The VEI-7 Millennium eruption, Changbaishan-Tianchi volcano, China/DPRK: New field, petrological, and chemical constraints on stratigraphy, volcanology, and magma dynamics

Abstract

Field relations, petrography, bulk and micro-scale chemistry reveal that the most recent history of hazardous Changbaishan-Tianchi volcano should be revised with important implications for volcanic hazard in NE Asia. Currently, the two most recent large eruptions are identified separately as a VEI 5 trachytic Baguamiao eruption (BGM) and the much heralded VEI 7, late 946CE (Common Era) “Millennium” eruption (ME) of comendite. However, we find that the former is part of the latter based on the following evidence: (1) trachytic fallout of the BGM lies directly on the comendite tephra of the ME without any indication of depositional hiatus; (2) abundant mingled trachyte-comendite pumice in the tephra deposits; (3) similar chemistry of mingled pumice and its components to those in the BGM and ME products; (4) correlation of bimodal glass shard compositions in the distal ‘B-Tm’ ash from the Japan Sea with comendite and trachyte glass from the BGM and ME products. Based on the above evidence, we suggest that the great Millennium eruption of 946CE should be revised finally to include the BGM trachyte as its final stage. Furthermore, deposits attributed to two other trachytic eruptions in 1668 and/or 1702CE (also called Baguamiao by some authors), and 1903CE referred to in historic accounts were also examined. Our field observations, petrography, bulk and micro-scale chemistry combined with previously published Ra/Th ages indicate that all these trachytes are either primary deposits of the ME or its reworked deposits. Thus our findings do not support two separate post-ME eruptions and require that volcanic hazard assessment at Changbaishan volcano include this new interpretation.Recently published geochronological data integrated with our new petrochemical and volcanological framework informs the magma dynamics leading to the ME. The ME comendite, derived from a parental trachyte similar to the BGM started accumulating at shallow levels around 12ka to 8ka. Around 4 to 1.6ka the BGM trachyte sensu strictu separated from its basaltic parent and started accumulating and crystallizing separately from the ME comendite. Just prior to the ME eruption, mingling of trace amounts of a third more mafic hybrid is implicated by the composition of mafic glass selvedges. The strong evidence for mixing of all these endmembers in the eruption products suggests that recharge mixing and overturn of this predominantly comendite-trachyte system occurred during the ~1ka ME. The common occurrence of comendite and trachyte in the last 100ka (also maybe 1Ma) at Changbaishan-Tianchi suggests that conditions for trachyte-comendite magma interactions are prevalent in the magma system of Changbaishan and maybe crucial in catastrophic eruptions there.