Abstract
Ata is the most southerly island in a chain of young volcanic islands in the Tonga Group lying to the west of the Tonga trench. It is presently inactive and now consists of a small remnant (1.8 km2) of a large composite volcano which had at least four main centres of eruption.
The lavas are mainly porphyritic basaltic andesites, with minor phenocryst - poor basalts, and are dominanted by phenocrysts of calcic plagioclase. Clinopyroxene, mainly augite, is also a significant: phenocryst phase and orthopyroxene, olivine and microphenocrysts of titanomagnetite occur in small amounts. The groundmass textures are intergranular and
hyalo-ophitic and consist of plagioclase, augite, pigeonite, titanomagnetite, cristobalite, devitrification products after glass, brown glass, and two coexisting glasses which represent immiscible silicate liquids. Analyses of the coexisting glasses show that they are very similar to coexisting glasses found in lunar basalts. Some very interesting blocks of medium and coarsegrained, mafic to silicic rocks and some small andesite pumice blocks occur in a block breccia .
The basalts and basaltic andesites have a very restricted range of compositions (50-55% SiO2) with a pumice block sample being the most siliceous (62.25% SiO2). Chemically the rocks are similar to rocks from the other Tonga Islands and are more similar to those from the Kermadec Islands. Most of the chemical features are indicitive of the island arc tholeiitic series. Although there is only a mild iron enrichment and the K/Rb ratios are more like those from calc-alkaline series rocks.
Low pressure crystallisation (< 2-3 kb) prior to eruption is suggested by the mineralogy and mineral chemistry. Crystal fractionation which produced the variation in the rock compositions, and crystal accumulation, to form the cumulate rocks, are also considered to have occurred at this stage. The pH20 was low, probably less than saturated, and the f O2 is likely to have been close to, or slightly higher than, values shown by the quartz - fayalite - magnetite buffer curve. The evidence suggests plagioclase was the liquidus phase and crystallised at temperatures of about 1150-1190°C with pyroxene crystallising at 20-30°C lower.
The chemical evidence indicates a mantle origin for the magmas but the mineralogical evidence suggests a high temperature, very low pH2O magma, and thus a model of anhydrous melting of subducted oceanic crust followed by intermixing and equilibration with mantle peridotite is prefered.