USING (U-TH)/HE THERMOCHRONOLOGY TO QUANTIFY THE VARIABILITY OF DIKE EMPLACEMENT ALONG STRIKE DURING THE MAIN PHASE OF THE COLUMBIA RIVER FLOOD BASALTS

Continental large igneous provinces (LIPs) are the most intense expressions of continental magmatism. Their voluminous deposits are temporally associated with global environmental crises, and the mantle melting that produces LIPs appears to have a variety of origins. A gap in our knowledge of these systems is the tempo of individual LIP eruptions, because geochronologic methods are not precise enough to measure the duration of ancient events that lasted months or decades. Here, we present new results from a novel approach to solving this problem in the ~16 Ma Columbia River Basalts (CRB): quantifying the timescale of magma transport through dike segments that fed the main phase of CRB using low-temperature thermochronology. In our study area, dikes that fed the voluminous Grande Ronde formation were emplaced into the ~125 Ma Wallowa batholith. We collected four new ~100-m long thermochronologic transects perpendicular to dike-wallrock contacts in the three-kilometer long Maxwell Lake Dike Complex (MLDC) to quantify the spatial extent of heating next to each dike segment, which scales with how long the dike was hot and transporting magma. Apatite and zircon (U-Th)/He (AHe, ZHe; nominal closure temperatures 40-80˚C and 140-220˚C, respectively) ages from samples far from each dike segment are ~90-110 Ma and unperturbed by Miocene heating, whereas samples close to the dike segments have ages that are reset to ~16 Ma. We observe ~16 Ma AHe ages at different distances from the dike-wallrock contact at the four dike segments: 3 m, 3.1 m, 5 m, and 16.9 m. This qualitative comparison suggests that each dike segment was actively transporting magma for different lengths of time. To quantify these timescales, we use a 1D thermal model for dike heating, linked to a chemical diffusion model of He thermochronometer resetting. Data are inverted with a Bayesian Markov-Chain Monte Carlo method to explore trade offs in variables, such as dike longevity and ambient temperature, to explain the spatial resetting patterns. Preliminary model results suggest that the four new dike segments were actively flowing for 0.7-2.2 yrs, which is a shorter duration than a previously studied MLDC dike segment with AHe ages reset 53.5 m from the dike-wallrock contact.