Department of Geology and Geological Engineering, Colorado School of Mines, Golden, CO 80401, USA.
Health Phys. 2011 Dec;101(6):722-30. doi: 10.1097/HP.0b013e31821ddb07.
A complete, historical dataset is presented of radionuclide resuspension-factors. These data span six orders of magnitude in time (ranging from 0.1 to 73,000 d), encompass more than 300 individual values, and combine observations from events on three continents. These data were then used to derive improved, empirical models that can be used to predict resuspension of trace materials after their deposit on the ground. Data-fitting techniques were used to derive models of various types and an estimate of uncertainty in model prediction. Two models were found to be suitable: a power law and the modified Anspaugh et al. model, which is a double exponential. Though statistically the power-law model provides the best metrics of fit, the modified Anspaugh model is deemed the more appropriate due to its better fit to data at early times and its ease of implementation in terms of closed analytical integrals.
给出了一个完整的、历史的放射性核素再悬浮因子数据集。这些数据跨越了六个数量级的时间范围（从 0.1 到 73,000 天），包含了 300 多个单独的值，并结合了来自三大洲事件的观测结果。然后，使用这些数据推导出了改进的经验模型，可用于预测在地面上沉积后痕量物质的再悬浮。使用数据拟合技术推导出了各种类型的模型，并估计了模型预测的不确定性。发现有两种模型是合适的：幂律模型和修正后的 Anspaugh 等人的模型，即双指数模型。尽管从统计学角度来看，幂律模型提供了最佳的拟合度量，但修正后的 Anspaugh 模型被认为更合适，因为它在早期数据拟合方面更好，并且在封闭解析积分方面易于实现。
