Naturally cooled samples in different types of lava flows, as well as Peleé’s hair, were collected. Additionally, we quenched hot glowing lava samples in a bucket of water in the field. All samples likely represent the opening stage of the July eruption at the Sundhnúkur crater row in 2025. The samples were analysed by electron microprobe at the Institute of Earth Sciences, University of Iceland.
They are composed of tholeiitic basalt glass and microphenocrysts and microlites of plagioclase (pl), olivine (ol) and clinopyroxene (cpx) (Fig. 1). The silicate glass has 4.8 to 6.3 wt% MgO and a K2O/TiO2 ratio of 0.130.01 (Fig. 2). As before, the MgO content reflects the extent of cooling and crystallisation of the magma and the K2O/TiO2 ratio records changes in the magma supply from depth. The Peleé’s hair has the highest MgO content, which reflects the rapid cooling and therefore crystal-poor nature of these thin glass fragments (Fig. 1). Compared to these, the other samples have higher crystallinity (Fig. 1) and more evolved (lower MgO) glass compositions.
Nevertheless, all analysed glass compositions are slightly more evolved compared to tephra collected during the 2025 April eruption and have somewhat lower K2O/TiO2 ratios (Fig. 2). Taken together, these preliminary data indicate that there has been no significant change in the conditions of magma storage beneath Svartsengi and that the July 2025 lava was not fed from deeper level in the crust.
Figure 1: Back scattered electron images of quenched aa lava (a), Pelée’s hair (b) and naturally cooled scoria (c). Pl-plagioclase, gl-silicate glass, cpx-clinopyroxene, ol-olivine.
The sulfur (S) content in several melt inclusions was also determined and reaches up to 1340 ppm. In contrast basaltic glass in the lava matrix contains only 200 to 550 ppm S, indicating that the matrix glass was variably degassed. The difference between the melt inclusion and matrix glass compositions gives us information about the sulfur release at the eruptive vents, whereas lava flow degassing can be estimated by assuming complete release of SO2 from the matrix glass. Based on these data we calculate, that each 1 m3 of basalt released 7.2 kg of SO2, which caused the significant air pollution experienced in the middle of July
Figure 2: Variation of MgO content (a) and K2O/TiO2 ratio (b) in basaltic glass since the start of the eruptions at Sundhnúkur. Data from December 2023 to April 2024 are from Matthews et al. (2024). Blue field indicates the variability measured in samples from the Reykjanes fires that occurred in the Middle Ages (Caracciolo et al. 2023).
References:
Caracciolo, A., Bali, E., Halldórsson, S. A., Guðfinnsson, G. H., Kahl, M., Þórðardóttir, I., Pálmadóttir, G. L., & Silvestri, V. (2023). Magma plumbing architectures and timescales of magmatic processes during historical magmatism on the Reykjanes Peninsula, Iceland. Earth and Planetary Science Letters, 621. https://doi.org/10.1016/j.epsl.2023.118378
Matthews, S. W., Caracciolo, A., Bali, E., Halldórsson, S. A., Sigmarsson, O., Guðfinnsson, G. H., Pedersen, G. B. M., Robin, J. G., Marshall, E. W., Aden, A. A., Gísladóttir, B. Y., Bosq, C., Auclair, D., Merrill, H., Levillayer, N., Low, N., Rúnarsdóttir, R. H., Johnson, S. M., Steinþórsson, S., & Drouin, V. (2024). A dynamic mid-crustal magma domain revealed by the 2023 to 2024 Sundhnúksgígar eruptions in Iceland. Science, 386(6719), 309-314.