7.3 Mantle Plumes & Hotspots
Deep-Rooted Upwellings
In 1971, W. Jason Morgan proposed that volcanic hotspots — isolated centers of persistent volcanism located far from plate boundaries — are surface expressions of narrow, hot upwelling plumes rising from deep in the mantle. This hypothesis elegantly explained the age-progressive volcanic chains like the Hawaiian-Emperor seamount chain: as a plate moves over a relatively fixed plume source, volcanism punches through the lithosphere at successive locations, leaving a trail of extinct volcanoes that records the plate's motion history.
The plume hypothesis has become one of the most debated topics in geodynamics. While the existence of deep-seated thermal plumes is supported by seismic tomographic imaging of low-velocity conduits beneath several hotspots, alternative explanations involving shallow mantle processes, lithospheric cracking, or chemical heterogeneities have been proposed for many supposed hotspots. The current consensus is that a subset of ~10 hotspots likely originate from the core-mantle boundary, while the remaining 20–40 proposed hotspots may have shallower or non-plume origins.
The Hawaiian-Emperor Chain
The Hawaiian-Emperor seamount chain is the classic example of a hotspot track and provides the most compelling evidence for the plume hypothesis. The chain stretches over 6000 km across the Pacific Ocean, from the active volcanoes of the Big Island of Hawaii (Kilauea and Mauna Loa, <1 Ma) to the 80 Ma Emperor seamounts near the Kamchatka-Aleutian junction.
The age of volcanic rocks increases monotonically with distance from the Big Island. The linear age-distance relationship gives a direct measurement of the Pacific plate velocity over the hotspot:
Hawaiian Chain (0–47 Ma)
The Hawaiian chain trends WNW (N60°W) from the Big Island through Midway Island (~28 Ma). The age-distance gradient yields a Pacific plate velocity of ~7 cm/yr over this interval. Islands become progressively more eroded, subsided, and eventually drowned as atolls and then guyots (flat-topped seamounts) with increasing age and distance from the hotspot.
Emperor Chain (>47 Ma)
North of the Hawaiian chain, the Emperor seamounts trend nearly due north (N10°W). The sharp ~60° bend at ~47 Ma was long attributed to a sudden change in Pacific plate motion direction. However, paleomagnetic evidence suggests the plume itself may have drifted southward by ~15° between 80 and 47 Ma, complicating the simple fixed-hotspot interpretation.
The Hawaii plume currently produces a volcanic output of ~0.2 km³/yr, making it one of the most volcanically productive hotspots on Earth. The total volume of the Hawaiian Ridge is estimated at ~106 km³ of basalt, all extracted from a mantle source region with an excess temperature of ~200–300 K above the ambient upper mantle.
Plume Head and Tail Model
Laboratory experiments and numerical simulations reveal that thermal plumes rising from a boundary layer develop a characteristic “mushroom” structure consisting of a large spherical head followed by a narrow conduit (tail). This structure has profound implications for surface volcanism.
Plume Head (1000–2000 km diameter)
The plume head entrains ambient mantle as it rises, growing to 1000–2000 km in diameter. When it impinges on the base of the lithosphere, it flattens and spreads laterally, causing uplift of ~1–2 km over a broad region. The massive volume of hot material undergoes partial melting, producing a flood basalt province (Large Igneous Province, LIP) — millions of cubic kilometers of basalt erupted over 1–3 Ma. The plume head arrival is often associated with continental rifting and breakup.
Plume Tail (100–200 km diameter)
After the head passes, the narrow conduit (tail) supplies a steady but much smaller flux of hot material. As the plate moves over this conduit, a linear volcanic chain is produced — the hotspot track. The volcanism is less voluminous than the initial flood basalt but persists for tens of millions of years.
The rise velocity of a buoyant plume head through the viscous mantle is estimated from the Stokes velocity for a viscous sphere:
\(v_{\text{rise}} = \frac{\Delta\rho \, g \, r^2}{3 \, \eta_{\text{ambient}}}\)
where Δρ is the density contrast (~20–30 kg/m³ for ΔT ~ 300 K), r is the plume head radius (~500 km), and ηambient is the surrounding mantle viscosity (~1021 Pa·s). This gives rise velocities of ~10–20 cm/yr, meaning a plume originating at the CMB can reach the surface in roughly 50–100 Ma.
The thermal buoyancy flux of a plume quantifies the rate at which buoyancy is transported upward:
\(B = \rho \, C_p \, \Delta T \cdot Q\)
where Cp is specific heat capacity (~1200 J/kg/K), ΔT is the excess temperature, and Q is the volume flux (m³/s). For the Hawaii plume, B is estimated at 6–9 Mg/s (megagrams per second), making it the most buoyant plume on Earth. In contrast, weaker plumes like the Cape Verde or Canary hotspots have B < 1 Mg/s.
Flood Basalts & Plume Head Arrivals
The plume head-tail model predicts that each long-lived hotspot track should begin with a massive flood basalt event. This prediction is remarkably well supported by the geological record:
| Flood Basalt (LIP) | Age (Ma) | Volume (km³) | Current Hotspot |
|---|---|---|---|
| Deccan Traps (India) | ~66 | >106 | Réunion |
| Paraná-Etendeka | ~132 | ~1.5 × 106 | Tristan da Cunha |
| North Atlantic (NAIP) | ~62 | ~6 × 106 | Iceland |
| Siberian Traps | ~252 | ~3 × 106 | None identified |
| Karoo-Ferrar | ~183 | ~2 × 106 | Marion / Crozet |
The coincidence of the Deccan Traps eruption at ~66 Ma with the end-Cretaceous mass extinction is widely studied. While the Chicxulub asteroid impact is considered the primary kill mechanism, Deccan volcanism likely contributed through CO2 and SO2 emissions that destabilized global climate in the hundreds of thousands of years bracketing the impact. Several other flood basalt events correlate with major extinction events (Siberian Traps with the end-Permian, CAMP with the end-Triassic), suggesting that plume head arrivals can trigger global environmental crises.
Seismic Evidence & the Plume Debate
The most direct test of the plume hypothesis is to image the predicted narrow, hot conduits in the mantle using seismic tomography. If plumes originate at the core-mantle boundary, they should appear as continuous low-velocity anomalies extending from the CMB to the surface beneath hotspot locations.
Recent advances in full-waveform tomographic inversions have provided increasingly compelling evidence for deep plume conduits. French and Romanowicz (2015) imaged broad low-velocity columns extending through the entire mantle beneath at least a dozen hotspots, many of which root at the edges of the two Large Low-Shear-Velocity Provinces (LLSVPs) in the deep mantle beneath Africa and the Pacific.
Evidence for Deep Plumes
- ▶Continuous low-velocity columns in tomographic models
- ▶High 3He/4He ratios indicating primitive, undegassed mantle source
- ▶Age-progressive volcanic chains with flood basalt initiation
- ▶Excess crustal thickness (e.g., Iceland, ~40 km vs normal ~7 km)
Arguments Against / Alternatives
- ▶Many hotspots show no age progression
- ▶Tomographic resolution may be insufficient for narrow conduits
- ▶Some volcanism may result from lithospheric extension and cracking
- ▶Shallow mantle heterogeneity or edge-driven convection as alternatives
The current consensus recognizes a spectrum of hotspot types: a small number (~10) of “primary” hotspots rooted at or near the CMB (Hawaii, Iceland, Louisville, Réunion, Kerguelen, Easter, Tahiti, Samoa, Afar, Tristan), and a larger number of “secondary” hotspots that may originate in the upper mantle or transition zone. The total number of proposed hotspots ranges from ~30 to ~50, depending on the classification criteria used.
Key Concepts Summary
ΔT ~ 200–300 K
Plume excess temperature above ambient
~47 Ma
Hawaiian-Emperor bend age
~10
Likely deep-seated primary plumes