Keynote: Do erupted mafic lavas accurately reflect mantle magmatic timescales?

Goldschmidt Conference Abstracts Mineralogical Magazine www.minersoc.org 1763 Do erupted mafic lavas accurately reflect mantle magmatic timescales? K.H. RUBIN Dept. of Geology & Geophysics, Univ. of Hawaii, Honolulu , HI 96822 USA ( [email protected]) Mantle melting and melt transport encompass a range of processes that vary in duration, magnitude and length scale, and depend upon variables such as tectonic setting, volcano maturity, volcano spacing, and mantle physical characteristics (e.g. thermal state, composition, lithology, upwelling rate). Multiple compositional attributes of mantle derived magmas provide information on melting processes and conditions. The U-series nuclides are nearly unique in their ability to provide information about rates of such processes because chemical fractionation creates radioactive disequilibria between decay chain nuclides, which then decay away on characteristic timescales once fractionation ceases. Some surprising results from this literature suggest that the duration of melting is highly variable over short temporal and spatial scales in the mantle, and that melt transport is generally quite fast. But how trustworthy are these interpretations? Early studies were content to use the presence or absence of radioactive disequilibria in magmas to crudely estimate the duration of magma generation and transport in the mantle. Yet since the 1970s, development of much more sophisticated melting models and of analytical methods for nuclide activity measurements with 100-fold precision improvements have led to more specified time scale assessments in the U-series literature. Despite these refinements, the geological errors introduced by the melting and melt transport processes in the mantle place severe and often underestimated limits on our ability to tell time with these tracers in erupted magmas. Situations such as multiple melting lithologies, magma mixing, fluid-addition (at convergent margins) and reactive transport create a large amount of non-uniqueness to the way U-series data sets can be interpreted. Add crustal/magma chamber processes to this list and the challenge of deriving a unique temporal solution to the duration of magmagenesis and transport using a combined U-series, major/trace element and radiogenic isotope data set is nearly insurmountable. Nevertheless, results from U-series studies have much to tell us about magmatic timescales so long as data are not overinterpreted or modeled with a false sense of precision to the parameterization. This presentation will discuss the current state of the art and successful strategies for limiting the uncertainty of timescale assessments from magma chemistry, and present examples where results are generally consistent with independent geological and geophysical observations. Rubin K (2011) Mineralogical Magazine, 75(3) 1763