SUMMARY When a mantle plume with elevated temperature underlies an oceanic spreading centre it affects the generation of oceanic crust by creating thicker crust. We map the variation in crustal thickness and seismic velocity along three long‐offset seismic profiles acquired over oceanic crust generated shortly after continental breakup in the North Atlantic: a 212‐km‐long flowline from the Faroes rifted continental margin across crust of 51–42 Ma age, where oceanic spreading developed close to the inferred centre of the Iceland mantle plume; a 256 km flowline extending from the Hatton rifted continental margin across crust of 52–40 Ma age, about 800 km south of the presumed centre of the mantle plume; and a 99 km strike line over oceanic crust formed at 43 Ma in the Iceland Basin off the Hatton continental margin. The crustal velocity structure along each profile is constrained by multichannel seismic reflection data, which is used primarily to map the sediments, and by densely spaced ocean‐bottom seismometers, which recorded wide‐angle reflections and refractions to offsets of more than 100 km. Over 56 000 crustal diving wave and Moho wide‐angle reflection arrivals were used in joint crustal refraction and reflection tomographic inversions. Quantitative error analysis shows that the seismic velocity of the crust is mostly constrained to within 0.1 km s−1 and the depth of the Moho to within ±250 m. We interpret the crustal thickness and velocity changes along the profiles as caused primarily by changes in the mantle temperature at the time of crustal formation. If all the oceanic crustal thickness variations are ascribed to mantle temperature changes, we infer that as mature seafloor spreading developed following continental breakup, the mantle cooled by ca. 75 °C over a 10 Myr period, although it still remained hotter than the global average of normal oceanic crust. The crust formed close to Iceland is at all times thicker than that formed further away, which we interpret as reflecting higher temperatures close to the centre of the thermal anomaly created by the mantle plume. Currently at the Reykjanes Ridge, south of Iceland, we interpret thicker than normal oceanic crust as being caused by the presence of hotter mantle, modulated by thickness variations of 1.5–2.0 km which are attributed to temporal variations in the mantle plume temperature of about 25 °C on a 3–6 Myr timescale. A 1.5 km increase in thickness of oceanic crust generated between 48 and 45 Ma on the Faroes line is similar in magnitude and duration to those occurring on the present day Reykjanes Ridge, which we suggest is due to a temperature pulse of ∼25 °C. Gravity lineations in the northern North Atlantic suggest that the oceanic crust has exhibited small thickness fluctuations of similar size throughout its history, interpreted as due to small fluctuations in the temperature of the Iceland mantle plume.

No abstract

S U M M A R YSeismicity shallows towards the south along the Tonga-Kermadec-Hikurangi margin, deep and intermediate seismicity being absent altogether in the South Island of New Zealand. Beneath the Taranaki region of the North Island the maximum depth of the main seismicity is 250 km, but very rare events occur directly below at 600 km. These could be associated with a detached slab or a vertical, aseismic continuation of the subducted Pacific Plate. Six small events that occurred in the 1990s were recorded extensively by digital instruments of the New Zealand National Network (NZNN) and temporary deployments. We relocate these events by a joint hypocentre determination (JHD) method and find their focal mechanisms using first motions and relative amplitudes of P and S arrivals. The earthquakes relocate to a remarkably uniform depth of 603 ± 3 km relative error (±10 km absolute error) in a line 30-km long orientated 40 • NE, roughly parallel to the strike of the intermediate-depth seismicity. The only consistent component of the focal mechanisms is the tension axis: all lie close to horizontal and tend to align with the line of hypocentres. We interpret this deep seismic zone as a detached sliver of plate lying horizontally with the same orientation as the main subducted plate above. Volume change caused by a phase change controlled by the pressure at 600 km and temperature in the sliver produces a pattern of strain that places the sliver under tension along its length.

No abstract

No abstract