• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Joint annual meeting of the International Society of Exposure Science and the International Society for Environmental Epidemiology >Development of Spatio-Temporal Land Use Regression Models for Gaseous Pollutants in London, UK, within the Steam Project
【24h】

Development of Spatio-Temporal Land Use Regression Models for Gaseous Pollutants in London, UK, within the Steam Project

机译：Steam项目中英国伦敦的气态污染物时空土地利用回归模型的开发

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Background/Aim: Epidemiological studies either use fixed site monitor measurements or modeled estimates at subjects' residential addresses to estimate health effects of short or long-term exposures. Recent developments in exposure modeling provide spatially resolved daily estimates enabling an integrated assessment of health effects arising from both long-and-short-term exposures. Our aim was to develop a spatio-temporal land use regression (LUR) model that estimates daily concentrations of nitrogen oxides (NOx), nitrogen dioxide (NO2) and ozone (O3), in London, UK. Methods: Data from the extensive network of fixed monitoring sites in London were collected for the years 2009-201$3while LUR variables were derived from Land Cover Map 2007.We obtained road geography and information on road traffic flows from the Department of Transport. We applied a semiparametric approach using spatial covariates, meteorological data, time varying variables and a bivariate smooth thin plate function. The final set of explanatory variables was selected based on the adjusted-R2. Moreover, we developed hybrid models by incorporating chemical transport modeling (CTM) predictions, obtained within the STEAM project for the same area and time-period, into the developed spatio-temporal LUR models. Results: The adjusted-R2 of the developed LUR models was 0.75 for NOx, 0.72 for NO2 and 0.69 for O3. We performed a ten-fold cross-validation and the adjusted-R2 were 0.74, 0.71 and 0.62 for NOx, NO2 and O3 respectively. Incorporation of the dispersion estimates in the LUR models as a predictor, improved the LUR model fit: CV-R2 increased to 0.77 for NOx, to 0.76 NO2 and to 0.75 for O3. Conclusions: Our study supports the combined use of LUR and CTM in a single spatio-temporal modeling framework to improve the accuracy of predictions for subsequent use in epidemiological studies.
机译：背景/目的：流行病学研究使用固定站点监测器测量值或对受试者住所地址的模型化估计来估计短期或长期暴露对健康的影响。接触模型的最新发展提供了空间分解的每日估计值，从而能够综合评估长期和短期接触所产生的健康影响。我们的目标是在英国伦敦开发一个时空土地利用回归（LUR）模型，以估算氮氧化物，二氧化氮和臭氧的日浓度。方法：从伦敦固定监控站点的广泛网络中收集数据，这些数据来自2009-201年$ 3美元，而LUR变量来自2007年《土地覆盖图》，我们从交通运输部获得了道路地理信息和道路交通流量信息。我们使用空间协变量，气象数据，时变变量和双变量平滑薄板函数应用半参数方法。根据调整后的R2选择最终的解释变量集。此外，我们通过将在STEAM项目中针对相同区域和时间段获得的化学迁移模型（CTM）预测合并到已开发的时空LUR模型中来开发混合模型。结果：已开发的LUR模型的调整后R2（NOx）为0.75，NO2为0.72，O3为0.69。我们进行了十次交叉验证，对于NOx，NO2和O3，调整后的R2分别为0.74、0.71和0.62。将色散估计值纳入LUR模型作为预测指标，可以改善LUR模型的拟合度：对于NOx，CV-R2增加至0.77，对于NO3，CV-R2增加至0.76，而对于O3增加至0.75。结论：我们的研究支持在单个时空建模框架中结合使用LUR和CTM，以提高预测的准确性，以供随后在流行病学研究中使用。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号