Institute of Statistical Science, Academia Sinica, Taipei, Taiwan, 128, Academia Road, Section 2, Nangang, Taipei, 11529, Taiwan.
Institute of Information Science, Academia Sinica, Taipei, Taiwan, 128, Academia Road, Section 2, Nangang, Taipei, 11529, Taiwan.
Environ Pollut. 2020 Sep;264:114810. doi: 10.1016/j.envpol.2020.114810. Epub 2020 May 13.
A widespread monitoring network of Airbox microsensors was implemented since 2016 to provide high-resolution spatial distributions of ground-level PM data in Taiwan. We developed models for estimating ground-level PM concentrations for all the 3 km × 3 km grids in Taiwan by combining the data from air quality monitoring stations and the Airbox sensors. The PM data from the Airbox sensors (AB-PM) was used to predict daily mean PM levels at the grids in 2017 using a semiparametric additive model. The estimated PM level at the grids was further applied as a predictor variable in the models to predict the monthly mean concentration of PM at all the grids in the previous year. The modeling-predicting procedures were repeated backward for the years from 2016 to 2006. The model results revealed that the model R increased from 0.40 to 0.87 when the AB-PM data were included as a nonlinear component in the model, indicating that AB-PM is a significant predictor of ground-level PM concentration. The cross-validation (CV) results demonstrated that the root of mean squared prediction errors of the estimated monthly mean PM concentrations were smaller than 5 μg/m and the R of the CV models of 0.79-0.88 during 2006-2017. We concluded that Airbox sensors can be used with monitoring data to more accurately estimate long-term exposure to PM for cohorts of small areas in health impact assessment studies.
自 2016 年以来，我们实施了广泛的 Airbox 微传感器监测网络，以提供台湾地区地面 PM 数据的高分辨率空间分布。我们通过结合空气质量监测站和 Airbox 传感器的数据，为台湾所有 3km×3km 的网格开发了估算地面 PM 浓度的模型。使用半参数加法模型，将 Airbox 传感器（AB-PM）的数据用于预测 2017 年网格的日平均 PM 水平。将网格的估算 PM 水平进一步应用为模型中的预测变量，以预测上一年所有网格的月平均 PM 浓度。对 2016 年至 2006 年的年份重复建模-预测过程。模型结果表明，当 AB-PM 数据作为模型的非线性成分时，模型 R 从 0.40 增加到 0.87，表明 AB-PM 是地面 PM 浓度的重要预测因子。交叉验证（CV）结果表明，2006-2017 年，估算的月平均 PM 浓度的均方预测误差的平方根小于 5μg/m，CV 模型的 R 在 0.79-0.88 之间。我们得出结论，Airbox 传感器可以与监测数据一起使用，以更准确地估计健康影响评估研究中小面积队列的 PM 长期暴露情况。
