Journal of Conference Abstracts

Temporal Variation in Magma Sources Related to the Impact of the Tristan Mantle Plume

S. A. Gibson¹ (sally@esc.cam.ac.uk), R. N. Thompson², A. P. Dickin³, J. G. Mitchell⁴ & S. C. Milner⁵

¹ Dept of Earth Sciences, University of Cambridge, UK.

² Dept of Geological Sciences, University of Durham, UK.

³ Dept of Geology, University of McMaster, Hamilton, Canada.

⁴ Dept of Physics, The University, Newcastle upon Tyne, UK.

⁵ Geological Survey of Namibia, Windhoek, Namibia.

The Early-Cretaceous impact of the Tristan mantle plume was associated with a diversity of magma types, including small-volume mafic potassic magmas and large-volume continental flood basalts (CFB). The earliest evidence of melt generation associated with the upwelling of the Tristan plume head is in northern Paraguay, where mafic potassic dykes yield K/Ar ages of 143 Ma on fresh phlogopite separates. These volatile-rich, small-volume melts are enriched in incompatible trace elements, have high La/Nb ratios (~2) and low ¹⁴³Nd/¹⁴⁴Nd_i (0.51182), suggesting that they were predominantly derived from ancient metasomatised subcontinental lithospheric mantle (SCLM). We believe that these magmas are evidence of early lithospheric extension and/or heat conduction, associated with the arrival at the Tristan plume on the base of the SCLM. The main eruption of the Parana-Etendeka CFB province occurred 137-132 Ma (Turner et al., 1994). Inversion modelling of the REE abundances in picrites (MgO=15 wt.%, Lu=0.18 ppm; epsilonNd=2.58) from the base of the lava pile suggests that initial melting of the plume-head occurred at ~120 km depth, beneath an 80 km thick lithosphere. Subsequent eruptions appear to have sampled plume-derived melts that had interacted with both fusible SCLM and crustal melts (Gibson et al., 1995). These constitute the major volume of the CFB pile. Post-CFB igneous mafic potassic igneous activity occurred in Namibia (133-129 Ma; Watkins et al., 1994), Paraguay (132-127 Ma) and southern Brazil (132-131 Ma; Renne et al., 1993; our unpubl. data). These 'late', relatively low-temperature, volatile-rich melts have variable incompatible trace-element (e.g. La/Nb=0.45-2) and isotopic ratios (¹⁴³Nd/¹⁴⁴Nd_i=0.51267-0.51163) due to the lateral SCLM chemical heterogeneity over this large area. Nevertheless, their widespread occurrence provides further evidence of localised rather than wholesale melting of the SCLM during the main CFB event.

Volcanic Plume Detection and Tracking Using Electric Fields

J. S. Gilbert (J.S.Gilbert@lancaster.ac.uk) & S. J. Lane

Environmental Science Division, Lancaster University, Lancaster LA1 4YQ, UK.

The presence of lightning in volcanic plumes alerts us to the fact that very high electric fields (or potential gradients) must be generated during explosive volcanic eruptions. Measurement of atmospheric potential gradients produced by vulcanian eruption plumes at Sakurajima in Japan suggests that, even in the absence of lightning, all particle-laden eruption plumes generate electric fields. We set up an array of electrostatic fieldmeters around Sakurajima volcano and collected potential gradient data simultaneously at four sites during an explosion. Our data suggest that, for the Sakurajima plumes, the ash carries a net negative charge and liquid droplets, sourced from condensed volcanic and entrained gases, carry a net positive charge. We have found that the form of the potential gradient perturbation is similar for each eruption. Highest perturbations of the ambient atmospheric potential gradients occur at the measurement sites directly downwind of the plume and increasingly negative potential gradients occur during fall of ash. We suggest that potential gradient monitoring may be used, in the absence of wind directional data and visual information, to indicate (1) the travel direction of a plume, and (2) the area of ash or mudrain fallout. Thus, the electrical properties of plumes have the untapped potential for use in the remote sensing of volcanic plumes. A simple theoretical model indicates that separation of volcanic gases and ash by a few tens of meters can generate the evolving potential gradients typical of vulcanian explosions.