HIGH-TEMPERATURE ULTRABASIC ERUPTIONS ON EARTH: VOLATILE CONTENTS, EMPLACEMENT MECHANISMS, AND THERMOMECHANICAL EROSION MECHAMISMS OF ARCHEAN KOMATIITES

Many interpretations of high-temperature ultramafic planetary volcanism are based on analogies with Archean komatiites, which are inferred to have had similar physical characteristics. Some komatiites contain vesicles and/or igneous amphibole, which have been used to suggest derivation by wet melting and eruption at lower temperatures and higher viscosities than normally assumed. However, most komatiites are strongly depleted in highly incompatible lithophile elements and have strongly positive eNd values, indicating derivation by high-degree partial melting of a long-term depleted, probably anhydrous source. Most komatiites likely contained insignificant amounts of volatiles and erupted as very high temperature (1360-1640ºC), very low viscosity (0.1-1 g cm sec-1), very mobile flows that incorporated volatiles during emplacement, accounting for the presence of vesicles and igneous amphibole in some units. Some workers have suggested that komatiites were inflationary and that cumulate zones grew by accumulation of primocrysts beneath thickening spinifex-textured crusts. Although some degree of endogenous growth must have occurred at some stage, textural, mineralogical, and geochemical variations in flows that have been studied in detail indicate that most cumulus olivines crystallized in situ and that most spinifex-textured zones crystallized after emplacement. Inflation probably occurred early and channelization in pre-existing or self-generated (thermomechanical erosional) topographic lows was more important in producing many thick komatiite flows. Field, geochemical, and isotopic studies of Ni-Cu-PGE mineralized komatiite flows indicate that they often occur in transgressive embayments in footwall rocks and exhibit evidence of local contamination and/or PGE depletion, consistent with thermomechanical erosion, assimilation of silicate xenomelts, and segregation of sulfide xenomelts. Whether or not komatiites flowed turbulently or not is an outstanding question that depended, at least in part, on eruption rates and slopes, but one that does not seem to be critical, as field, theoretical, and analog models of laminarly-flowing basalts have shown that they may also thermomechanically erode their substrates.