Crystal cargo perspectives on magma assembly and dynamics during the 2021 Tajogaite eruption, La Palma, Canary Islands

The 2021 eruption of Tajogaite was the longest duration eruption, most voluminous, and had the largest human impact in recorded history on La Palma, Canary Islands. Extensive geophysical and geochemical data were collected during both the preceding unrest and eruptive event. Petrological monitoring was largely restricted to rapid stereo microscope observation and a few supporting in-depth studies using analytical instruments off-island. Here, we utilise time-series samples of lava and tephra from the Tajogaite eruption collected with near-daily frequency to understand the magmatic processes responsible for changes in petrological, geochemical, and geophysical observations. We combine published whole-rock major and trace element data with new QEMSCAN textural and mineral abundance data, major element analyses of the major macrocryst phases, and trace element data from clinopyroxene to illustrate magma plumbing system processes, supported by pressure and temperature modelling of mineral growth. Finally, we calculate olivine Fe-Mg diffusion timescales from early erupted tephra, and compare them with timescales of the climactic unrest period. Our data indicate that more-evolved and mineralogically-diverse magmas were tapped during the first week of the eruption, with little evidence for magma mixing. Magma mixing only becomes apparent when more primitive magmas erupted after the first ~10 days, exemplified by reverse-zoned olivines. Nonetheless, clinopyroxene barometry suggests that much of the erupted material is fed from the upper mantle at all stages of the eruption. Timescales of this process overlap with, and extend, the record of climactic geophysical unrest, suggesting that destabilisation of the magma system started before geophysical methods alone could detect and resolve variations from background. The chemical compositions of the crystal cargo are surprisingly uniform from Stage 2 (~5 - 10 days of activity) to eruption cessation (after 85 days of activity), and changes in whole-rock and tephra glass compositions observed by previous studies are not obviously mirrored in the mineral record. We highlight the importance of combining both whole-rock and mineral scale observations to understand how eruptions progress, and ultimately end.