Lava Flow Emplacement Processes and Eruptive Characteristics of the Ontong Java Plateau: Inferences from High-Precision Glass Analysis

Abstract:

High-precision major and volatile element analyses were performed on natural basaltic glass from ODP Leg 192 Sites 1185 and 1187 of the Ontong Java Plateau (OJP) as a way to correlate lava flows within and between ODP drill sites. The ultimate goal is to estimate the dimensions, emplacement style, and eruption characteristics of the high-MgO Kroenke-type lavas: the youngest known flows at the two sites. The 122-Ma Ontong Java Plateau is the largest known magmatic event in Earth’s history, yet little is known of the emplacement style (e.g. flow dimensions and durations) of OJP lavas due to its submarine nature and burial beneath hundreds of meters of sediment. Basalt samples were recovered from 110- and 130-m thick core sections from Sites 1185B and 1187A, respectively. Total Kroenke-type lava thickness is 125 m at 1185B and >136 m at 1187. Site 1187A is located 146 km north of Site 1185B and lies ≈50 m shallower than Site 1187. Remarkably, all of the glass compositions from both sites fall on a common liquid line of descent, suggesting that all lavas were the product of a single eruption from a common magma chamber. The range of MgO compositions reflects a 20ºC range in temperature, representing ~1.9% crystallization of olivine + spinel. Using measured phenocryst abundance, we examine whether this crystallization occurred within the magma chamber or during long transport of lavas on the seafloor. More primitive lavas are present in the upper 30 m of Site 1185B (average of ~9.54 wt. % MgO), overlying more fractionated lavas (average of ~9.06 wt. % MgO). Lavas from Site 1187A bridge the gap between the high- and low-MgO groups of 1185B. In contrast to MORB, OJP glasses have no vesicles, suggesting they remained liquid for much longer during flow. Paleoeruption depths calculated from H₂O and CO₂ contents of glasses show no systematic variation with depth in Core 1185B, and range from ~2130-2650 mbsl, while Site 1187 shows deeper eruption depths of ~2410-3040 mbsl even though the corrected basement depth is ≈50 m shallower. Based on OJP’s 0.3° slope in this area, these depth differences indicate that 1185B flows may have traveled up to 55 km longer than those of 1187. We suggest that lavas flowed for >10s of km after erupting from widely spaced vents of a huge, shallow magma chamber that was compositionally zoned or was recharged over time.