| Summary: | Because of extensive volcanic production in Iceland during Holocene, many of the early postglacial large-volume fissure lavas cannot be traced easily to their eruptive crater solely by observations in the field. For example, the Bárðarbunga volcanic system has been suggested as a likely eruptive source of the large Holocene lava flows found in Bárðardalur valley, but this idea mainly relies on petrographic observations. A detailed mineralogical, chemical and isotopic study of the lavas in Bárðardalur is presented. For comparative purposes, we also targeted basement rocks of Bárðarbunga as well as several eruptive units in the region north of Vatnajökull (e.g., from Dyngjuháls). Disequilibrium features and Sr-isotopic ratios of plagioclase macrocrysts confirmed that the macrocrysts, which are found in various proportions in these lavas, are unrelated to their host melt. The macrocrysts were likely incorporated up by their carrier melt as it passed through a crystal mush zone on its way towards the surface. Based on detailed comparison, involving both chemical and radiogenic isotopic data, between the lavas from Bárðardalur and the eruptive units north of Vatnajökull, it appears almost inevitable that the Bárðardalur lavas belong to the Bárðarbunga volcanic system. In this respect, the Dyngjuháls region north of Vatnajökull is suggested as the most likely eruptive site of the lavas in Bárðardalur. Evaluation of the new data presented here, and a compiled dataset constructed for other volcanic systems of the NRZ, sheds light on possible limitations when assigning erupted material to its source volcano by means of chemical data. Furthermore, this study also demonstrates that further improvements in our understanding on the relative importance of different processes at play during petrogenesis of Icelandic basalts, are likely to be highly dependent on multi-parameter datasets generated on geologically well-characterized eruptive units.
|
|---|