The Hawaiian mantle plume forms the longest oceanic island chain on Earth, running approximately 6,000 km. It represents the typical inner workings of intraplate volcanism and the Hawaiian mantle plume has the largest buoyancy flux of any other plume in the world (e.g. Sleep, 1990; Ribe & Christensen, 1999; DePaolo et al., 2001).
For the past 10 years, Dominique has studied various Hawaiian volcanoes using four isotopic systems (Sr, Nd, Hf, Pb) and elemental geochemistry to gain deeper insight on the Hawaiian plume (e.g. Hanano et al., 2010). Initially, she worked on the Hawaiian Scientific Drilling Project (HSDP, Blichert-Toft et al., 2003; Nobre Silva et al., 2013) and then focused on Mauna Loa (the largest volcano on Earth) among other key volcanoes. Approximately 120 shield lavas of Mauna Loa were analyzed, corresponding to over 500,000 years of volcanic activity. The high-precision data that was generated confirm the clear geochemical difference between the Loa and Kea volcano geographical trends and also clearly document the Loa source to be both more enriched and more heterogeneous than the Kea source. The EM-I component in the Loa source has been connected to the presence of the Pacific LLSVP (large low shear velocity province) at the core-mantle boundary on the Loa side of the volcano trends (Weis et al., 2011), while the Kea component is compositionally similar to “C” and is interpreted as being a common, widespread composition within the Pacific deep mantle (Nobre Silva et al., 2013).
Isotopic data for Hawaiian shield lavas. a, Plot of 208Pb/204Pb against 206Pb/204Pb for all Hawaiian shield lavas (normalized to same standard values). Mauna Loa and Mauna Kea samples have larger symbols. Thick black line indicates boundary separating volcanoes on the Mauna Loa (diamonds and cool colours) and Mauna Kea (circles and warm colours) trends based on Pb isotopic composition (Abouchami et al., 2005). Insets: Pb–Pb arrays among some of the individual volcanoes. Solid lines indicate Mauna Loa (HSDP-I and Mile High Section, MHS) and Mauna Kea (low-, mid- and high-8, Eisele et al., 2003). Dashed lines indicate other Hawaiian volcanoes or features (Li, Lō‘ihi; H, Hualālai; L, Lāna‘i; K, Kaho‘olawe; Ha, Hana Ridge; Ki, Kīlauea; Ko, Kohala; M, Maui; EM, East Moloka‘i; WM, West Moloka‘i). b, Plot of 208Pb*/206Pb* against εNd for Hawaiian shield lavas. Inset: plot of 208Pb*/206Pb* against εNd detail of Mauna Loa lavas (this study, plus prehistoric lavas <3 kyr old, radial vents (Marske et al., 2007; Wanless et al., 2006), where older samples (red, yellow and orange diamonds; see also Supplementary Fig. S1 in Weis et al., 2011) sampled along the submarine southwest rift zone define a unique mixing trend. Darker outline indicates triple-spike literature data. See Supplementary Information in Weis et al. (2011) for all data sources.
Temporal evolution of isotopic compositions from Hawaiian volcanoes. a, Plot of εNd against distance from Kīlauea in kilometres along the Hawaiian chain. b, Plot of 208Pb*/206Pb* against distance from Kīlauea. Symbols, colour coding and data source (plus Wai‘anae) as in the figure above. See Supplementary Information for data sources in Weis et al. (2011). Individual volcanoes are represented by the average of the analyses of shield lavas (thermal ionization triple- spike or multicollector inductively coupled plasma mass spectrometry data), with the two standard deviation on the mean used to indicate the variation around the average. Shaded fields indicate the total range of variation covered for each volcano.
Deep mantle velocity anomalies and hotspot locations. a, Global map showing the lowermost-mantle Vs perturbations from tomographic model TXBW (Grand, 2002, figure adapted from Garnero, 2000). Continents, thin black outline; Pacific ‘Ring of Fire’, thick blue line; circles, hotspot locations (Steinberger, 2000). Red and blue colours indicate lower and higher velocities than global average, respectively. The peak-to-peak value for this model is 5.5. Hawai’i and Pitcairn (Pacific Ocean), Kerguelen (Indian Ocean) and Tristan (Atlantic Ocean) have EM-I-type geochemical signatures. b, Histogram of 208Pb/206Pb (used to minimize analytical noise on 204Pb) for Hawaiian shield lavas. c, Cross-section through the Hawaiian mantle plume down to the CMB, showing a schematic representation of the low-velocity structures beneath the central Pacific (after Bréger et al., 1998, modified after Garnero, 2000). Hawaiian Island surface locations and relative position of the Loa and Kea trends in the mantle plume are indicated.
This study is being extended back in time, along the Northwestern Hawaiian Ridge, up to the bend at ~41 Ma, 3,400 km of the Hawaiian Chain, to establish the source characteristics (components involved and depth) of the Hawaiian plume (see Garcia et al., 2015 for a review of the sample settings).
Map of the Northwest Hawaiian Ridge based on the SRTM30_PLUS (Shuttle Radar Topography Mission) bathymetry data set (Becker et al., 2009). Red triangles show locations where shield stage lavas (tholeiites) have been recovered; yellow triangles show where post-shield stage transitional and alkalic lavas have sampled; white triangles show where rejuvenation-stage rocks were obtained. Note that many Northwest Hawaiian Ridge volcanoes have no geochemical data. The location of the Kīlauea reference point at the center of its summit caldera is shown by the K at 155.273840°W, 19.413017°N. The data were gridded at 1 km spacing (~30 arc s). A Mercator projection was used for map. Inset: Map of the north-central Pacific Basin showing the Hawaiian-Emperor Chain, Mid- Pacific Mountains and Aleutian Trench. A transverse Mercator projection was used (white box shows location of main map). From Garcia et al., 2015.
This information is also used to model the variations in space and time (create a geochemical-isotopic map), and to determine when the Loa component appeared and when the Loa and Kea plume trends became independent.
Dominique has also been working on the “What, When, Where and Why of Secondary Volcanism?” She has participated in a series of JASON2 ROV dives on the Northern Hawaiian Islands (lead by M.O. Garcia and G. Ito, University of Hawai’i at Manoa), sampling the various phases (pre-shield, shield, post-shield and secondary, i.e. a record of over 5 myr) of volcanic activity of a Hawaiian island (Garcia et al., 2008).
Bathymetric map of the seafloor around the islands of Kaua`i, Ni`ihau, and Ka`ula, based on new results from the Kilo Moana expedition. Inset photos: (a) R/V Kilo Moana. (b) JASON2. (c) Members of the science team.(d) Slab of hollow pillow lava. (e) JASON2 collecting pillow lava with white sponge. (f) South Kaua`i swell columnar jointed outcrop. (g) Mechanical arm of JASON2 collecting pillow lava. (h) Slab of pillow lava lobe shown in Figure 1g. (i) Ka`ula pillow lavas. (j) Outcrop of pillow lavas on ocean floor. (k) Slabbed vesicular pillow lava. (l) Tripod fish. Figures 1d–1f are slabbed rocks or outcrops being sampled by JASON2 from the east Kaua`i area, Figures 1g–1i are from Ka`ula, and Figures 1j and 1k are from the Middle Bank. From Garcia et al., 2008.
Specifically, the studies involving these samples investigate (i) the relationship between small-scale heterogeneity at the volcano scale and large-scale zoning in the plume source and (ii) the origin of the post-shield and then secondary volcanism and the reasons why they carry more geochemical anomalies than shield-stage lavas.
Over the last year, two new projects have started: the sampling and thorough geochemical characterization of 1) the Makapu’u enriched end-member (Ko’olau) on O’ahu and 2) the shield-stage lavas on Kaua’i.