Deep drilling into a Hawaiian volcano

Although the Hawaiian volcanoes are the most comprehensively studied volcanoes on earth, there is a fundamental limitations to how much can be learned about the long-term history of a Hawaiian volcanic system by studying surface. This limitation for Hawaii (and indeed for any oceanic hot spot volcano) is that the major volume of each volcano is inaccessible because it is below sea level. Even for those parts of oceanic volcanoes above sea level, erosion typically exposes only a few hundred meters of buried lavas (out of a total thickness of 6-20 kilometers). For example, although the Hawaiian-Emperor chain has been active for at least 70 million years, all we can generally examine for any individual volcano is that small fraction (5-10%) of its history that is now exposed subaerially. Thus, although the late stages of Hawaiian volcanoes can be studied and viewed as a time sequence, the evolution of a single volcano during its ~1 million year passage across the plume is almost entirely inaccessible. If sequences of lava flows from ocean island volcanoes spanning sufficiently long time periods could be collected, they could be uniquely valuable as probes of plume structure and related magmatic processes. Continuous core drilling through a lava sequence on the flank of an oceanic volcano is probably the only way to obtain such a stratigraphic sequence.

In recognition of the unique opportunities to characterize processes at a Hawaiian volcano by drilling, an international group of scientists have undertaken to core-drill up to 5 km into Mauna Kea volcano on the big island of Hawaii. This project, called the "Hawaiian Scientific Drilling Project" (or HSDP) is led by me, Don DePaolo (UC Berkeley) and Don Thomas (U. Hawaii). In fall 1993, we core drilled a 1,056 m “pilot hole” into the flank of Mauna Kea volcano in Hilo, Hawaii (see figure). In 1999, we drilled to a depth of 3,098 meters below sea level in a nearby location. A third phase of drilling was begun in 2005, with an ultimate goal of a depth of > 4 km. Core recovery has typically been 90% or higher, yielding a time series of fresh volcanic rocks extending back to ~600,000 years before present. Petrological, geochemical, geomagnetic, and volcanological characterization of the recovered core and downhole logging and fluid sampling have provided a unique view of the evolution and internal structure of a major oceanic volcano unavailable from surface exposures. Detailed information can be found at updated HDSP website.


The 1999/2005 bore hole is located equidistant from the Mauna Loa and Mauna Kea rift zones. This location was chosen in part to reduce the likelihood of encountering intrusive or altered rocks. After penetrating a short (~0.2 km) sequence of subaerial Mauna Loa lavas, the drill core encountered ~3.1 km of material from the flank of Mauna Kea (~0.8 km of subaerial lavas; ~2.3 km of hyaloclastites and pillow lavas; minor intrusives are also present). The stratigraphic column at the end of the 1999 phase of drilling is shown in the accompanying figure at right. This figure shows a simplified lithologic column of the core hole indicating key depths: the Mauna Loa-Mauna Kea boundary; the occurrence of alkalic lavas at the top of the Mauna Kea section; the subaerial-submarine boundary; the first occurrences of intrusives and pillows; and the division of the submarine section into zones 1-4.

Excluding post-shield subaerial lavas, which form a veneer at the top of the subaerial Mauna Kea section, all of the Mauna Kea lavas are tholeiitic, as expected for a Hawaiian volcano in its shield-building phase. What was not expected, however, was the wide range of SiO 2 contents found in the over 500 fresh glasses from the submarine part of the section. The SiO2 contents of tholeiites from a given Hawaiian volcano have often been regarded as "fingerprints" of the volcano; i.e., each volcano has tholeiites with a narrow (~1-2%) range in SiO2, varying from ~48% (Loihi) to 54% (Koolau). What is striking about the HSDP Mauna Kea tholeiites is that they cover almost the entire range in SiO2 contents previously observed in Hawaiian tholeiites (see figure). Although not necessarily apparent in this figure, the glass compositions define a strongly bimodal distribution in SiO2. This suggests that that there are two distinct tholeiitic magma types at Mauna Kea. Moreover, the two distinct magma types alternate back and forth in the section, especially in its deeper parts, indicating that both types were being produced simultaneously and that they preserved their identities en route from their mantle sources to eventual eruption. Other major- and minor-element compositions and isotopic ratios co-vary with SiO2. The differences in isotopic ratios of the low- and high-SiO2 magmatic groups indicate that their formation must have involved distinguishable mantle sources. Attempting to combine these chemical and isotopic heterogeneities into a unified model of the compositional structure of the Hawaiian plume is presently a very active area of research.

The availability of samples and data from the HSDP project and the testing of hypotheses relating to the data obtained from this project are important focuses of my research at this time. For example, we are examining experimentally hypotheses for the relationship between the high- and low-SiO2 magma types, causes of the high and variable Ni contents of olivines from these lavas, and the origin of complex zoning patterns in phenocrysts from these samples.

Another aspect of our research related to the drilling project is an effort to model the growth and subsidence of Hawaiian volcanoes as the move over the underlying mantle plume. The goal of this effort is to provide a three-dimensional context for interpreting the one-dimensional data from the drill core. The movie shown below (Quicktime format, 29MB) gives an example of one such model for the growth of the island of Hawaii. With relatively simple assumptions about the slopes of the submarine and subaerial parts of the volcano, for the temporal variation of the flux of magma as volcano moves over the hot spot, for the interactions between lavas from adjacent volcanoes, and for the subsidence of the volcanic load, the model reproduces well the topography and bathymetry of Hawaii and the distribution of surface deposits. Our goal is to feed back between the results of the drilling efforts and such models of oceanic islands.

Note that in the upper panel of the movie, N-S longitude lines (in map view) have been rotated ~55° counter-clockwise; this places the inferred trace of the plume axis beneath Hawaii on the horizontal and simplifies "bookkeeping" in several of the program's subroutines. Volcano color scheme is as follows: Mahukona, tan; Kohala, dark blue; Hualalai, yellow; Mauna Kea, gray; Mauna Loa, orange; Kilauea, dark green; Loihi, light green; light blue denotes subaerial regions that receive lava from more than one volcano in a given time step (2000 years). Purple contours are subaerial, black contours are submarine, and the white contour is subaerial outline of the island. The red 100 km diameter circle represents the outer edge of the plume in plan view. The horizontal black at y = 131 shows the location of the cross-section in the lower panel. In the lower panel, volcano colors are the same as listed above, however overlap regions are now black (light blue in the top panel), and the curved blue line denotes the trace of the submarine-subaerial transition.

The final figure compares, at the same scale, the subaerial coverage of lavas from the different volcanoes that make up the Big Island of Hawaii with that predicted by the growth model. Note that the volcano color scheme is different than that used in the movie: Kohala, yellow; Hualalai, red; Mauna Kea, brown; Mauna Loa, green; Loihi, blue.