Measurement and Discrepancies of Tsunami Flow Directions Using Anisotropy of Magnetic Susceptibility (AMS) and Dip Orientations of Heavy Mineral Layers in Tsunami Deposits

Sediments preserve information about depositional processes that can be used to reconstruct flow orientation and velocity. Anisotropy of magnetic susceptibility (AMS) has been used successfully in determining the preferred orientation of magnetic minerals and is an easy and quick technique to measure sediment orientation. This technique has been used on recent tsunami deposits, which may have responded to several flows in a single event, such as incoming and return flow. However, it is unclear if the magnetic orientations measured by AMS are an accurate proxy for the flow direction. To test the AMS results, we estimated the flow directions by the dip orientation of heavy mineral layers (dHML) within known tsunami deposits in accordance with bedform theories using CT scan images and compared these directions with measured AMS orientations. For half of the tsunami deposit, flow directions from AMS and dHMLs were the same. However, for the other half of the tsunami deposit we found a 90° orientation discrepancy between dHML and AMS results, suggesting that AMS does not always accurately record the flow direction. To determine why a 90° orientation discrepancy exists in some of our samples, we employed anisotropy of anhysteretic remanence magnetization (AARM), a more sensitive measurement that is able to capture the magnetic mineral alignments of finer-grained particles. The results of the AARM measurements indicate a variety of orientations that differ from the AMS ellipsoids, suggesting that orientations change as a function of magnetic particle size. Based on these results, AMS methods alone are not always sufficient to accurately measure flow directions from tsunami deposits. However, as AMS measurement techniques improve and as they continue to be tested on other tsunami deposits, we may start to understand why these 90°-discrepancies exist and how to better interpret AMS results.