USING MULTIPLE THERMOCHRONOMETERS ACROSS THE MAXWELL LAKE DIKE COMPLEX TO INFER MAGMA TRANSPORT DURATIONS OF THE MAIN PHASE COLUMBIA RIVER FLOOD BASALT ERUPTIONS

The Columbia River flood basalt (CRB) province is the best exposed Large Igneous Province (LIP) on Earth, containing massive packages of rapidly emplaced lava flows. The largest of these, the Wapshilla Ridge Member, is one of the largest known eruption sequences on the planet. The emplacement mechanics of massive eruptions such as the Wapshilla remain debated, as does the influence of flood basalt eruptions on mass extinctions and climatic shifts. Part of the uncertainty surrounding the lifecycle of LIPs is due to poor constraints on the durations of individual eruptions, and thus, poor constraints on the volatile flux to the atmosphere during eruption. Here we present a novel use of thermochronology to quantify the duration of magma transport through individual dike segments of the CRB. We directly investigate a shallowly emplaced (~2 km paleodepth) feeder dike complex for the Wapshilla, exposed in the Wallowa Mountains of Oregon (the Maxwell Lake Dike Complex, MLDC), to determine how long the system was active, flow localization along strike, and the role of structural dike segmentation on emplacement. Using thermochronologic transects collected perpendicular to 5-8 m wide dike segments exposed in the MLDC we measure the width of the dike’s thermal aureole. Then, the thermochronologic results are combined with geochemical and structural analysis of the dikes and a 1-D thermal model to quantitatively estimate the duration of magma transport in each dike segment. We employed 2-6 thermochronometers, U-Th/He, fission-track, and Ar systems, at >5 segments along ~3 km of the MLDC. We compare thermochronometer profiles with a novel paleomagnetic geothermometer that leverages reset near-dike magnetic directions in combination with a thermal model to estimate dike longevity. We found segments with a bimodal distribution of geochemistry distributed in a structurally complex sequence of en echelon strands that suggest more than one episode of intrusion and fracturing. Co-inversion of thermochronometers and the magnetic geothermometer suggest that no segment was active for more than 10 years, and most segments were active for less than one year. Our observations suggest high variability in magma flow and flow localization within the MLDC; eruptions sourced from this dike complex were likely short lived but high intensity.