Hawaii's Volcanic Earthquakes: Kīlauea and the Big Island

On May 3, 2018, residents of the Leilani Estates subdivision in the lower East Rift Zone of Kīlauea felt the ground rumble and watched cracks open in their streets. Within hours, lava began fountaining from fissures that opened along their neighborhood roads. Over the following four months, the eruption destroyed 716 homes, buried 35 square kilometers of land under new lava flows, and forced 2,500 people from their residences in the most destructive volcanic event in Hawaii in at least 200 years. What made 2018 remarkable from a scientific standpoint was not just the eruption's scale — it was how thoroughly the Hawaiian Volcano Observatory (HVO) had documented its approach. The seismic and deformation precursors that preceded the lower East Rift Zone (LERZ) eruption were visible weeks in advance, building in intensity in a pattern consistent with magma migrating from Kīlauea's summit reservoir down the rift zone toward the point of eventual eruption. 2018 was, in this sense, the most thoroughly predicted major volcanic eruption in American history — not because anyone could say exactly where or when, but because the seismological and geodetic monitoring network provided a clear, real-time narrative of what the volcano was doing underground.

This is what makes Hawaii — and the Big Island specifically — uniquely valuable to earthquake science. The island experiences roughly 20,000 earthquakes per year, making it one of the most seismically active places in the United States. But Hawaiian seismicity is not the monotonous background of plate boundary fracturing. It is a richly varied catalog of distinct phenomena: volcanic tremor generated by magma flowing through conduits; long-period earthquakes from pressurized fluid movement in the volcanic plumbing; swarms of high-frequency earthquakes marking the tip of an advancing dike; brittle tectonic earthquakes in the flanks of the volcanic edifice as it flexes under the weight of new lava; and occasional large flank earthquakes from slip on low-angle detachment faults at the base of the island — the type that produced the devastating 1975 M7.2 Kalapana event.

Understanding this diversity — why each type of seismicity exists, what physical process it reflects, and what monitoring agencies watch for to distinguish background from eruption precursor — is both intrinsically fascinating and directly practical. The Big Island has a resident population of approximately 185,000 people, with hundreds of thousands of tourists visiting the active volcanic zones each year. The relationship between the seismicity catalog and the volcano's current state is not academic for the communities living on its flanks.

The Geological Setting: A Hotspot Atop an Ocean Plate

The Hawaiian Islands sit atop the Pacific plate, which is moving northwestward over a fixed mantle hotspot at approximately 7–8 cm per year. The hotspot has been punching through the Pacific plate for at least 80 million years, creating the Hawaiian-Emperor seamount chain — a 6,000-kilometer-long track of volcanic islands and submarine seamounts whose ages increase from zero at the Big Island to 80+ million years at the Meiji Seamount near the Aleutian trench. The Big Island is the youngest and still-active expression of the hotspot, with Kīlauea and Mauna Loa currently receiving the bulk of the magma supply from the underlying mantle plume.

Unlike arc volcanoes fed by fluid-fluxed partial melting of the mantle wedge above a subducting slab, Hawaiian volcanoes are fed by deep mantle plume material. The magma is basaltic — low in silica, low in viscosity, poor in dissolved volatiles compared to arc magmas — which is why Hawaiian eruptions are typically effusive rather than explosive, producing lava flows rather than pyroclastic blasts. This low gas content and low viscosity also means that the seismic signals of Hawaiian magmatic activity differ from those of arc volcanoes: the absence of high-pressure volatile-driven explosive fragmentation means that the dominant seismic signatures are fluid-flow tremor and low-frequency resonance rather than the high-frequency impulsive signals of hydrofracture in more explosive volcanic systems.

The Sources of Hawaiian Seismicity: A Six-Way Classification

The approximately 20,000 earthquakes recorded on the Big Island each year come from six fundamentally distinct physical processes, each with characteristic seismic signatures that trained seismologists and automated classification algorithms can distinguish in the HVO catalog.

1. Volcanic Tremor

Volcanic tremor is the most distinctive seismic signal at Kīlauea — a continuous, sinusoidal ground vibration that persists for minutes to hours (or longer during active eruptions), with energy concentrated at specific narrow frequency bands between 1 and 10 Hz. Unlike the impulsive, broadband arrivals of tectonic earthquakes, tremor has no identifiable P-wave or S-wave onset — it simply appears and continues. On a seismogram, it resembles a sustained oscillation rather than a spike, and its amplitude waxes and wanes over time in patterns that correlate with eruption rate at the surface.

Tremor is generated by the flow of magma or magmatic gas through the conduit system — the resonance of fluid-filled cracks and channels in the rock at their natural frequencies, excited by the flowing fluid. The specific frequencies observed depend on the geometry of the resonating crack or conduit (its length, width, and fluid properties), and changes in tremor frequency during an eruption reflect changes in the physical state of the volcanic plumbing — changes in the gas content of the magma, changes in the conduit geometry from thermal expansion or assimilation of wall rock, or changes in the pressure of the flowing fluid. At Kīlauea, tremor is particularly well-developed during sustained lava lake activity at the summit and during active lava tube transport in the rift zones.

2. Long-Period (LP) Earthquakes

Long-period earthquakes at Kīlauea occur at depths of 1–5 km beneath the summit and rift zones, with dominant frequencies of 1–5 Hz and emergent, rather than impulsive, waveforms. Like tremor, LP events are thought to represent fluid resonance in the volcanic plumbing — the sudden crack opening or closing in response to pressure transients in the magma system, generating a resonant oscillation rather than a pure shear failure signal. Individual LP events are discrete, lasting 30–90 seconds, and occur in swarms during periods of volcanic unrest and inflation.

LP earthquake swarms at Kīlauea have been used operationally since the 1950s as an indicator of magma movement at shallow depth. A marked increase in LP seismicity rate — particularly accompanied by ground deformation — is one of the primary monitoring indicators that HVO watches for as a potential precursor to an eruption or intrusion event. The 2018 eruption precursor sequence included a significant increase in LP earthquake rate at Kīlauea's East Rift Zone beginning weeks before the LERZ eruption, consistent with pressurization of the rift zone plumbing.

3. High-Frequency (HF) or Volcano-Tectonic (VT) Earthquakes

High-frequency volcano-tectonic earthquakes are the most numerous type in the Kīlauea catalog — conventional shear failure events with clear P-wave and S-wave arrivals, located at depths of 0–10 km along the rift zones and beneath the summit caldera. They are caused by brittle fracture of the rock surrounding the magma plumbing — both the stress changes from magma pressure and the thermal fracturing of rock being assimilated or heated by adjacent magma. In the context of volcanic monitoring, VT earthquakes that occur in swarms migrating toward the surface along a rift zone are particularly significant: they trace the tip of an advancing dike — the seismic equivalent of hearing rock breaking ahead of the magma as it forces its way through the crust.

The 2018 LERZ eruption precursor sequence showed exactly this pattern: a swarm of VT earthquakes that began beneath the Pu'u 'Ō'ō cone in late April 2018 and migrated progressively eastward — downrift — over two weeks, tracing the path of a magma dike intruding from the established rift zone conduit system toward the lower rift zone where the eruption eventually began. The spatial migration rate of the VT swarm — approximately 2–5 km per day — allowed HVO to project the approximate trajectory of the intrusion days before surface eruption began.

4. Dike Intrusion Earthquakes

When a dike — a vertical or near-vertical sheet of magma intruding into the crust — propagates through rock, it generates a characteristic swarm of small to moderate VT earthquakes at its propagating tip, often accompanied by tremor behind the tip where magma is already flowing. The combination of a migrating VT swarm at the leading edge with sustained tremor in the trailing section is the classic seismic signature of dike propagation — distinguishable from purely tectonic earthquake swarms by the migration pattern and the accompanying tremor, and from pure tremor by the presence of the VT events at the advancing front.

Dike intrusions at Kīlauea do not always reach the surface to produce eruptions — some stall at depth and solidify, releasing their accumulated seismic energy in a final burst of VT activity as the magma crystallizes and contracts. These failed intrusions — detectable seismically but ultimately non-eruptive — are more common than successful intrusions and provide HVO scientists with a continuous stream of data on the sub-surface plumbing without the destruction of a surface eruption. The 1997 East Rift Zone intrusion, the 2007 Father's Day intrusion, and numerous smaller events documented in the catalog provide a rich comparative dataset for understanding what distinguishes intrusions that erupt from those that do not.

5. Flank Earthquakes: Tectonic Seismicity on the Volcanic Edifice

As magma is injected into Kīlauea's rift zones over thousands of years of volcanism, the volcanic edifice progressively expands toward the ocean — the rift zones act as jacks forcing the flanks outward. This outward pushing creates a system of gravitational instability: the south flank of Kīlauea is sliding seaward over a décollement fault at the base of the volcanic edifice (the basal sediment layer between the young basaltic shield and the older oceanic crust beneath it), driven by both the magmatic injection pressure from the rift zones above and the gravitational pull of the sloping flank.

This seaward sliding produces a class of tectonic earthquakes — flank earthquakes — that are fundamentally different from the volcanic seismicity discussed above. Flank earthquakes are conventional brittle-failure events on the décollement or on secondary faults within the volcanic edifice, with thrust-faulting focal mechanisms reflecting the seaward compression as the sliding flank is arrested by friction. They occur at depths of 7–15 km in the basal décollement and can reach magnitudes of M5–7, producing felt shaking across the island and, in the largest events, triggering tsunamis.

6. Deep Earthquakes: The Mantle Plume Zone

Beneath the volcanic edifice and the oceanic plate it sits on, the Big Island experiences a distinct population of deep earthquakes at depths of 30–60 km — within the oceanic mantle lithosphere and possibly extending into the mantle plume material itself. These deep events have tectonic focal mechanisms reflecting the stress state of the oceanic lithosphere being loaded by the weight of the volcanic edifice — essentially flexural earthquakes analogous to the outer rise earthquakes at subduction zones, driven by the bending of the oceanic plate under the enormous mass of Mauna Loa and Mauna Kea above it. The largest historical Big Island earthquake — the 2006 M6.7 Kiholo Bay earthquake — was a deep event of this type.

The 2018 LERZ Eruption: A Seismological Case Study

The 2018 lower East Rift Zone eruption provides the most comprehensive and best-documented modern example of how the different types of volcanic seismicity combine into a coherent precursor narrative that allows monitoring agencies to track volcanic activity in near-real time. It deserves detailed examination as a case study in operational volcano seismology.

The Precursor Sequence: April–May 2018

The 2018 sequence began not with the LERZ eruption itself but with weeks of gradual change in the state of Kīlauea's magma system. In late April 2018, the lava lake at Kīlauea's Halema'uma'u crater — which had been maintained continuously since 2008 — began rising anomalously rapidly, reflecting an increase in the magma supply from depth. Simultaneously, GPS stations along the East Rift Zone recorded accelerating ground inflation, indicating that the rift zone plumbing was being pressurized by the increased magma supply.

On April 30, the Pu'u 'Ō'ō cone — which had been the site of nearly continuous eruption since 1983 — suddenly deflated and collapsed as its lava lake drained rapidly, indicating that the magma supply to Pu'u 'Ō'ō had been diverted. Within hours, a swarm of VT earthquakes began beneath the East Rift Zone between Pu'u 'Ō'ō and the coast — the seismic signature of a new dike propagating eastward into the lower rift zone. The swarm migrated progressively eastward over the following two days at rates of 2–5 km per day, tracking the dike's advance through the seismogenic crust in near-real time on HVO's monitoring screens.

By May 3, the dike had reached Leilani Estates. The first surface fissures opened that evening, and the eruption that would ultimately produce one of the most significant lava flow events in Hawaiian history had begun. The seismic network had provided approximately 72 hours of advance warning that a rift zone intrusion was underway and moving toward the lower rift zone — time that allowed partial evacuation planning even if the exact eruption location could not be precisely predicted until the fissures opened.

Summit Collapse Earthquakes: May–August 2018

As the LERZ eruption progressed and enormous volumes of magma were exported from Kīlauea's magma system to the eruption site, the shallow reservoir beneath the summit caldera drained progressively. The roof of the emptying reservoir — the floor of Halema'uma'u crater — began subsiding in discrete increments, each subsidence event generating a distinctive seismic signal unlike anything in the routine Kīlauea catalog.

The summit collapse earthquakes of 2018 occurred roughly every 25–30 hours during the peak of the LERZ eruption, each generating M5.0–5.4 events with a very specific seismic signature: an implosive first motion (compression on all azimuths, indicating volumetric collapse rather than shear failure), long-period waveforms consistent with a large collapse event rather than a shear earthquake, and ground deformation measured by nearby GPS stations showing 0.3–0.5 meters of subsidence per event. Over the course of the eruption, the summit caldera subsided by approximately 500 meters as 0.8 km³ of material collapsed into the draining magma reservoir below — one of the largest volcanic collapse events in modern Hawaiian history, generating more than 60,000 earthquakes over four months and visible from space in satellite imagery.

The Hilina Slump: Hawaii's Largest Geological Hazard

The flank earthquake system at Kīlauea includes one of the most watched geological features in the Pacific Ocean: the Hilina slump. This enormous mass of the Big Island's southern flank — approximately 5,000 km³ of material, comparable in scale to the entire island of Oahu — is slowly sliding seaward at rates of approximately 6–10 cm per year, measured by GPS networks on the flank surface. The sliding is accommodated by a system of seaward-dipping normal faults (the Hilina fault system) at the surface and by the basal décollement at depth — the same fault system that generates the flank earthquakes discussed above.

The geological record shows that the Hilina slump has experienced dramatic, rapid seaward collapses in the past — submarine landslide deposits mapped on the ocean floor south of the Big Island include debris avalanche deposits that can only have been produced by catastrophic failure of the volcanic flank. The largest of these prehistoric collapses — the Alika 2 debris avalanche, dated at approximately 127,000 years ago — displaced an estimated 1,000–2,000 km³ of material into the Pacific in what was probably a geologically rapid event, generating a mega-tsunami with estimated run-up heights of 100 meters or more on neighboring islands.

The 1975 Kalapana Earthquake: The Modern Flank Event

On November 29, 1975, a M7.2 earthquake struck the south flank of Kīlauea near Kalapana — the largest earthquake in Hawaii in more than a century and the clearest modern example of what a major Hilina-system flank earthquake looks like. The earthquake occurred on the basal décollement at approximately 10 km depth, with a thrust-faulting mechanism reflecting seaward motion of the volcanic flank. It generated a local tsunami with maximum run-ups of 14 meters on the Ka'ū coast — killing two people at a beach campsite and damaging several coastal communities. The seafloor south of the earthquake's epicenter dropped by 3–8 meters and moved seaward by 5–8 meters, measured by comparison of pre- and post-earthquake bathymetric surveys.

The Kalapana earthquake was preceded by approximately 36 hours of elevated seismicity on the south flank — a swarm of events that, in retrospect, traced the aseismic slow slip that propagated across the décollement before the final dynamic rupture. This precursory slow slip pattern is analogous to the foreshock and slow slip sequences observed before some large subduction zone earthquakes, and it has subsequently been recognized in the instrumental record of earlier large flank events. Whether the monitoring network's current sensitivity would provide actionable warning of an impending large flank earthquake remains an open research question — the 1975 sequence suggests that some precursory signal exists, but its reliability as a short-term warning indicator has not been firmly established.

Mauna Loa: The Other Giant

Kīlauea receives most of the scientific attention because it is the more continuously active volcano and because its plumbing system is shallower and better instrumented. But Mauna Loa — the Earth's largest volcano by volume, rising 4,169 meters above sea level and extending below the ocean surface to total height of approximately 9 km above the ocean floor — poses its own seismic hazard profile and has its own distinct monitoring picture.

Mauna Loa last erupted in November–December 2022, producing lava flows that advanced toward Hilo before stalling. The eruption was preceded by weeks of elevated seismicity and ground inflation that allowed HVO to elevate the alert level before the eruption began — a practical success of the monitoring system analogous to the 2018 Kīlauea case. Mauna Loa's rift zones and flank also experience their own seismicity populations, including flank earthquakes on its northeast and southwest rift systems and deep earthquakes from lithospheric flexure beneath the enormous volcanic load. The 1950 Mauna Loa eruption — which sent lava flows to within 10 kilometers of Hilo in less than two weeks — illustrates the potential for rapid lava flow hazard from this volcano that its current state of apparent quiescence does not eliminate.

The Hawaiian Volcano Observatory: Monitoring at the Limit

The Hawaiian Volcano Observatory, established in 1912 by Thomas Jaggar of MIT — making it one of the oldest volcano observatories in the world — operates one of the densest seismic monitoring networks on any active volcanic system globally. The current network includes approximately 60 seismograph stations on the Big Island, with station spacing of 5–15 km in the most active areas near Kīlauea's summit and rift zones, achieving detection thresholds of approximately M0.5–1.0 across the entire volcanic system.

The seismic network is complemented by a similarly dense GPS and tiltmeter network (approximately 50 continuous GPS stations), a volcanic gas monitoring network (measuring SO₂ emission rates as a proxy for magma supply), and periodic airborne surveys. The integration of these multiple data streams into a coherent real-time picture of volcanic state is the operational core of HVO's monitoring mission — and the 2018 experience demonstrated both the capability of this integrated system and the fundamental limitations of forecasting individual eruption events even with world-class monitoring in place.

The Seismology of Volcanic Hazard: What HVO Watches For

For residents and visitors on the Big Island, translating the continuous seismicity of Kīlauea and Mauna Loa into an intuitive understanding of what is and is not hazardous is genuinely difficult. Most of the 20,000 earthquakes per year are entirely routine — the background seismicity of a continuously active volcanic system going about its normal business of supplying and redistributing magma through its plumbing. The signals that actually indicate elevated hazard are specific and require context to interpret correctly.

HVO scientists watch for several specific patterns that distinguish routine background from genuine elevated hazard:

• Rapid increase in VT earthquake rate beneath a rift zone, particularly if accompanied by ground inflation and migration of seismicity toward shallower depths — indicating dike intrusion
• Onset of tremor beneath the rift zone in a location without active eruption — indicating magma movement into a new area
• Rapid deflation of the summit after a period of inflation — indicating drainage of the summit reservoir toward a rift zone eruption site
• Unusual LP earthquake swarms at depth beneath the summit — potentially indicating magma recharge from below
• Anomalous acceleration of GPS velocities on the south flank — potentially indicating increased stress on the Hilina system
• Large M5+ flank earthquakes — both as hazards in themselves and as potential triggers for Hilina slump acceleration

None of these signals is a perfect predictor of eruption — volcanic systems are inherently probabilistic, and many dike intrusions that produce VT swarms and ground inflation do not ultimately erupt. But the combination of multiple concurrent signals — inflation plus VT swarms plus tremor plus rift zone migration — is a strong indicator that an eruption is likely and imminent, as 2018 demonstrated.

The Visitor Dimension: Seismicity and Public Safety

The Big Island's status as a major tourism destination creates a specific challenge for communicating volcanic seismicity to a population that changes completely every few days. Hundreds of thousands of visitors per year enter Hawai'i Volcanoes National Park — many of them seeking the visceral experience of active volcanism with little scientific background for interpreting what they observe or feel. The park's visitor center and ranger staff invest heavily in communicating the distinction between the routine seismicity of a continuously active system and the elevated-hazard signals that would justify evacuation or altered behavior.

The 2018 eruption, which began in the lower East Rift Zone outside the national park boundary but ultimately impacted communities that had lived adjacent to the rift zone for decades, illustrated both the challenge of communicating volcanic hazard to long-term residents and the importance of having a monitoring system capable of providing the hours of advance warning that allowed partial evacuation of Leilani Estates before fissures opened in the streets. For volcanic seismology, 2018 was not a failure — it was a validation that continuous, multi-parameter monitoring of an active volcanic system can provide meaningful advance warning of major eruptive events even when the precise eruption location and timing cannot be predicted with precision.

Conclusion

The Big Island of Hawaii is, from the perspective of seismology, one of the richest natural laboratories on Earth. It produces a continuous, varied catalog of earthquake types that collectively illuminate nearly every aspect of how volcanic and tectonic processes interact — from the fluid-resonance tremor of magma flowing through conduits to the brittle fracture of rock at dike tips to the flank earthquakes of a volcano slowly collapsing into the sea under its own weight. Kīlauea specifically, with its world-class monitoring network and its history of accessible, well-documented eruption precursor sequences, has provided more empirical data on volcanic seismology than any other volcano in history.

What that data has shown — most clearly in 2018 — is that the seismic and deformation signals of a major eruption precursor are detectable days to weeks in advance, interpretable in near-real time by a trained monitoring team, and translatable into public hazard communication that can save lives when combined with effective emergency management. It has also shown that the same monitoring network that provides eruption warnings sits atop one of the most structurally unstable large volcanic edifices on Earth — the Hilina slump — whose long-term behavior makes Kīlauea's south flank one of the most carefully watched geological features in the Pacific. The seismicity of the Big Island is not just a scientific curiosity. It is the living record of a planet reshaping itself in real time, watched by a monitoring network that translates each tremor and shockwave into an incrementally better understanding of when and how the next eruption will come.