Geodynamic setting of Upper Miocene to Quaternary alkaline basalts from Harrat al 'Uwayrid (NW Saudi Arabia): Constraints from K-Ar dating, chemical and Sr-Nd-Pb isotope compositions, and petrological modeling

The volcanic field of Harrat ar Rahah-'Uwayrid (NW Saudi Arabia) consists of an older plateau basalt sequence that overlies Cambrian sandstone and younger cinder cones with smaller flows that are concentrated in the central part of this field. Petrographic and whole rock geochemical data indicate that the older plateau basalts are alkali olivine basalts, while the younger volcanic products are basanites and tephrites with some phonotephrites and tephriphonolites. K-Ar model ages on 22 bulk-rock samples were obtained for variable grain size fractions (125-250, 250-500 mu m). These dates range from 8.2 to 0.34 Ma, whereby three phases of volcanic activity during the Upper Miocene, Pliocene and Quaternary can be distinguished. Sr-Nd-Pb isotopic ratios are similar for all age groups with Sr-87/Sr-86 = 0.70307-0.70324, Nd-143/Nd-144 = 0.512912-0.512957, (206)pb/(204)pb = 19.360-19.717, (207)pb/(204)pb = 15.603-15.633 and Pb-208/Pb-204 = 39.083-39.521 (present-day ratios are indistinguishable from calculated initial ratios) suggesting that chemical differences among the lavas were probably produced by different degrees of partial melting rather than by different source compositions. Trace element ratios indicate an origin of all volcanics by small degrees of partial melting of amphibole-spinel-garnet peridotite. Geochemical and isotopic data of the Harrat ar Rahah-'Uwayrid are similar to those from the small volcanic fields of Harrat Kura (about 200 km SSE of Harrat al 'Uwayrid) and of Wadi Jizan in SW Saudi Arabia, but clearly more enriched than those of the volcanic fields that are located in the more central parts of the eastern shoulder of the Red Sea Rift. The lithospheric thickness underneath Harrat ar Rahah-'Uwayrid is estimated to about 60 km based on published results from seismic tomography. Our data support the model that the metasomatized lithospheric mantle was the primary source of the magmas, while the asthenospheric mantle or a plume-type component played an unsignificant role. The composition of the lithospheric mantle source was similar to prevalent mantle (PREMA). There is no need to postulate the participation of a HIMU component and/or enriched mantle components (EM-1, EM-2) in the source of the investigated magmas. (C) 2019 Elsevier B.V. All rights reserved.