While measurements >5000 Bq/kg account for only 0.0012% of the measurements made in this work, the possibility of high concentration particulate matter being present in these areas must be considered. It is clear that extensive sampling of the identified regions is necessary to determine the cause of these anomalies. While it has been demonstrated that the well documented affinity of 137Cs to fine-grained sediments determines the overall distribution of 137Cs on the seafloor (Otosaka and Kobayashi, 2013), it has also been pointed out that sediment mineralogy alone cannot completely account for the spatial distributions observed along the east coast of Japan (Kusakabe et al., 2013).

With regard to this point, the influence of the original distribution of 137Cs in the water column has been identified as a potential cause by Oikawa et al., (2013), who describe a scenario for rapid Roscovitine concentration downward migration of 137Cs in the water column. While the majority of 137Cs in the water column is known to be in the form of dissolved ions (Stanners and Aston, 1981, Nies et al., 1991, Knapinska-Skiba et al., 1994, Lujaniene et al., 2004 and Lujaniene et al., AG-014699 mw 2010), it has been shown that once 137Cs is incorporated into particulate matter, it is rapidly removed from seawater to the

bottom sediment with reported settling velocities ranging from 29 to 190 m day−1 (Fowler et al., 1987 and Kusakabe Sulfite dehydrogenase et al., 1988). The unusual sedimentary environment resulting from suspended load carried back from land by the tsunami may also have contributed to rapid removal of nuclides from seawater (Kusakabe et al., 2013). Fig. 6 shows a conceptual model for the mechanisms thought to be responsible for the observed relation between features of the terrain and the high level 137Cs anomalies recorded in this work. Fig. 6A and C show two snapshots of the situation shortly after contamination, where the underwater currents

flow in opposite directions normal to a vertical feature of the terrain. Field observations of currents at a depth of 20 m, 30 km off F1NPP by Miyazawa et al. (2012) indicate that the mean currents in the region are relatively weak with velocities typically less than 0.4 m/s. Diurnal cycling of the currents along the north–south direction occurs due to wind effects, and simulations performed in their study demonstrate that tidal currents and river discharge flows also have a moderate impact on the transport of dissolved 137Cs. Since rapidly sinking particles are thought to be responsible for transport of 137Cs from the water column to the sediments (Fowler et al., 1987, Kusakabe et al., 1988, Oikawa et al., 2013 and Kusakabe et al., 2013), it is reasonable to assume that the horizontal distribution of sinking 137Cs particles in the water column would be relatively homogeneous over the scale of a few 100 m at any moment in time.