Shen Youchen, de Hoogh Kees, Schmitz Oliver, Clinton Nicholas, Tuxen-Bettman Karin, Brandt Jørgen, Christensen Jesper H, Frohn Lise M, Geels Camilla, Karssenberg Derek, Vermeulen Roel, Hoek Gerard
Division of Environmental Epidemiology, Institute for Risk Assessment Sciences, Utrecht University, Utrecht, the Netherlands.
Swiss Tropical and Public Health Institute, Basel, Switzerland; University of Basel, Basel, Switzerland.
Environ Int. 2022 Oct;168:107485. doi: 10.1016/j.envint.2022.107485. Epub 2022 Aug 24.
Previous European land-use regression (LUR) models assumed fixed linear relationships between air pollution concentrations and predictors such as traffic and land use. We evaluated whether including spatially-varying relationships could improve European LUR models by using geographically weighted regression (GWR) and random forest (RF). We built separate LUR models for each year from 2000 to 2019 for NO, O, PM and PM using annual average monitoring observations across Europe. Potential predictors included satellite retrievals, chemical transport model estimates and land-use variables. Supervised linear regression (SLR) was used to select predictors, and then GWR estimated the potentially spatially-varying coefficients. We developed multi-year models using geographically and temporally weighted regression (GTWR). Five-fold cross-validation per year showed that GWR and GTWR explained similar spatial variations in annual average concentrations (average R = NO: 0.66; O: 0.58; PM: 0.62; PM: 0.77), which are better than SLR (average R = NO: 0.61; O: 0.46; PM: 0.51; PM: 0.75) and RF (average R = NO: 0.64; O: 0.53; PM: 0.56; PM: 0.67). The GTWR predictions and a previously-used method of back-extrapolating 2010 model predictions using CTM were overall highly correlated (R > 0.8) for all pollutants. Including spatially-varying relationships using GWR modestly improved European air pollution annual LUR models, allowing time-varying exposure-health risk models.
以往的欧洲土地利用回归（LUR）模型假定空气污染浓度与交通和土地利用等预测因子之间存在固定的线性关系。我们评估了纳入空间变化关系是否能够通过使用地理加权回归（GWR）和随机森林（RF）来改进欧洲LUR模型。我们利用欧洲各地的年度平均监测观测数据，为2000年至2019年的每一年分别构建了LUR模型，用于分析一氧化氮（NO）、臭氧（O₃）、细颗粒物（PM₂.₅）和粗颗粒物（PM₁₀）。潜在的预测因子包括卫星反演数据、化学传输模型估算值和土地利用变量。使用监督线性回归（SLR）来选择预测因子，然后用GWR估计潜在的空间变化系数。我们利用地理和时间加权回归（GTWR）开发了多年模型。每年的五折交叉验证表明，GWR和GTWR对年度平均浓度的空间变化解释程度相似（平均R²：NO为0.66；O₃为0.58；PM₂.₅为0.62；PM₁₀为0.77），优于SLR（平均R²：NO为0.61；O₃为0.46；PM₂.₅为0.51；PM₁₀为0.75）和RF（平均R²：NO为0.64；O₃为0.53；PM₂.₅为0.56；PM₁₀为0.67）。GTWR预测值与之前使用化学传输模型（CTM）对2010年模型预测值进行反向推算所采用的方法，对于所有污染物总体上高度相关（R＞0.8）。使用GWR纳入空间变化关系适度改进了欧洲空气污染年度LUR模型，并使得构建随时间变化的暴露-健康风险模型成为可能
