Osmium isotope variations accompanying the eruption of a single lava flow field in the Columbia River Flood Basalt Province

详细信息    查看全文

• 作者：C. Vye-Brown^a ; ^b ; ^{cvye@bgs.ac.uk} ; A. Gannoun^a ; ^c ; ^{m.gannoun@opgc.univ-bpclermont.fr} ; T.L. Barry^a ; ^d ; ^{tlb2@le.ac.uk} ; S. Self^a ; ^{stephen.self@open.ac.uk} ; K.W. Burton^a ; ^e ; ^{kevin.burton@durham.ac.uk}
• 关键词：osmium isotopes ; Columbia River Flood Basalts ; crustal contamination
• 刊名：Earth and Planetary Science Letters
• 出版年：2013
• 出版时间：15 April, 2013
• 年：2013
• 卷：368
• 期：Complete
• 页码：183-194
• 全文大小：1732 K

文摘

Geochemical interpretations of continental flood basalts usually assume that individual lava flows represent compositionally homogenous and rapidly erupted products of large well-mixed magma reservoirs. However, inflated p¨¡hoehoe lavas may develop over considerable periods of time and preserve chemical variations that can be temporally linked through flow formation to eruption sequence thus providing an understanding of magma evolution over the timescale of a single eruption. This study presents comprehensive major, trace element and Re-Os isotope data for a single eruption that formed the 2660 km³ Sand Hollow flow field in the Columbia River Basalt Province, USA. Major and trace element variations accompanying flow emplacement (e.g. MgO 3.09-4.55 wt % , Ni 17.5-25.6 ppm) are consistent with fractional crystallisation, but other petrogenetic processes or variable sources cannot be distinguished. However, there is a systematic shift in the initial ¹⁸⁷Os/¹⁸⁸Os isotope composition of the magma (age corrected to 15.27 Ma), from 0.174 (lava core) to 1.444 (lava crust) within a single 35 m thick sheet lobe. Lava crust values are more radiogenic than any known mantle source, consistent with previous data indicating that neither an enriched reservoir nor the sub-continental lithospheric mantle are likely to have sourced these basalts. Rather, these data indicate that lavas emplaced during the earliest stages of eruption have higher degrees of crustal contamination. These results highlight the limitations of applying chemostratigraphic correlation across continental flood basalt provinces, the use of single data points to define melt sources and magmatic processes, and the dangers of using conventional isochron techniques in such basalt sequences for absolute chronology.