The Origin of Younger Volcanism in Western Saudi Arabia: Insights from Geochronology, Isotope Geochemistry, and Mineral Chemistry

Abstract

The origin of many intraplate volcanic fields and the relationship between intraplate volcanism and mantle flow dynamics are still poorly constrained. The long history of intraplate volcanism fields in western Saudi Arabia appears broadly related to the arrival and continued activity of the Afar plume and the opening of the Red Sea, other processes – such as the possible impact of “fossil” plumes and lithospheric extension have also been argued to be the cause of volcanism. Post-12 Ma volcanic fields, also known as “harrats”, extend across western Saudi Arabia and have generated alkali basalt magmas that are compositionally different from those associated with older tholeiitic regional volcanism (> 20 Ma). The largest volume of these younger harrats is concentrated along an N-S structural trend called the Makkah-Madinah-Nufud-line (MMN-line). In addition, other harrats – typically associated with lower volume eruptions -- occur peripherally to the MMN line. Here, we conduct a regional study by comparing three harrats that occur along the MMN line, Rahat, Khaybar, and Ithnayn, with three that occur peripherally to the MMN line: Kura, Lunayyir, and Hutaymah. Our aim is to achieve a more complete understanding of the origin of younger Harrats volcanism by examining spatial-temporal variations in mantle source compositions and contributions to magmatism through various whole-rock and mineral geochemical analyses. We also test the hypothesis that volcanism along the MMN line has migrated northward over time. In the first study, we present new 40Ar – 39Ar age determinations for 23 lava flows from Harrats Ithnayn, Kura, and the northern part of Harrat Khaybar. Our new 40Ar – 39Ar data indicate an age range of 8.39 – 5.90 Ma for Harrat Kura and 467 – 33.4 Ka for Harrat Ithnayn. These results suggest that both Kura and Ithnayn lavas are much younger and formed over a shorter period than previously reported from K-Ar ages. We also observe a slightly older age (2.3 Ma) for the northern part of Harrat Khaybar. Our data confirm that Harrat Ithnayn is the most recently formed lava field along the MMN volcanic line system, consistent with a south-to-north volcanic age progression (Rahat > 10 Ma, Khaybar > 2.3 Ma, and Ithnayn > 0.46 Ma) of ~ 59 mm/yr. Volcanism along the MMN line cannot be explained by Plate motion over a fixed sub-lithospheric hotspot due to discrepancies between the rate and direction of observed Plate motion compared to the resulting MMN line trend. Instead, progressive volcanism is consistent with northward sub-lithospheric channeling of the Afar mantle plume flow beneath the Arabian Plate. The second study investigates variations in mantle source compositions and contributions using major and trace elements and Sr-Nd-Pb-He isotopic measurements. Modeling of mantle melting suggests that the MMN-line harrats are derived from a relatively higher degree of melting (> 3%), higher mantle temperatures (TP > 1420 C), and shallower melting regions, while Kura and Hutaymah magmas are mostly derived from relatively lower degree of melting (< 3%), lower mantle temperature (TP < 1420 C), and deeper melting regions from within the garnet peridotite stability field. Harrat Lunayyir is unique in that it shows a relatively lower TP but a higher degree of melting, shallower melting regions (and thinner lithosphere thickness), and is similar to MMN-line harrats. These results are also broadly consistent with recent seismic constraints on the depth of the lithosphere-asthenosphere boundary in this area. Helium measurements show a narrow range of 3He/4He values for the selected harrats (6.3 – 8.3 RA, where RA is the atmospheric ratio), except for Harrat Kura (~ 6 Ma) and Northern Rahat (< 1.4 Ma), where we see elevated 3He/4He values up to 12.7 RA and 10.8 RA, respectively. Harrats that are located on the flank of the MMN-line (e.g., Harrat Kura and Hutaymah) also show the most radiogenic enriched mantle source signature of Pb, Nd, and Sr compositions (e.g., 206Pb/204Pb up to 19.53, and εNd 1.9 – 4.5) relative to the selected MMN-line harrats (e.g., 206Pb/204Pb < ~ 19.0 and εNd > ~ 4.5). The role of crustal contamination in the chemical variation is minimal, except for Harrat Ithnayn, where several lavas have lower Nb/U (<30) and elevated 87Sr/86Sr ratios. Although radiogenic isotope compositions broadly define a two-component array, a comparison of trace element and isotopic compositions shows that the peripheral and MMN-line harrats sampled two different enriched endmembers: the peripheral harrats sampled the enriched metasomatized SCLM, whereas the MMN-line harrats sampled an enriched asthenospheric mantle component, more likely an Afar plume-like component. Asthenospheric isotopic signatures are apparent in harrats lavas, which erupted when the magma flux increased. Overall, we suggest that a mixture of multiple mantle sources contributed to the younger magmatism, including the Afar mantle plume (through northward asthenospheric flow), with some involvement of the SCLM. Finally, the third study utilizes olivine and spinel inclusion compositions to provide insights into the mantle heterogeneity and geodynamics beneath western Saudi Arabia. The variations in olivine minor element concentrations and the first-row transition elements provide insights into the mantle sources lithologies beneath the younger harrats. Our olivine data indicate that the selected harrats predominantly derive from olivine-rich source assemblages rather than pyroxenitic sources. However, the influence of lithospheric thickness on the minor element concentrations further emphasizes the role of the "lid" effect suggested by whole-rock trace element modeling. Also, variations of olivine first-row transition elements show an indication of changing the oxidation state due to the Fe-Ti oxide fractionation or the result of hydrous melting. This confirms the involvement of the SCLM sources, especially in the harrats located in the periphery of the MMN-line axis.