In October 2020, we collected sediment samples within the area of the coastal forest inundated by the 2011 tsunami along two transects roughly perpendicular to the shoreline (Figs. 1c–e, 2). Sediment samples were taken directly from the walls of excavated pits using a 18-cm-long and 10-cm-wide plastic box (Fig. 2b, c) at seven locations along Transect A (A1–A7) and five along Transect B (B1–B5) (Fig. 1d, e). The sediment samples were visually described and photographed. We also collected a reference sediment sample (R, Fig. 1c) beyond the known inundation limit in June 2021 using the same sampling approach. In addition, pine bark, needles, and cones on the ground surface were collected for geochemical reference data. Topography was measured using a Leica Viva TS15 imaging total station (Leica Geosystems, St. Gallen, Switzerland) and a Leica Viva GS10 high-precision GNSS receiver.
In the laboratory, sediment samples were scanned using a computed tomography (CT) scanner (Supria Grande, Hitachi) in the Geological Survey of Japan laboratory (National Institute of Advanced Industrial Science and Technology). They were then subsampled in 1-cm-thick vertical intervals. The subsamples were dried at 60°C for 24 h, and their water contents (%) were calculated from the sample weights before and after drying. A portion of the dried samples was then homogenized with an agate mortar and pestle, weighed, and ashed at 550°C for 4 h to estimate loss on ignition (LOI550, %), which is indicative of organic content (Dean 1974; Santisteban et al. 2004).
Before grain-size analysis, organics and carbonates were respectively removed from the remaining dried samples by dissolution in hydrogen peroxide (H2O2) and hydrochloric acid (HCl). The samples were then sieved at 63 µm (4 phi) to remove the mud fraction, and the sand content (%) was calculated from the sample weights before and after sieving. Grain-size analysis was performed on the residual sand samples using an image analyzer (Camsizer, Retsch Technology GmbH, Haan, Germany) with an effective measuring range of − 5.25 phi (pebble) to 6.25 phi (silt) at 0.25-phi intervals. The mean grain size of each sample was then calculated from the measurements using the logarithmic graphical method of Folk and Ward (1957).
We kept about 1.9–2.4 g of each subsample from location A6 for radioactive cesium measurements using a gamma spectrometer at the Geological Survey of Japan (GCW2022, Canberra Industries Inc., USA). The 137Cs (t1/2 = 30.2 years) data were used to detect horizons of specific ages (e.g., 1950–1964 CE; He and Walling 1997; Wren and Davidson 2011).
Biomarkers were measured in samples A1, A4, A6, A7, and the reference sample, as well as in the pine materials taken from the ground surface. Because large plant fragments constitute a large portion of the organic matter present in these sediments (Shinozaki 2021), terrestrial plants were mostly removed from each subsample using a 250-µm sieve, and fraction finer than 250-µm in each subsample was then dried and homogenized. Lipids were extracted from the homogenized subsamples with dichloromethane and methanol using an accelerated solvent extractor (ASE350, Dionex, Sunnyvale, USA). Hydrocarbons (N1 fraction; locations A1, A2, A4, A6, R, and pine materials), ketones (N3; locations A4 and A6 only), and alcohols and sterols (N4; locations A4 and A6 only) were extracted as described by Shinozaki et al. (2015). These organic fractions were measured using a gas chromatograph with flame-ionization detection (Agilent 7890B or Agilent 6890N, Agilent Technologies Inc., USA) and an Agilent DB-5HT nonpolar column (30 m long, 0.25 mm internal diameter). Individual compounds were identified using an Agilent 5973 network mass selective detector gas chromatograph/mass spectrometer (GC-MS).