Open University Uranium-Series Laboratory    
Earth and Environmental Sciences, The Open University, Milton Keynes, UK    

Axial Volcanic Ridges (AVRs) are common features of slow-spreading mid-ocean ridges and provide an accessible record of MOR volcanism and its evolution over time and space. However, the time-scale of AVR development is poorly constrained and is the focus of the U-series component of this project.

Oblique AVR fledermaus
Figure 1. Oblique view of the AVR at 45oN with 6x vertical exaggeration. AVR is under the cross-hairs, Median Valley mostly in blue.  

The great majority of spreading segments on slow-spreading Mid Ocean Ridges (90% on the Mid Atlantic Ridge) are characterised by AVRs. Generally, AVRs are a few kilometres wide and tens of kilometres long, are located above the plate boundary and constitute the youngest crust (Figure 1.).

Data and samples have been collected for this project during Cruise JC24 with RRS James Cook to a typical AVR near 45°30’N on the Mid-Atlantic Ridge and used among other platforms, the state-of-the-art ROV Isis (Figure 2.).

ROV Isis clamped
Figure 2. ROV Isis, clamped to its 'A'-frame before launch. Folded manipulator arm and sample basket on the front with cameras and lights.  
‘Field’ evidence, based on observations using Isis’ camera system and the bathymetry generated with its sonar survey tool EM2000 (Figure 3.), have revealed unparalleled detail of this volcanic terrain. Numerous open faults or ‘fissures’ are evident and indicate that seafloor tectonism accounts for a significant proportion of the total extension at the 45°30’N AVR. However, all observed rocks are volcanic pillow basalts or sheet-flows, or their talus, variably covered by pelagic sediment; no plutonic or tectonised rocks were encountered.
EM2000 bathymetry dive 91
Figure 3. EM2000 bathymetry of the area surveyed by ROV Isis on Dive 91, contours at 10m, max depth 3266m. Red silhouette of RRS James Cook for scale.  
Volcanic activity of the AVR is characterised by parallel N-S trending lineaments of cones (‘hummocks’). Cones have a typical diameter 100-200m and are 60-100m high and consist of various types of pillow basalt; individual cones may be monogenetic. The high relief terrain is approximately 4 km wide and 30 km long. Around the flanks of the AVR the topography is more subdued, typical of ‘sheet flow’ lavas, and bounded by the near-vertical faults of the ridge wall.

Cone sector collapse

Figure 4. Inset from Figure 3. showing recent sector collapse of a monogenetic cone. Red silhouette of RRS James Cook for scale.  

The long-term average spreading rate at this section of the Mid-Atlantic Ridge is ~10mm/y but the episodic nature of both volcanic and tectonic activity renders this average rather meaningless for a detailed discussion of the age relations of the observed phenomena. The observationally ‘monogenetic’ character of individual cones as well as observations of active pillow lava formation on Hawaii, indicate that volcanic activity of individual cones lasts maybe decades. Evidence of mass wasting and sector collapse (Figure 4.) and the observation that cones are build on relatively flat surfaces between the lineaments, indicates that cones are quite young. The age span of a substrate covered by a 200m diameter cone, assuming the average spreading rate for the substrate, is of the order of the 20,000y and this could be argued as the maximum age of an individual cone.
There is clear evidence of 1 or 2 extremely young kilometre-long volcanic lineaments on either side and parallel to the central lineament which has the shallowest bathymetry. Some of these lineaments diverge away from the general trend, other, much shorter 'spurs' are at right angles to the main trend.

U-series methodology can be used to constrain the age of a trace element fractionation process, or the rate of process progression. Evaluations of the results of U-series modelling based on dynamic melting for U and Th have shown:

Th-U fractionation is significant only at small melt fractions, <4% in the presence of garnet during a porous flow regime,
the earliest melt can take >800,000y to traverse a 10 km melt column (at 10mm/y mantle upwelling rate) from the solidus to the garnet-out isograd,
melt generation rates of 30-150 mg/m3 per year are quite small, (10-50ppb).
Melt transfer from the garnet-out isograd to the surface is quite fast and magma ponding is short relative to 234U-230Th half-lifes.
It is highly likely that melt generation and extraction are ‘steady-state’ processes and that lavas have similar U-series characteristics on extrusion. Standard radioactive decay equations can then the used to calculate the age difference between the youngest and older lavas. The age range that is covered by 234U-230Th half-lifes is from centuries to 350,000y.  

The age of the youngest lavas can be calculated using Ra disequilibrium, similarly, with the assumption that the 226Ra/Ba ratio of the lavas is constant on extrusion. The age difference can be calculated relative to the youngest lavas. The age range covered by 226Ra/Ba is from decades to 8000y. The age of even younger lavas can be calculated with similar assumptions using 210Pb/204Pb for samples <100y.


Samples that show 210Pb/204Pb disequilibrium are ‘zero-age’ with respect to 226Ra/Ba ages and samples with 226Ra/Ba disequilibrium are effectively ‘zero-age’ compared to 234U-230Th ages.

235U-231Pa calculated ages can serve as confirmation of 234U-230Th ages but the parent 235U isotope has only 0.7% abundance while the 238U parent of 234U has 99.3% abundance resulting in significantly larger uncertainty in calculated ages.  

The presence or indeed effect of magma chambers can be assessed using 226Ra-230Th to model the magma fractionation rate. This is based on the analysis of feldspar which has a high partition coefficient for Ra and is ubiquitous in some samples. The fractionation rate depends on power output of the system and if the steady-state heat-flow is known, magma-chamber dimensions can be calculated.

Samples for U-series dating were taken on board usually by gently chiselling the glass rind off protrusions of basalt pillows.
To Top

© Peter van Calsteren
Last updated: 23 December, 2011 10:57