China is facing serious heavy metal pollution in farmland soil, which is a major pressing concern for food safety and human health. This research delivers an integrated methodology for pollution source apportionment and a soil-rice-human model to predict heavy metal transfer in the farmland soil, rice grain and human blood chain. The source identification integrated positive matrix factorization (PMF), cluster analysis (CA) and the life cycle assessment (LCA) survey of agricultural and industrial production and consumption. Based on the case analysis of Shaoxing, this method showed very good performance through the illustration of the source contributions by PMF and LCA at county level and the identification of the pollution sources using CA and LCA at field scale. According to the overall evaluation, the integrated method was superior for the farmland metals pollution source identification comparing to existing source apportionment methods. To predict metal transformation in soil-rice-human chain, a set of models of metals (As, Pb, Cd, Hg, Cr) accumulation ability in rice grain and human blood has been established by literature review and monitoring data. The models showed adequate predictability for the metal content of rice grains at both the field and regional scale, and plausible simulation of the metal concentration in human blood throughout the whole study region. Therefore, this study provides valuable tools for farmland soil heavy metal pollution source identification and for the prediction of heavy metal transformation in soil-rice-human chain; and it can highlight the need to take mitigating action to reduce farmland metal pollution risks in specific regions.

中文翻译：

---

农田土壤金属源解析及土壤-水稻-人链中金属迁移预测[J].

我国农田土壤重金属污染严重，是食品安全和人类健康的重大紧迫问题。该研究提供了一种污染源解析的综合方法和土壤-水稻-人体模型，以预测农田土壤、水稻籽粒和人体血链中的重金属迁移。源识别集成了农业和工业生产和消费的正矩阵分解（PMF）、聚类分析（CA）和生命周期评估（LCA）调查。基于绍兴的案例分析，该方法通过县级PMF和LCA的源贡献说明和现场尺度CA和LCA的污染源识别显示出非常好的效果。根据综合评价，综合方法在农田金属污染源识别方面优于现有的源解析方法。为预测土壤-水稻-人体链中的金属转化，通过文献回顾和监测数据，建立了一套水稻籽粒和人体血液中金属（As、Pb、Cd、Hg、Cr）积累能力模型。该模型显示了在田间和区域尺度上稻谷中金属含量的充分可预测性，并在整个研究区域内对人体血液中的金属浓度进行了合理的模拟。因此，本研究为农田土壤重金属污染源识别和土壤-水稻-人链重金属转化预测提供了有价值的工具；