Co-inversion of multiple thermochronometers and a paleomagnetic thermometer at a Columbia River flood basalt feeder dike (Oregon, USA) demonstrates the sensitivity of thermal history results to diverse data constraints

Abstract The high-volume, short-duration magmatism that forms large igneous provinces (LIPs) has been repeatedly linked to past perturbations of Earth's climate and biosphere. To better understand the dynamics of LIPs, and thus accurately assess their impacts, the durations of individual LIP eruptions need to be measured at the scale of years to decades. Here, we quantify how long one dike segment in the Miocene Columbia River flood basalt province (northwestern USA) actively transported magma during a LIP eruption using an unusually large and diverse thermochronologic, paleomagnetic, and stable isotopic dataset collected in the dike's Cretaceous country rocks. We expand a published Bayesian Markov-chain Monte Carlo approach for systematically predicting dike emplacement conditions (the duration of magma flow as well as ambient temperature and thermal conductivity of the country rocks) and use it to jointly invert six different combinations of new and published (U-Th)/He, 40Ar/39Ar, fission-track, and paleomagnetic data, which collectively have temperature sensitivities ranging from ~60 °C to 580 °C. All inversion results suggest that the feeder dike was actively transporting magma for <10 years, and the results are not sensitive to noble gas diffusion kinetics. We find that jointly inverting all available datasets narrows the range of acceptable dike lifetimes (1.4–2.9 years) and documents anomalously hot ambient temperatures (77–92 °C). New apatite δD data document an isotopic depletion that supports previous δ18O evidence of a fossil hydrothermal system next to the dike. This work demonstrates the flexible utility of our approach for quantifying the emplacement conditions and active lifetimes of conduits that fed LIP eruptions.