Every seismograph on Earth picks up a hum that nobody has fully explained: a faint, precise vibration repeating exactly every 26 seconds, without stopping, without seasonal variation, traced to a single point in the ocean off West Africa and recorded continuously since 1962. Findings published in Communications Earth and Environment in May 2023 quantify the signal’s characteristics in detail, confirming its fixed source in the Bight of Bonny, a shallow coastal bay sitting at the edge of the Gulf of Guinea, and documenting the frequency jumps that make it stranger still.
The signal itself has a quality that puzzles seismologists. Most seismic noise is broadband and messy, the background rumble of waves, wind, and human activity spread across many frequencies simultaneously. The 26-second hum is nothing like that. It sits at a single, extraordinarily precise frequency of 0.038 Hz, meaning it completes 0.038 full cycles per second, or one full cycle every 26.3 seconds. Its spectral purity, meaning how tightly the energy is packed around that one frequency rather than spreading across a range, is high enough to place it in a category more commonly associated with tuning forks than with geological systems. It has been radiating from approximately the same location on the continental shelf of West Africa for over six decades without interruption, and the strongest evidence pins its source to a 20-kilometre stretch of seabed in water roughly 50 to 100 metres deep.
What makes the Bight of Bonny location significant is its geology. The continental shelf there is unusually steep relative to the surrounding seafloor, dropping from shallow coastal water to deep ocean over a short horizontal distance. Beneath that shelf break, the crust is crossed by fracture zones and hydrothermal conduits, places where superheated, mineral-laden water circulates through cracks in the rock at temperatures between 350 and 450 degrees Celsius, under pressures several hundred times greater than at the surface. Those fluids travel at speeds around 1,540 metres per second, faster than water at normal surface conditions but consistent with the physics of hot, pressurised brine deep in the crust. The combination of fluid speed and fracture length is where the 26-second period comes from, under at least one proposed explanation.
The acoustic resonance model works like a pipe organ. When a column of fluid is trapped inside a fracture of fixed length and subjected to continuous forcing from outside, it vibrates at a frequency set by its own dimensions. For a 20-kilometre fluid column at 1,540 metres per second, the calculation gives 0.03850 Hz, a period of 25.97 seconds. The match to the observed 0.0385 Hz signal is exact to the precision of the measurement. The model predicts not just the frequency but also the signal’s spectral bandwidth, the narrow spread around the central frequency, which requires the resonating system to have a quality factor above 20. A quality factor of 22 gives a predicted bandwidth of 0.00175 Hz, consistent with the observed value of less than 0.002 Hz. Ocean swells from Southern Ocean storms provide the continuous mechanical energy that drives the resonator, with wave heights between 1 and 5 metres coupling to the shelf break and feeding power into the fracture system below.
The Southern Ocean connection explains the energy supply but not the full picture. Seismic power at the Bight of Bonny increases by a factor of 16 during major storm periods compared to calm conditions, corresponding to a fourfold increase in wave amplitude. That amplitude variation tracks wave height precisely. The frequency of the signal does not shift at all in response to wave height changes, which is a key discriminating fact. If the ocean waves were directly setting the period of the hum, bigger waves would produce a different pitch. They do not. Frequency and amplitude respond to completely different drivers, which is what an acoustic resonator model predicts: the pitch is fixed by the physical dimensions of the resonator, while the volume responds to how hard the ocean pushes on it.
The stranger feature of the signal is what happens during its frequency glide events. Roughly every few days, the hum abruptly shifts pitch, moving from its baseline 0.038 Hz upward to as high as 0.050 Hz over a period of hours to a few days, then returning to baseline. These glides have no seasonal preference, ruling out seasonal storm patterns as a trigger. They occur with no consistent relationship to wave conditions in the Bight of Bonny. The mechanism that changes the pitch without changing the amplitude pattern, and does so episodically rather than continuously, has remained unresolved since the glides were first systematically documented.
A preprint submitted to EarthArXiv in March 2026 by independent researcher Paul Nicholas Hermatz proposes that the pitch shifts are driven by solar activity. The proposed mechanism runs as follows. When the sun ejects a coronal mass, a burst of charged particles and magnetic field that travels outward through the solar system, the leading edge strikes Earth’s magnetic shield within one to three days. The compression of that shield generates ultra-low-frequency electromagnetic pulsations, oscillations in Earth’s magnetic field in the 0.002 to 0.1 Hz range. Those pulsations penetrate the crust and modify the pressure in the fluid-filled fracture system beneath the Bight of Bonny. A fluid under higher pressure has a different acoustic velocity. A different acoustic velocity in the same 20-kilometre fracture produces a different resonant frequency. The size of the frequency shift corresponds mathematically to the size of the pressure perturbation. A major geomagnetic storm, rated above Kp 5 on the standard 0-to-9 index of magnetic disturbance, produces a perturbation to the fluid’s bulk modulus of approximately 19 percent, sufficient to shift the signal from 0.038 Hz to 0.042 Hz. The largest glides on record, reaching 0.050 Hz, correspond to a bulk modulus perturbation of 68.8 percent, consistent with a severe coronal mass ejection event.
The Hermatz model extends deeper than the crust. It proposes that the baseline level of geomagnetic activity, the average strength of Earth’s field at the boundary between the outer liquid core and the solid mantle above it, is itself modulated by waves inside the outer core on a six-year cycle. The outer core is a shell of liquid iron roughly 2,200 kilometres thick, conducting electric current and generating Earth’s magnetic field through its motion. That motion is not smooth. Torsional Alfvén waves, a specific type of oscillation in which cylindrical shells of the liquid core vibrate back and forth like a series of nested tubes, carry momentum across the core on a six-year period. The wave speed, called the Alfvén velocity, depends on the strength of the magnetic field inside the core and the density of the liquid iron. Using published values of 4 millitesla for the internal field strength and 11,000 kilograms per cubic metre for the core fluid density gives an Alfvén velocity of 3.4 centimetres per second. Crossing the outer core radius of 2.26 million metres at that speed and accounting for the geometry of the oscillation produces a torsional period of 6.32 years, consistent with the six-year periodicity observed in Earth’s rotation rate and magnetic field variations. Beneath even that, the model links to a 70-year oscillation in the inner core, the solid iron sphere at the planet’s centre, whose rotation rate has been documented to vary over multidecadal timescales.
This five-layer structure, ocean swell to shelf coupling to crustal resonance to solar electromagnetic modulation to core dynamics, is the Hermatz Effect. It treats the Sun not as an occasional external disturbance to Earth’s magnetic system but as a continuous participant in a feedback loop that runs from the solar wind down to the innermost solid layer of the planet. The Bight of Bonny fracture system, in this framework, functions as a transducer, a device that converts geomagnetic variability into a seismic signal detectable on every continent. Several of the model’s predictions are testable against data that already exists. If frequency glide events correlate with southward-oriented solar magnetic field conditions, which maximise the coupling efficiency between solar wind and Earth’s magnetosphere, the solar trigger hypothesis is supported. If they do not, the model is falsified at that specific point. The Swarm satellite constellation, operated by the European Space Agency and measuring Earth’s magnetic field continuously from low orbit, produces data that can be searched for electromagnetic anomalies above the Bight of Bonny in the one-to-seven-day window before glide events, a test that requires no new fieldwork.
What the model does not do, and what its author is careful to note, is claim peer-reviewed confirmation. The paper is a preprint, meaning it has not yet passed external scientific review. The core equations and the parameter values plugged into them are drawn from published literature, and the match between computed and observed frequencies is genuine rather than fitted, but the causal chain from solar wind to crustal fluid pressure has not been confirmed by direct simultaneous measurement. The existing catalog of frequency glide events from the 2023 peer-reviewed study can be cross-referenced against solar wind data from the NASA OMNI database and IMF Bz orientation records, a comparison that any research group with access to those open-access datasets can perform without additional cost. If that cross-reference produces a statistically significant correlation between southward IMF Bz events and glide onset times in the two to twenty-four hour lag window the model predicts, the solar hypothesis will have passed its first quantitative test. The 26-second hum has been logged on global seismic networks since 1962, and the full record at 0.03850 Hz remains active and uninterrupted at the Bight of Bonny today.
Source:
Hermatz, P.N. (2026). The Hermatz Effect: A Five-Layer Solar-Geo Dynamo Model for the Persistent 0.038 Hz Global Seismic Signal. EarthArXiv preprint. https://eartharxiv.org
