Validation of accumulation rates in Teluk Banten (Indonesia) from commonly applied 210Pb models, using the 1883 Krakatau tephra as time marker

Abstract

Teluk Banten is a tropical shallow marine embayment on the north coast of Java, close to the Krakatau volcanic complex. The tsunami generated by the 1883 Krakatau eruption resulted in the deposition of a tsunamite tephra layer on the bottom of Teluk Banten. The Krakatau Event Layer has been used as a time marker to validate 210Pb geochronological dating models. This validation was performed to answer the question: Why should we believe excess 210Pb derived mass accumulation rates? The commonly used models have therefore been applied to 210Pb activity profiles from 25 gravity cores. Models that have been validated are the Constant Flux and Constant Sedimentation (CF/CS) (with and without a surface mixed layer), the Constant Initial Concentration (CIC), the Constant Rate of Supply (CRS), and the Flux Balance model. After comparison of the mass accumulation rates (MARs) of the different models with the MARs obtained from the Krakatau-layer, the following could be concluded. Firstly, activity profiles without a surface mixed layer gave CF/CS–MARs which were on average a factor 1.4 ± 0.3 ( n = 8) too high. Including a mixing-rate of 0.5 ± 0.4 cm 2 yr − 1 into the model could explain this overestimation. Secondly, for 210Pb activity profiles with a surface mixed layer, the MARs were in good comparison (1.0 ± 0.3, n = 12) with the MARs derived from the Krakatau Event Layer. Thirdly, increased accumulation instead of reworking had to be the main cause of the comparatively thick upper layer of five cores with very high inventories. Fourthly, assuming a variable accumulation rate and applying the CIC model and CRS model for these five cores, the weighted average CIC–MARs were in general more accurate than the CRS–MAR. Finally, the Flux Balance model, which is not often used, gave MARs that were a factor 1.3 ± 0.2 ( n = 25) too high. The advantage of this model is that it is independent of the profile shape. The disadvantage is that a reliable and accurate initial activity over longer time scales has to be obtained.