1.2 Seafloor Spreading
Harry Hess & the Seafloor Spreading Hypothesis
In 1962, Harry Hess of Princeton University published what he modestly called "an essay in geopoetry" — a paper titled History of Ocean Basins that would become one of the most influential contributions in the history of Earth science. Drawing on his wartime experience commanding a Navy transport ship equipped with a continuously recording echo sounder, Hess had mapped thousands of flat-topped seamounts (which he named guyots) and observed the morphology of the ocean floor firsthand.
Hess proposed that the ocean floor is not permanent but is continuously created at mid-ocean ridges by upwelling mantle material, spreads laterally away from the ridge axis, and is ultimately recycled back into the mantle at deep-sea trenches. In this model, the ocean floor acts as a conveyor belt — new crust is born at ridges and destroyed at subduction zones, explaining why no ocean crust is older than approximately 200 Ma despite the Earth being 4.54 billion years old.
Robert Dietz independently proposed a similar model in 1961, coining the term "seafloor spreading." Hess's contribution was more detailed and mechanistic, linking spreading directly to mantle convection cells as envisioned by Holmes. The concept elegantly solved several outstanding puzzles: the youth of the ocean floor, the topographic prominence of mid-ocean ridges, and the existence of deep-sea trenches as sites of crustal destruction.
Mid-Ocean Ridges & Ocean Floor Bathymetry
The discovery of the global mid-ocean ridge system was a pivotal step toward understanding seafloor spreading. The Mid-Atlantic Ridge was first identified during the laying of transatlantic telegraph cables in the 1850s, but its true extent was only revealed by systematic echo-sounding surveys conducted by Bruce Heezen, Marie Tharp, and Maurice Ewing at Columbia's Lamont Geological Observatory in the 1950s.
Tharp and Heezen mapped a continuous rift valley running along the crest of the Mid-Atlantic Ridge, extending over 65,000 km as a globe-encircling system of submarine mountain ranges. The ridge system stands 2–3 km above the abyssal plains and is characterized by high heat flow, shallow-focus seismicity, and active volcanism.
65,000 km
Total Ridge Length
2–3 km
Ridge Elevation Above Abyssal Plains
~2.5 km
Typical Ridge Crest Depth
The topographic profile of a mid-ocean ridge is controlled by thermal isostasy. As newly formed lithosphere moves away from the ridge axis, it cools, thickens, and subsides. The depth of the ocean floor as a function of its age follows the half-space cooling model prediction:
Ridge flank subsidence (half-space cooling model):
\[ d(t) = d_r + 2 \rho_m \alpha_v (T_m - T_s) \sqrt{\frac{\kappa t}{\pi}} \cdot \frac{1}{\rho_m - \rho_w} \]
where $d_r$ is the ridge crest depth, $\rho_m$ is mantle density,$\alpha_v$ is the volumetric thermal expansion coefficient, $T_m - T_s$ is the temperature contrast between mantle and surface, $\kappa$ is thermal diffusivity, and $\rho_w$ is water density. This predicts depth proportional to $\sqrt{t}$.
The Vine–Matthews–Morley Hypothesis
In 1963, Frederick Vine (a Cambridge graduate student) and Drummond Matthews published a landmark paper proposing that the pattern of magnetic anomalies on the ocean floor could be explained by the combination of seafloor spreading and periodic reversals of Earth's magnetic field. Lawrence Morley independently arrived at the same conclusion but had his paper rejected by Nature and the Journal of Geophysical Research before Vine and Matthews published.
The key insight was elegant: as basaltic magma erupts at a mid-ocean ridge and cools through the Curie temperature (~580°C for magnetite), the ferromagnetic minerals in the rock acquire a thermoremanent magnetization (TRM) aligned with the ambient geomagnetic field. If the field is in its normal polarity, the rock records a normal magnetization; if the field has reversed, the rock records a reversed magnetization.
The Magnetic Tape Recorder Analogy
As spreading carries newly magnetized crust away from the ridge axis on both sides symmetrically, the ocean floor acts as a magnetic "tape recorder," preserving a continuous record of geomagnetic polarity reversals. Normally magnetized crust produces positive magnetic anomalies (reinforcing the present field), while reversely magnetized crust produces negative anomalies (partially canceling the present field). The result is a symmetric, striped pattern of alternating positive and negative anomalies centered on the ridge.
The magnetic anomaly observed at the ocean surface is the sum of the induced and remanent contributions. Over normally magnetized crust, the remanent field reinforces the ambient field, producing a positive anomaly; over reversely magnetized crust, the remanent field opposes the ambient field, producing a negative anomaly. The anomaly amplitude depends on the magnetization intensity, the geometry of the magnetized body, and the survey altitude.
Geomagnetic Polarity Timescale (GPTS)
The geomagnetic polarity timescale provides the temporal framework essential for interpreting marine magnetic anomalies. It was built through two independent approaches:
Radiometric Dating of Lavas
Allan Cox, Richard Doell, and Brent Dalrymple at the USGS used K-Ar dating of young volcanic rocks combined with paleomagnetic polarity determination to establish the polarity sequence back to ~5 Ma. They identified the Brunhes (normal), Matuyama (reversed), Gauss (normal), and Gilbert (reversed) chrons, as well as short polarity events (e.g., Jaramillo, Olduvai).
Marine Magnetic Anomaly Profiles
The numbered marine magnetic anomaly sequence (Anomaly 1, 2, 3, ... 34) extends the GPTS back through the Cenozoic and Mesozoic. The current standard timescale (Cande & Kent 1995, updated by Ogg 2012) covers the last ~170 Ma with sub-million-year resolution, calibrated by tying anomaly identifications to biostratigraphic and radiometric age constraints.
The average reversal frequency has varied over geological time: the field reversed roughly every 0.2–0.5 Myr during the Cenozoic, but did not reverse for approximately 40 Myr during the Cretaceous Normal Superchron (~124–84 Ma). The physical causes of this variability remain an active area of geodynamo research.
Spreading Rate from Magnetic Anomalies
The fundamental measurement in marine magnetic studies is the spreading rate, calculated by correlating identified magnetic anomalies with their ages from the GPTS:
Half-spreading rate:
\[ v_{\text{half}} = \frac{\Delta x}{\Delta t} \]
Full spreading rate (both flanks):
\[ v_{\text{full}} = \frac{x_{\text{left}} + x_{\text{right}}}{\Delta t} \]
where $\Delta x$ is the distance from the ridge axis to an identified anomaly and $\Delta t$ is the age of that anomaly from the GPTS. Typical values range from ~10 mm/yr (ultra-slow, e.g., Gakkel Ridge) to ~160 mm/yr (fast, e.g., East Pacific Rise).
< 20 mm/yr
Ultra-slow (Gakkel, SW Indian Ridge)
20–55 mm/yr
Slow (Mid-Atlantic Ridge)
55–160 mm/yr
Intermediate to Fast (EPR)
Asymmetric spreading — where the half-rate differs on opposite flanks — is observed at many ridges, particularly the Mid-Atlantic Ridge. This can arise from ridge migration relative to the underlying mantle, asymmetric magma supply, or the capture of ridge segments by rift propagation.
Age of the Ocean Floor
One of the most profound consequences of seafloor spreading is that the ocean floor is geologically young. The oldest oceanic crust, found in the western Pacific (Pigafetta Basin) and western Atlantic, dates to only ~200 Ma (Jurassic). This is less than 5% of Earth's age. In contrast, continental crust preserves rocks as old as 4.03 Ga (Acasta Gneiss, Canada) and detrital zircon crystals as old as 4.4 Ga (Jack Hills, Australia).
The age distribution of the ocean floor is not uniform. Because subduction preferentially consumes the oldest oceanic lithosphere, the age-area distribution is strongly skewed toward younger ages. The mean age of the present ocean floor is only about 60 Ma. The total area of ocean crust as a function of age can be approximated by a triangular distribution modified by the time-varying history of plate boundary lengths and spreading rates.
Implications for Earth's Heat Budget
The continuous creation and subduction of oceanic lithosphere is the primary mechanism by which Earth loses its internal heat. The global rate of heat loss through the ocean floor is approximately 32 TW (out of a total of ~46 TW), making mid-ocean ridge volcanism and lithospheric cooling the dominant modes of planetary heat transfer.
Marine Heat Flow
Heat flow measurements on the ocean floor provided early quantitative support for seafloor spreading. Near ridge crests, heat flow is extremely high (often >200 mW/m²), reflecting the proximity of hot mantle material. Heat flow decreases systematically with distance from the ridge (and hence with crustal age), following a predictable cooling curve.
The half-space cooling model treats the oceanic lithosphere as a semi-infinite solid cooling from an initially uniform temperature. It predicts that surface heat flow decays as the inverse square root of age:
Half-space cooling prediction for heat flow:
\[ q(t) = \frac{k(T_m - T_s)}{\sqrt{\pi \kappa t}} \]
where $q(t)$ is the surface heat flux at age $t$, $k$ is thermal conductivity (~3.1 W/m/K), $T_m$ is the mantle potential temperature (~1350°C),$T_s$ is the surface temperature (~0°C at the seafloor), and $\kappa$ is thermal diffusivity (~10² m²/s, specifically ~1 × 10²² m²/s, i.e., ~10<sup>-6</sup> m²/s).
For young ocean floor (<80 Ma), the half-space model fits the observed heat flow data well. However, measured heat flow near ridges is systematically lower than predicted, due to hydrothermal circulation that extracts heat convectively through fractures in the young, permeable crust. For lithosphere older than ~80 Ma, heat flow (and ocean depth) flatten out relative to the $1/\sqrt{t}$ prediction, which is better explained by the plate cooling model with a fixed lithospheric thickness (~125 km):
Plate model asymptotic heat flow:
\[ q_{\infty} = \frac{k(T_m - T_s)}{a} \]
where $a$ is the plate thickness (~125 km). This yields $q_\infty \approx 33$ mW/m², consistent with observations of old oceanic lithosphere.
Symmetric Magnetic Anomaly Patterns
The definitive confirmation of the Vine–Matthews–Morley hypothesis came from detailed magnetic surveys across the Juan de Fuca Ridge (Raff & Mason 1961) and the Reykjanes Ridge south of Iceland (Heirtzler et al. 1966). These surveys revealed strikingly symmetric patterns of magnetic anomalies on opposite sides of the ridge axis, exactly as predicted by the tape-recorder model.
In 1968, Heirtzler, Dickson, Herron, Pitman, and Le Pichon published a comprehensive magnetic anomaly timescale extending back to Anomaly 32 (~73 Ma) by assuming a constant spreading rate on the South Atlantic Ridge. This was later confirmed and extended by Deep Sea Drilling Project (DSDP) results, which showed that the age of the oldest sediment resting on oceanic basement increased linearly with distance from the ridge — exactly as predicted by seafloor spreading.
DSDP/ODP Confirmation
Leg 3 of the Deep Sea Drilling Project (1968–1969) drilled a transect across the South Atlantic, recovering basal sediments whose biostratigraphic ages increased systematically with distance from the Mid-Atlantic Ridge. This provided the first direct, independent verification of the magnetic anomaly timescale and the seafloor spreading hypothesis, transforming it from hypothesis to established fact.