To investigate the main factors that control the dissolved radiocesium concentration in river water in the area affected by the Fukushima Daiichi nuclear power plant accident, the correlations between the dissolved 137Cs concentrations at 66 sites normalized to the average 137Cs inventories for the watersheds with the land use, soil components, topography, and water quality factors were assessed. We found that the topographic wetness index is significantly and positively correlated with the normalized dissolved 137Cs concentration. Similar positive correlations have been found for European rivers because wetland areas with boggy organic soils that weakly retain 137Cs are mainly found on plains. However, for small Japanese river watersheds, the building area ratio in the watershed strongly affected the dissolved 137Cs concentration. One reason for this would be because the high concentrations of solutes, such as K+ and dissolved organic carbon, discharged in urban areas would inhibit 137Cs absorption to soil particles. A multiple regression equation was constructed to predict the normalized dissolved 137Cs concentration with the topography, land use, soil component, and water quality data as explanatory variables. The best model had the building land use as the primary predictor. When comparing two multiple regression models in which the explanatory variables were limited to (1) the land use and soil composition and (2) the water quality, the water quality model underestimated the high normalized dissolve 137Cs concentration in urban areas. This poor reproducibility indicates that the dissolved 137Cs concentration value in urban areas cannot be solely explained by the solid-liquid distribution of 137Cs owing to the influence of the water quality, but some specific 137Cs sources in urban areas would control the dissolved 137Cs concentration.
Keywords: Fukushima Daiichi nuclear power plant accident; Land use; Multiple regression model; River water; Topographic wetness index; Water quality.
Copyright © 2019 The Authors. Published by Elsevier B.V. All rights reserved.