A statistical approach to estimating soil-to-plant transfer factor of strontium in agricultural fields

Abstract

Soil-to-plant transfer factor (TF) is one of the important parameters in radiation dose assessment models for the environmental transfer of radionuclides. Since TFs are affected by several factors, including radionuclides, plant species and soil properties, development of a method for estimation of TF using some soil and plant properties would be useful. In this study, we took a statistical approach to estimating the TF of stable strontium (TFsr) from selected soil properties and element concentrations in plants, which was used as an analogue of 90Sr. We collected the plant and soil samples used for the study from 142 agricultural fields throughout Japan. We applied a multiple linear regression analysis in order to get an empirical equation to estimate TFsr. TFsr could be estimated from the Sr concentration in soil (C Srsoil1) and Ca concentration in crop (Ccacrop) using the following equation: log TFSr- -0.88-logCsrsoil+0.93- logCcacrop-2.53. Then, we replaced our data with Ca concentrations in crops from a food composition database compiled by the Japanese government. Finally, we predicted TFsr, using Sr concentration in soil from our data and Ca concentration in crops from the database of food composition. © Atomic Energy Society of Japan.