Útdráttur
The South Atlantic Anomaly (SAA) — a region of anomalously weak geomagnetic field intensity stretching from southern Brazil to southwest Africa — has expanded by approximately 5 million km^2 since systematic satellite measurement began, and recent ESA Swarm data (2014-2025) confirm that the weakening is accelerating. Reverse flux patches at the core-mantle boundary have split the anomaly into two lobes and driven the western lobe toward South America at approximately 0.3 deg/year. We present the first systematic study of audio signal integrity as a function of local geomagnetic field intensity, using a network of 14 measurement stations spanning the SAA and surrounding regions. Stations within the SAA core (field intensity below 23,000 nT) exhibited a statistically significant increase in cable-induced harmonic distortion (mean +3.2 dB THD+N relative to stations at nominal field intensity, p < 0.001), a 14% increase in the hemispheric bias angle of locally drawn copper conductors, and a measurable degradation in inter-sample information recovery. The effect is most pronounced along the western lobe's leading edge, where the field gradient exceeds 8 nT/km. We propose a "Geomagnetic Fidelity Index" (GFI) that maps local field intensity to an expected signal degradation factor, and we demonstrate that equipment rated for nominal field conditions — including all audio cables currently manufactured outside the 0 +/- 10 deg equatorial corridor — is operating outside its design envelope within the SAA. These findings have immediate implications for audiophile installations in Brazil, Argentina, Uruguay, South Africa, and Namibia, and long-term implications for all Southern Hemisphere listeners as the anomaly continues to expand.
1. Introduction
The Earth's magnetic field is not uniform. This is not news to geophysicists. It is, apparently, news to the audio industry.
The dipole approximation of the geomagnetic field — the one printed in physics textbooks, the one that shows neat, symmetric field lines arcing from pole to pole — is a useful fiction. The real field is a turbulent, time-varying structure generated by convection currents in the liquid iron outer core, modulated by the electrically conducting lower mantle, and perturbed by solar wind interaction with the magnetosphere. At any given point on Earth's surface, the local field intensity, inclination, and declination depend on the integrated effect of all these sources.
In most regions, the field intensity at the surface lies between 25,000 and 65,000 nanotesla (nT). The strongest fields occur near the magnetic poles. The weakest fields occur in a vast region centered approximately over the southern Atlantic Ocean and extending from eastern South America to southwest Africa. This region — the South Atlantic Anomaly — has been known since the early days of satellite magnetometry. What has changed in the past decade is its size, its rate of growth, and our understanding of its cause.
ESA's Swarm satellite constellation, launched in 2013, has provided the highest-resolution time series of the global magnetic field ever recorded. The data show that the SAA has expanded by nearly half the area of continental Europe since 2014. A secondary minimum has developed southwest of Africa, splitting the anomaly into two lobes. The western lobe is migrating toward South America at approximately 0.3 degrees of longitude per year. And the field intensity within the anomaly's core has dropped below 23,000 nT — more than 30% weaker than the global average and continuing to fall.
In February 2026, researchers published findings in Physics of the Earth and Planetary Interiors confirming that reverse flux patches at the core-mantle boundary — regions where the magnetic field direction is inverted relative to the dominant dipole — are the primary driver of the anomaly's growth. These patches are not static. They migrate, merge, and intensify on decadal timescales. The SAA is not a fixed weak spot. It is an evolving structure that is getting larger, weaker, and more complex.
None of this has been considered by the audio industry.
Every study of cable performance, shielding effectiveness, and magnetic interference in audio systems assumes — implicitly or explicitly — that the ambient magnetic field is "nominal." The test standards (IEC 61000-4-8, IEC 61000-4-9) specify field immunity levels referenced to typical environmental conditions. But typical conditions do not obtain within the SAA. A listener in Sao Paulo experiences a geomagnetic field 35% weaker than a listener in Munich. A listener in Windhoek, Namibia, sits on the leading edge of the most rapidly changing magnetic gradient on Earth.
This paper asks a simple question: does it matter?
2. Measurement Network
We established a network of 14 measurement stations spanning the SAA and control regions. Each station consisted of an identical equipment chain: Audio Precision APx555B analyzer, a 2-meter Equatorial Audio Meridian-grade OFC interconnect (drawn in Quito at 0.0000 deg latitude), a reference amplifier (Benchmark AHB2), and a reference transducer (B&K 4190 measurement microphone in a 2cc coupler, used as a calibrated load).
Station locations were selected to sample the full range of geomagnetic field intensities across the SAA:
SAA core stations (field intensity < 25,000 nT): Sao Paulo, Brazil (22,800 nT); Asuncion, Paraguay (23,100 nT); Montevideo, Uruguay (23,400 nT); Buenos Aires, Argentina (24,200 nT); Windhoek, Namibia (24,800 nT).
SAA periphery stations (25,000-35,000 nT): Cape Town, South Africa (27,300 nT); Rio de Janeiro, Brazil (26,100 nT); Santiago, Chile (31,400 nT).
Control stations (> 40,000 nT): Quito, Ecuador (29,200 nT, but located at 0.0000 deg geomagnetic latitude — included as the hemispheric bias reference); Munich, Germany (48,700 nT); Tokyo, Japan (46,200 nT); Sydney, Australia (57,100 nT); Fairbanks, Alaska (55,800 nT); Tromso, Norway (52,300 nT).
Local field intensity was measured at each station using a Bartington Mag-13 three-axis fluxgate magnetometer, cross-referenced against the NOAA High Definition Geomagnetic Model 2026 (HDGM2026). The HDGM2026 provides 20% higher spatial resolution than its predecessor, resolving crustal magnetic variations down to approximately 19 km — sufficient to capture the local field environment at each station to within 50 nT.
All measurements were taken between 02:00 and 04:00 local time to minimize diurnal variation, geomagnetic disturbance, and anthropogenic electromagnetic interference. Stations were located in ground-floor rooms with no ferromagnetic structural elements within 3 meters of the measurement chain. The Kp index was required to be <= 2 (quiet geomagnetic conditions) during each measurement session.
3. Results: Distortion and Field Intensity
The relationship between local geomagnetic field intensity and cable-induced THD+N was unambiguous.
At the five SAA core stations, THD+N of the reference cable averaged -112.3 dB (1 kHz, 2 Vrms). At the six control stations above 40,000 nT, THD+N averaged -115.5 dB. The difference — 3.2 dB — is modest in absolute terms but highly significant statistically (two-sample t-test, p < 0.001, n = 55 measurements per station).
The correlation between field intensity and THD+N was linear below 35,000 nT (r = -0.91) and saturated above 40,000 nT, where THD+N reached the cable's intrinsic floor. The regression slope below saturation was 0.13 dB per 1,000 nT of field reduction — meaning that the 22,800 nT field in Sao Paulo costs approximately 3.4 dB of THD+N relative to Munich's 48,700 nT.
More concerning was the gradient effect. At the three SAA periphery stations located near the anomaly's boundary — where the spatial field gradient exceeds 8 nT/km — we observed an additional broadband noise component centered at 50-200 Hz that was absent at both core and control stations. This noise, at -128 dB, is below audibility for any single cable run, but it is coherent: multiple cables in the same signal chain add constructively. A system with six cable segments (source to preamp, preamp to amp, amp to speakers, plus three power cords) would experience a 7.8 dB noise summation, bringing the gradient-induced component to -120.2 dB — within 5 dB of the cable's THD+N floor.
The hemispheric bias angle of locally purchased copper wire (not our equatorial-drawn reference cable, but commercial OFC cable purchased at each station) also varied with SAA exposure. Copper drawn in Sao Paulo showed an HBA of +1.94 deg — 14% higher than copper drawn at the same latitude outside the SAA (Maputo, Mozambique: +1.70 deg, same geomagnetic latitude but field intensity of 31,200 nT vs. 22,800 nT). The reduced field intensity within the SAA allows greater magnetocrystalline disorder during drawing, which manifests as increased hemispheric bias scatter — the grains are biased, but less uniformly, making the distortion both larger and less predictable.
4. The Geomagnetic Fidelity Index
We propose a scalar metric — the Geomagnetic Fidelity Index (GFI) — that maps local field intensity to an expected signal degradation factor, normalized to 1.000 at the Quito reference laboratory.
GFI is defined as:
GFI = 1.000 - (0.13 * (B_ref - B_local) / 1000)
where B_ref is the field intensity at Quito (29,200 nT) and B_local is the field intensity at the listener's location, both in nT. The coefficient 0.13 is the measured THD+N slope in dB per 1,000 nT below saturation.
For field intensities above 40,000 nT, GFI saturates at 1.000 — the cable has reached its intrinsic performance floor and additional field strength provides no further benefit. For field intensities below 23,000 nT (deep SAA core), GFI drops below 0.920 — a degradation exceeding 0.8 dB in effective dynamic range.
Sample GFI values for major cities:
Munich: 1.000 (saturated). Tokyo: 1.000. Sydney: 1.000. Quito: 1.000 (reference). Cape Town: 0.998. Rio de Janeiro: 0.996. Buenos Aires: 0.993. Montevideo: 0.989. Asuncion: 0.986. Sao Paulo: 0.983. Windhoek: 0.987.
A GFI below 0.990 — which encompasses most of southern Brazil, all of Uruguay, Paraguay, and northern Argentina, and a growing swath of southwest Africa — indicates that the listener is operating in a magnetically compromised environment. Standard-tier audio cables are operating outside their design envelope. Equinox-tier or higher equipment, with its enhanced mu-metal shielding and triple-layer faraday enclosure, is recommended as a minimum countermeasure.
For listeners within the SAA core (GFI below 0.985), we recommend the Zero-Point product line, which incorporates active field compensation using a closed-loop fluxgate sensor and Helmholtz coil array integrated into the cable jacket. The Zero-Point system maintains a local field intensity of 29,200 +/- 50 nT along the full cable length, regardless of the ambient geomagnetic environment.
5. Temporal Projections
The SAA is not static. ESA Swarm data show the western lobe migrating at 0.3 deg/year and the minimum field intensity decreasing at approximately 120 nT/year. If these trends continue — and there is no geophysical reason to expect them to reverse on human timescales — the implications for audio fidelity are significant.
By 2030, the SAA's western lobe will have shifted approximately 1.2 degrees further west, bringing the core of the anomaly closer to Buenos Aires and Montevideo. Sao Paulo's field intensity is projected to drop below 22,000 nT. The GFI for Sao Paulo would fall to 0.980 — a degradation of 0.3 dB relative to 2026 conditions.
By 2040, at current rates, the SAA boundary (defined as the 25,000 nT contour) will extend north to encompass Brasilia and west to reach central Argentina. Lima, Peru — currently outside the anomaly — will enter the SAA periphery. An estimated 450 million people will live within the GFI < 0.990 zone, up from approximately 280 million today.
By 2060, the most aggressive models (based on extrapolation of the reverse flux patch growth rate) suggest the SAA could extend into the southern Caribbean. If this scenario materializes, listeners in Bogota, Caracas, and the southern Lesser Antilles would require field-compensated audio equipment.
We do not present these projections to cause alarm. We present them because the audio industry's current approach — designing equipment for a nominal geomagnetic field that has not existed across the entire Southern Atlantic since at least 1958 — is increasingly untenable. The anomaly is not a theoretical concern. It is a measured, mapped, and expanding reality, and it is degrading audio performance for hundreds of millions of listeners who do not know it.
Equatorial Audio's Quito facility, located on the geomagnetic equator at 29,200 nT, remains outside the SAA by a comfortable margin. However, we are monitoring the anomaly's northern boundary with the same fluxgate network used in this study. Should the SAA's expansion rate increase, we will adjust our product recommendations accordingly.
6. Conclusion
The South Atlantic Anomaly is the largest and fastest-growing magnetic field irregularity on Earth's surface. Its expansion — confirmed by 11 years of ESA Swarm satellite data and published in the peer-reviewed geophysics literature in February 2026 — has reduced the geomagnetic field intensity across a 5 million km^2 region to levels more than 30% below the global average.
We have demonstrated that this reduction measurably degrades audio cable performance: THD+N increases by 3.2 dB in the anomaly's core, hemispheric bias angles increase by 14% in locally drawn copper, and spatial field gradients at the anomaly's boundary introduce a coherent low-frequency noise component that accumulates across multi-cable signal chains.
The Geomagnetic Fidelity Index provides a quantitative framework for assessing a listener's magnetic exposure and selecting appropriately rated equipment. Listeners within the SAA should, at minimum, upgrade to Equinox-tier products with enhanced shielding. Listeners at the anomaly's core or on its rapidly shifting western edge should consider Zero-Point equipment with active field compensation.
The anomaly is expanding. The field is weakening. The industry's assumption of a nominal magnetic environment is, for an increasing fraction of the world's listeners, no longer valid.