Spatial variability of vertical profiles of radiocesium in a paddy field measured by conventional core sampling

Our analyses revealed the spatial variability of the vertical distribution of fallout radiocesium measured in small cylindrical soil cores taken in a non-disturbed paddy field contaminated by the Fukushima Dai-ichi Nuclear Power Plant (FDNPP) accident in 2011. The field is in Iitate Village 40km northwest of FDNPP and had been designated as an off-limit area. In December 2012, soil samples were taken along a longitudinal and transverse line in the field at intervals of 1 or 2 m. In May 2013, 36 additional soil samples were taken from a 1m2 plot at 20cm intervals, where the ground undulated considerably. The core samples were carefully cut into slices, dried, and subjected to measurement of the radiocesium concentration. The coefficient of variation in the total radiocesium inventory (i.e. the radioactivity of total radiocesium in the unit area of the soil profile) and the averaged radiocesium concentration at 0~10 cm were 47.8 % and 53.5 %, respectively, demonstrating that the mean of a few representative points in a paddy lot provides huge uncertainty. Our analysis of semivariograms for the mean radiocesium concentration revealed no spatial correlation even between the shortest lag distance (2 m). It should thus be noted that the mean concentration determined from a small core sample cannot represent the tendency around the sampled point. However, the result of the 20 cm2 grid samplings suggested that undulations on a 10-cm scale can cause radiocesium to migrate to a concave spot. Among the 82 sampling points, 50 % of the vertical profiles of the radiocesium concentration had a concentration peak at the ground surface, but the other points had an s-shape or bell shape, suggesting variation in the water flow regime among the sampled profiles. The apparent convective velocity of the radiocesium ranged widely from 0.00 to 3.38 cmy-1 with a mean of 0.75 cm y-1.