Geochemical Variability at Krafla, Iceland and the Great Rift, Usa Reveals Multiple Magmas in Basaltic Fissure Eruptions

51st Lunar and Planetary Science Conference (2020) 2698.pdf

GEOCHEMICAL VARIABILITY AT KRAFLA, ICELAND AND THE GREAT RIFT, USA REVEALS MULTIPLE MAGMAS IN BASALTIC FISSURE ERUPTIONS. S.S. Hughes1, E.H. Christiansen2, R. Huang2, W.B. Garry3, A. Sehlke4, and J.L. Heldmann4. 1Dept. of Geosciences, Stop 8072, Idaho State University, Pocatello, ID, 83209 ([email protected]); 2Brigham Young University, Provo, UT, 84602; 3Planetary Geology, Geophysics, and Geochemistry Laboratory, Code 698, NASA Goddard Space Flight Center, Greenbelt, MD, 20771; 4NASA Ames Research Center, Moffett Field, CA 94035.

Introduction: Chemical analyses of basaltic lavas series of phreatic explosions produced a central pit derived from fissure eruptions on Earth [e.g., 1–7] in- crater (Fig. 1, right) and numerous smaller blow-out dicate that significant geochemical diversity apparently pits along the fissure system [7,8]. occurs in many small monogenetic eruptions. The diversity may be related to relatively simple assimilation / fractional crystallization (AFC) processes or more complex scenarios involving multiple magma batches, with or without concomitant mixing. In order to better understand possible relations between chemical diversity and the possibility that some basaltic fissures erupt from more than one magma reservoir, we examine the geochemical variations in basaltic lavas within two fissure systems on Earth: (1) Krafla, Iceland (Fig. 1 left), the 1975-1984 (Krafla Fires) eruptions along a compound fissure, and (2) the closely related (~2.2 ka) eruptions of Kings Bowl (Fig. 2 right) and Wapi (south of KB) lava fields on the southern part of the Great Rift in Idaho, USA. Each of these fissure systems be- longs to a much greater volcanic province; Krafla lies within the Icelandic Oceanic Rift Zone, and the Great Rift is on the eastern Snake River Plain, USA and both are influenced by rifting and hot spot dynamics. Samples collected during 2014 – 2019 field seasons included accessible parts of each lava field (60 from the Great Rift lavas, mostly the Kings Bowl fissure system; and 41 from Krafla lavas and spatter ramparts) to obtain representative analyses of proximal and distal Fig. 1. Basaltic lava flows and eruptive fissures (in flow surfaces, pyroclastics, and older lavas known to red) of the Krafla Fires (left) and Kings Bowl (right) have erupted in the viscinity (e.g., pre-Kings Bowl and systems. Sample locations shown as yellow-filled stars. Base maps are grayscale Google Earth images. nearby Inferno Chasm lava; older Krafla lava flows). All samples were analyzed by XRF for major and trace Krafla Fires: Observations confirm much of the elements. Additional data, where available, were ob- 1975-1984 eruption sequence occurring in 9 events tained from literature sources for comparison. along fissures in the Krafla Caldera [9,10]. Approxi- Wapi and Kings Bowl: Wapi lava field, a basaltic mately 0.25 km3 is estimated to have erupted [9], low shield dated at 2,270 ± 50 B.P., erupted ~5.5 km3 which produced a lava field ~16 km long that partially of lava covering 325 km2 area [4]. Kings Bowl, which covered a 90 km long fissure zone extending north lies on the same volcanic rift, represents an “aborted” (through the caldera) from the Leirhnjúkur fumarole eruption dated at 2,220 ± 100 B.P. that, occurring in area (Fig. 1, left). The initial three eruptions (Dec 1975 multiple episodes, produced no more than ~0.0125 km3 – Sep 1977) from fissures ~1.5 – 3 km north of the of lava covering ~3.3 km2 area [4,7,8]. Although the fumarole area began near the south end, each one last- fissure system of Wapi is obscured by lava, the Kings ing only hours or a few days. The last six eruptions Bowl eruption occurred along a 7 km series of ~12 en (Mar 1980 – Sep 1984) extended the eruptive activity echelon eruptive fissures, with concomitant extension along several fissure segments northward nearly 11 cracks on either side of the fissure system [8]. Lava km. that erupted from the fissure length partially filled in a topographic depression, creating a small lava lake, as well as proximal spatter ramparts and small cones. A 51st Lunar and Planetary Science Conference (2020) 2698.pdf

A more significant issue is the comparison of Kings Bowl geochemistry to neighboring lava fields (Fig. 2). Separate geochemical trends possibly derived from different magma bodies are apparent for older pre-Kings Bowl, Wapi and Inferno Chasm lava flows, illustrating the complexities along the southern Great Rift. A few “outliers” from the clusters and trends may represent mixed magma, or some may be older non- juvenile lava blocks ejected during Kings Bowl phreatic explosions. Trace element co-variations tentatively illustrate similarly complex relations. The geochemistry of Krafla Fires lavas also reveal a tight major element trend (Fig. 2) that is overall less mafic than Kings Bowl lavas. The data confirm a strong bimodality recognized previously [9], which separates a rather primitive (mafic) series from a more chemically evolved series. Ranges of key major oxides are MgO (5.9–8.0), TiO2 (1.4–2.1), FeO (11.5– 15.2), CaO (10.2–12.4) and P2O5 (0.15–0.22) wt. %. The chemically evolved tight cluster (lower Mg; higher Ti, K and Fe) represents samples from the southern- most fissure near the fumarole hill (Fig. 1), whereas the mafic series erupted ~9-11 km north from two fissures that produced much of the lava that flowed north. We suggest that the separate geochemical clusters in both systems reflect multiple magma reservoirs, as proposed for Kings Bowl [7] and the Krafla Fires [9,10,12]. Monogenetic fissure eruptions on the Moon, Mars or other planetary bodies may have similar mag- matic complexities. Petrologic models are currently being investigated to determine possible scenarios for Fig. 2. Major element co-variations in lavas in the the variations within each eruptive episode. Great Rift and Krafla fissure systems. Separate geo- Acknowledgements: This research is supported by chemical trends, especially Ti vs. Mg, are likely re- the FINESSE project (Field Investigations to Enable lated to fractional crystallization (Ol, Pl, Cpx), with Solar System Science and Exploration), PI J.L. possible magma mixing or other complex processes Heldmann, a SSERVI research grant to NASA Ames exemplified by K vs. Fe. Research Center. Geochemistry: Geochemical analyses of Kings Bowl References: [1] Thornber C.R., 2003, USGS Prof. [7] reveal a moderate compositional range of olivine Pap. 1676 Ch.7. [2] Eason D.E. and Sinton J.M., 2009, tholeiites. Ranges (in wt. %) of major oxides are MgO JVGR 186, 331-348. [3] Volynets A.O. et al., 2015, (8.5–10.2), TiO2 (2.3–2.6), total Fe as FeO (12.3– JVGR 307, 120-132. [4] Kuntz M.A. et al. 1992, GSA 13.4), CaO (9.7–10.8) and P2O5 (0.52–0.61). These are Mem. 179, 227-267. [5] Hughes S.S. et al. 2002, GSA unique for a small eruptive volume, yet not unusual for Spec. Pap. 353, 151-173. [6] Miller M.L. & Hughes the Great Rift [4,11]. Some trace element variations (in S.S. 2009, JVGR 188, 153-161. [7] Hughes S.S. et al., ppm) are also notable, with ~20 percent or more varia- 2018, JVGR 351, 89-104. [8] Greeley R. et al., 1977, tion (e.g., Ba = 253–304, Cr = 336–400, and Ni = 119– NASA Contract. Rep., CR-154621, 171–188. [9] Grön- 166 ppm), while other trace elements are more uniform vold K., 2006, Eos Trans. AGU, 87(52), Abstract (e.g., Cu = 45–51, V = 282–314, Sr = 244–263, Zn = T33E-08. [10] GVP, 1989, McClelland L., (ed.), Sci. 101–117, and Zr = 196–220 ppm). Broad trends in Event Alert Network Bull. 14, 5. [11] Kuntz M.A. et geochemical variability (Fig. 2) are not spatially relat- al. 2007, USGS Sci. Invest. Map 2969. [12] Gud- ed, which suggests that either the fissure system tapped mundsson A., 1995, JVGR 64, 1-22. a single incompletely mixed magma source or that variable mixing occurred along the entire fissure.