Abstract
In an earlier study, Ninkovich used paleomagnetic methods to date five mega-scopically distinguishable tephras in Pleistocene abyssal sediments east of the North Island, New Zealand. This has been extended by extraction and counting of the atmospherically transported volcanic glass fraction between 88 and 11 /ira in five cores, four of which have been paleomagnetically dated. Sedimentation rates are between 2 and 8 mm per thousand years. Two of the four cores feature discon-formities at the core surfaces, with the youngest sediments in excess of 1•5 m.y. age. A total of 17 separate tephras, only 8 of which are visible, are identified; the oldest was deposited 4 m.y. ago. An apparent volcanic glass accumulation rate of at least 100 mg per thousand years per cm2 is required in the 88—11 µm particle size range before any ash horizon is readily visible in any of the cores. The finely dispersed volcanic glass constitutes a much greater portion of the total glass content in the cores than do the visible tephras, placing limits on the tephra stratigraphy of the source region which can be inferred from the coarse ash fractions. Variations in apparent accumulation rates of the volcanic glass by up to two orders of magnitude in a traverse, normal to the suspected wind direction, permits identification of the probable average cloud axis for the eruptions. A net paleowind velocity of 40 km per hour is computed as an average for the Brunhes epoch. Microfeature analysis of the glass shards throughout one core, and use of recently developed empirical relationships between microfeatures and volcanic explosivities, suggest maximum explosive volcanic cloud heights of 25 km. This is equivalent to an explosion of at least 10 megatons of TNT. No conclusion can be made about the variation of volcanic activity over the 4 m.y. period represented in the cores, because low amounts of volcanic glass in the older cores can be caused by geographic as well as temporal factors. Climatic variation causing a change in ambient wind direction and velocity may also effect ash distribution.