Magma residence and transfer rate from U-series disequilibria: the case of Hekla volcano, Iceland

Rate of magma transfer and differentiation processes can be estimated from radioactive disequilibria between radionuclides in the ²³⁸U-chain if the mechanism that fractionate the nuclides can be constrained. Once constrained, the variable half-lives of the radionuclides allow to restrict the time elapsed from the fractionation. Over the last centuries, eruptions from Mt. Hekla have terminated with emission of phenocryst-poor Fe-rich basaltic andesite (or icelandite) of uniform composition, produced by approximately 50% fractional crystallisation from basalt. The crystallising mineral assemblage is composed of 8-13% olivine, 34-40% clinopyroxene, 32-41% plagioclase (An_50-65) og 15-17% Fe-Ti oxides (Sigmarsson et al., 1992; Chekol et al., 2011).

Both ²³⁸U-²³⁰Th disequilibria and Th isotope ratios are identical in the basaltic andesite and basalt erupted around the volcano consistent with magma differentiation by fractional crystallisation. Modest radioactive disequilibrium is observed between thorium and radium that decreases from the basalt to the basaltic andesite with (²²⁶Ra/²³⁰Th) from 1.16 to 1.01. Such a decrease may represent decay of ²²⁶Ra (T_1/2: 1600 yrs) if the fractional crystallisation took longer than 200 years, which is the minimum time for measurable effect of ²²⁶Ra disintegration. On the other hand, if Ra enters the extracted mineral phases, the time of fractionation would be much shorter.

Radium is an incompatible element which concentration increases with magma differentiation. Its bulk partition coefficient (D_Ra; min-melt)) between the minerals fractionating and the derived melt can be estimated from the variations of Ra and Th concentrations as 0.12 ± 0.03. Since Ra only enters plagioclase of the fractionating mineral assemblage, the plagioclase D_Ra (plag-melt) is 0.12/0.4 = 0.3. Melting experiments result in partiioning of Ra between plagioclase and melt close to 0.3 for An₅₀ plagioclase (Fabbrizio et al., 2009), consistent with measured Ra and Th variations in Hekla lavas. Consequently, the lower (²²⁶Ra/²³⁰Th) in the basaltic andesite is fully explained by plagioclase fractionations on a timescale significantly shorter than 200 years.

Radium disintegrates to ²²²Rn, which in turn decays with a half-life of only 3.8 days to ²¹⁰Pb. The volatile behaviour of the inert gas Radon is thus the main fractionation process generating ²²⁶Ra-²¹⁰Pb disequlibrium, whereas the D (min-melt) for Pb and Ra is comparable. Hence, outgassing or accumulation of gas containing Radon is an effective fractionation mechanism on a short timescale causing Ra-Pb disequilibria that can be studied for approximately a century. Indeed, the first emitted tephra of the last Hekla eruptions display excess ²¹⁰Pb over ²²⁶Ra suggesting volatile accumulation in a hermetic magma chamber explaining the initial explosive character of Hekla eruptions (Garance and Sigmarsson, 2023). Taken together, U-series disequilibria suggests that magma residence and transfer to surface occurs on a decadal timescale beneath Hekla volcano with important role for gas accumulation in the basaltic andesite affecting the repose time between eruptions.

 

References

Chekol et al. 2011

Fabbrizio et al. 2009

Hervé and Sigmarsson 2023

Sigmarsson et al. 1992