A GEODETICALLY-CONSTRAINED PETROGENETIC MODEL FOR EVOLVED LAVAS FROM THE JANUARY 1997 FISSURE ERUPTION OF KILAUEA VOLCANO

Within a volcanic edifice, unerupted magmas may persist as mushy bodies for some time. New activity may remobilize these magmas, impacting lava compositions and eruption style, potentially leading to more violent eruptions in some low-viscosity volcanic systems. Here we test whether geochemical and petrological signatures of lavas erupted along the East Rift Zone of Kilauea Volcano on 30-31 January 1997 (Episode 54) can be explained by mixing between basaltic magmas and a partially crystallized intrusion from an earlier eruption, and if calculated mixing proportions are consistent with geodetic inversions of ground deformation and intrusion growth.

We use open-system phase-equilibria models to constrain the composition, degree of differentiation, and thermodynamic state of a rift-stored low-MgO magma body before mixing with degassed Pu’U ’O’o drainback and resident mafic magmas, shortly before fissure activity in Napau Crater began on 30 January 1997. Magma Chamber Simulator mixing models reproduce Episode 54 mineral and lava compositions within uncertainties, suggesting the identity of the low-MgO magma to be either variably differentiated remnants of unerupted magmas intruded into Napau Crater in October 1968, or a spatially and compositionally similar magma body.

Fissure A-E lavas from Episode 54 can be modeled as a mixture of 43% mafic magma and 57% residual melt left after 23% fractionation of the 1968 intrusion. Magmas formed by 35% fractionation of the 1968 intrusion, mixed with the same mafic recharge composition in a 60:40 ratio, replicate Fissure F lavas. Model mineral assemblages and compositions suggest that the residual melt in the rift-stored magma body was 40-50% crystalline at the time of mixing, and likely compositionally stratified. Our model results corroborate field and geochemical relationships and demonstrate that the pre-eruptive state of some partially crystallized magma bodies can be recovered by examining mineral compositions within mixed lavas. Discrepancies between geodetically-consistent mixing proportions and our mixing models highlight the uneven and complex nature of incomplete mixing—localized scales are recorded in erupted lavas, whereas geodetically-determined mixing proportions likely reflect large spatial scale contributions to the entire magma volume.