This study presents an alternative method to the traditional reference dose approach for the determination of health risk-based soil standards for arsenic. The model combines multimedia equations for air, soil and dietary exposure, a toxicokinetic component, a probabilistic output, a reference distribution for urine arsenic in the general population and exposure parameters values traceable in US-EPA or EFSA reference studies. The model calculates a mean inorganic As urine concentration AsU (sum of inorganic arsenic Asi and its metabolites) from environmental data and exposure parameters, which is attributed to the central value of a lognormal distribution. Risk is assessed by comparing a high percentile of the modelled distribution to the target AsU attributed to the reference value of 10 µg/g_CREA (microgram As per gram of creatinine), the upper confidence interval of the 95^th percentile from the AsU distribution in the French population. A soil standard value of 40 mg/kg is determined as being the arsenic concentration in soil not giving rise to more than 5% probability of having a AsU concentration above the target value of 10 µg/g_CREA. Once soil As concentrations above 40 mg/kg are measured, further environmental investigations should be carried, involving an assessment of As bioaccessibility to address health risks and decide of regulatory measures in residential setting.

本研究提出了一种替代传统参考剂量方法的方法，用于确定基于健康风险的土壤砷标准。该模型结合了空气、土壤和饮食暴露的多媒体方程、毒代动力学成分、概率输出、一般人群中尿砷的参考分布以及可在 US-EPA 或 EFSA 参考研究中追溯的暴露参数值。该模型根据环境数据和暴露参数计算平均无机砷尿浓度 AsU（无机砷 Asi 及其代谢物的总和），这归因于对数正态分布的中心值。_{通过将模型分布的高百分比与归因于 10 µg/g CREA}参考值的目标 AsU 进行比较来评估风险（微克每克肌酐），法国人口中 AsU 分布的第 95^{个百分位数的上置信区间。}土壤标准值 40 mg/kg 被确定为土壤中的砷浓度不会导致 AsU 浓度高于 10 µg/g _CREA目标值的概率超过 5% 。一旦测量到土壤砷浓度超过 40 毫克/千克，应进行进一步的环境调查，包括评估砷的生物可及性，以解决健康风险并决定居住环境中的监管措施。