Two years ago, at the Exploration and Science Forum, we proposed that volcanism on the Big Island of Hawaii could possibly be dated by a method using thermoluminescence (TL), the light produced as the sample is heated. However, unlike the many proposals to date, we do not rely on growth on the natural TL signal (which is prone to certain difficulties), but rather we measure increases in induced TL caused by the lavas equilibrating on the thousands-of-years timescale. (“Natural” TL is the signal in the “as received” sample. “Induced” TL is the signal produced by a standard test dose of radiation.) Last year we obtained data for 24 lavas from Craters of the Moon and surrounding areas in Idaho and reported the results at the fall 2015 AGU. This year we report initial results from 15 samples collected on Hawaii. Six samples from Hawaii and several samples from Maui, Oahu, and Kauai remain to be examined.
The samples we report here were collected mostly from flows of the Kilauea and Kohala volcanoes. The Kilauea samples are modern to about 10,000 years while the Kohala samples are 140,000 to 400,000 years. The Kilauea samples had induced TL values of 42 to 503 while the Kohala samples had induced TL values of 1203-4433. (These TL values are arbitrary units for the moment; we will publish absolute values in due course. Typical uncertainties are ~20%). So, a first order observation is that the Kohala samples have considerably greater induced TL as expected for their much greater ages. However, the oldest Kohala sample has an induced TL value of 1203 and an age of 400,000 years, indicating considerable scatter. The Idaho samples showed a similar trend between induced TL and age, but again with considerable scatter. Trend lines through the Hawaii and Idaho samples appear to be very similar based on these preliminary data.
Our working hypothesis is that lavas freshly formed are highly nonequilibrium systems with considerable amorphous material and minerals containing incompatible elements. The luminescent mineral is assumed to be feldspar, the most luminescent of the common rock-forming minerals. With time, after solidification, the lava assemblages approach equilibrium through (1) crystallization, as feldspar forms from the amorphous (glassy) material, and (2) diffusion, as incompatible elements leave the feldspar. Both these processes should produce an increase in induced TL levels. The scatter on the data suggests a second process. Either the variable abundance of feldspar phenocrysts is scattering the data, or varying amounts of crystallization of feldspar during flow is scattering the data. Petrographic and thermal studies are planned to help resolve this and quantify the induced TL vs. time relationship.
Whatever is causing the scatter, it seems that the induced TL of basalts in various environments can in principle be dated from their induced TL.
We appreciate help of Darlene Lim, PI of the BASALT project, for essential logistical support, members of the FINESSE team for sample collection, and Jennifer Heldmann, PI of the FINESSE node of SSERV Institute, for financial support.