggesting a high triggering susceptibility.

South Pacific with black stars near Japan, Papua New Guinea and Figi
Map showing the epicenter (black stars) of recent large (greater than magnitude-8.0) earthquakes that potentially triggered moderate (greater than magnitude-5.0) earthquakes near the source of the magnitude-7.4 earthquake (within the area limited by the red square). Also, the epicenter of the three earthquakes discussed in this study are shown (red stars).


Static triggering of the magnitude-8.1 earthquake

The magnitude-7.4 earthquake appears to have triggered the magnitude-8.1 earthquake by static triggering. The distance between their epicenters is only about 30 miles (50 kilometers) and they occurred about 100 minutes apart. In order to investigate how the magnitude-7.4 and magnitude-8.1 earthquakes increased the probability of future earthquakes, we estimated the stress transferred by these two earthquakes to nearby faults. This requires knowing the location, geometry and orientation of the faults — information that can be obtained from the analysis of the seismic waves of past earthquakes that occurred on those faults. Faults or fault segments are represented in maps using focal mechanisms, often referred to as “beachballs,” which indicate the orientation and direction of slip of the fault section that generated the earthquake. A long fault is more accurately represented by a series of slightly different beachballs rather than by a simple plane. We calculated the stress imparted by the magnitude-7.4 and magnitude-8.1 earthquakes on surrounding beachballs. For the magnitude-7.4 earthquake, we find a dense ‘halo’ of red beachballs near the epicenter, indicating that this earthquake transferred significant stress to surrounding active faults, bringing them closer to failure. A simpler way of looking at the static stress transfer is shown in the inset in the lower right. The future magnitude-8.1 rupture surface has a larger area of red than blue, indicating a net increase in its failure stress.

Below, we can see that after the magnitude-8.1, a core of blue beachballs highlights where the stress has dropped. But there are plenty of red beachballs as well, particularly north and south of the ruptured area. So, this sequence might not be over.

Maps with colored beachballs
Map shows stress transferred to beachballs by the magnitude-7.4 (left panel) and the magnitude-8.1 earthquakes (right panel). The orientation of each beachball provides information on the orientation of the fault (or fault section) to which the stress is transferred, and the direction of the potential slip if an earthquake is triggered. Color represents the amount of stress transferred. Inset shows the stress resolved on the future plane of the magnitude-8.1 caused by the magnitude-7.4 earthquake.



Parsons, T. (2005). A hypothesis for delayed dynamic earthquake triggering. Geophysical Research Letters, 32(4). doi:10.1029/2004GL021811.

Hill, D. P. (2008). Dynamic stresses, Coulomb failure, and remote triggering. Bulletin of the Seismological Society of America, 98(1), 66-92. doi:10.1785/0120070049.

Gonzalez‐Huizar, H., & Velasco, A. A. (2011). Dynamic triggering: Stress modeling and a case study. Journal of Geophysical Research: Solid Earth, 116(B2). doi:10.1029/2009JB007000.

Shelly, D. R., Peng, Z., Hill, D. P., & Aiken, C. (2011). Triggered creep as a possible mechanism for delayed dynamic triggering of tremor and earthquakes. Nature Geoscience, 4(6), 384-388.

Castro, R. R., González‐Huízar, H., Ramón Zúñiga, F., Wong, V. M., & Velasco, A. A. (2015). Delayed dynamic triggered seismicity in northern Baja California, México caused by large and remote earthquakes. Bulletin of the Seismological Society of America, 105(4), 1825-1835. doi:10.1785/0120140310.

Hirose, F., Maeda, K., & Kamigaichi, O. (2019). Tidal forcing of interplate earthquakes along the Tonga‐Kermadec Trench. Journal of Geophysical Research: Solid Earth, 124(10), 10498-10521.

Further Reading

Stein, R. S., Rollins, C., Sevilgen, V., and Hobbs, T. (2019), M 7.1 SoCal earthquake triggers aftershocks up to 100 mi away: What’s next?, Temblor,