EVIDENCE OF FLOOD BASALT ALTERATION FROM FE-OXIDES AND ITS IMPLICATIONS FOR GEOCHRONOLOGIC STUDIES

The age of a flood basalt event must be well constrained in order to determine any correlation between the eruption of the lavas and global climate shifts. However, such ages are often difficult to determine due to low potassium contents or a lack of available zircons. Recent studies of the Columbia River Basalts (CRBs) have used U-Pb TIMS zircon ages from interbedded ashes to determine that 95% of the CRBs erupted between 16.7 and 15.9 Ma. This directly contradicts decades of ⁴⁰Ar/³⁹Ar studies, which collectively show that the CRBs did not reach 95% of their final erupted volume until 15.0 Ma. The discrepancy is well outside the nominal uncertainties of the U-Pb TIMS and ⁴⁰Ar/³⁹Ar methods for rocks of the age and composition in question.

Here we use a combination of Fe-oxide geochemistry, rock magnetism, and whole-rock ⁴⁰Ar/³⁹Ar geochronology to show that the CRBs were aqueously altered during and after cooling, and that this may explain erroneous ⁴⁰Ar/³⁹Ar ages. We present results from 100+ Fe-oxide samples taken from throughout the CRBs. In some samples, presence of maghemite suggests high-temperature oxidation of basalts. In other samples, the presence of goethite and Fe-oxyhydroxides suggest long-term alteration via groundwater. In either case, alteration products can be present even in ‘fresh’ samples from the massive interiors of flows.

Overall, our results show that pervasive aqueous alteration of the CRBs is common. Such alteration may explain erroneous ⁴⁰Ar/³⁹Ar ages, as K and Ar can be mobilized in high-temperature aqueous environments. This alteration is not easily identified in the field, and is heterogeneous at the outcrop scale, making it difficult to avoid when sampling for geochronology and geochemistry.