Neha Singh , Joe Van Buskirk , Sagnik Dey , Luke D. Knibbs
{"title":"National, satellite-based land-use regression models for estimating long-term annual NO2 exposure across India","authors":"Neha Singh , Joe Van Buskirk , Sagnik Dey , Luke D. Knibbs","doi":"10.1016/j.aeaoa.2024.100289","DOIUrl":null,"url":null,"abstract":"<div><p>In India, scarcity of ground-based measurements of nitrogen dioxide (NO<sub>2</sub>) is a major challenge for estimating long-term exposure and associated health impacts. This study aimed to develop and validate a national-scale annual NO<sub>2</sub> exposure model for India for 2019 and determine if model cross-validation predictive ability was improved by including non-continuous (manual) measurements along with reference-grade, continuous measurements.</p><p>We used a supervised forward-addition linear regression method to fit land use regression (LUR) models developed with up to 804 Central Pollution Control Board ground monitoring stations (n = 157 continuous, n = 647 manual) and 209 spatial predictor variables, including satellite-based tropospheric NO<sub>2</sub> columns. Two models were developed: one using continuous sites only and one using continuous and manual sites, with standard diagnostics and cross-validation (CV) methods. We also assessed if the kriging of final model residuals reduced spatial autocorrelation and improved model CV results. LUR coefficients for the best-performing model were applied to predictors for 2015–2021 and gridded at 100 m to estimate population-weighted exposure.</p><p>The continuous sites-only model and combined continuous and manual sites models had CV-R<sup>2</sup> values of 0.59 (root-mean-square error [RMSE]: 9.4 μg/m<sup>3</sup>) and 0.54 (RMSE: 8.3 μg/m<sup>3</sup>), respectively, and both included the satellite NO<sub>2</sub> predictor. Kriging residuals increased the CV-R<sup>2</sup> of the combined model to 0.70 (RMSE: 7.2 μg/m<sup>3</sup>) but offered no improvement for the continuous site model. National population-weighted average NO<sub>2</sub> was 22.1 μg/m<sup>3</sup> in 2019. We estimated over 92% of the Indian population was exposed to annual NO<sub>2</sub> exceeding the WHO air quality guideline (10 μg/m<sup>3</sup>). In Delhi, Mumbai, and Kolkata, an estimated 45%, 100%, and 100% of the population, respectively, experienced annual NO<sub>2</sub> levels that surpassed Indian standards (40 μg/m<sup>3</sup>). To our knowledge, this is the first such long-term NO<sub>2</sub> LUR model specific to India, and predictions are available to interested researchers.</p></div>","PeriodicalId":37150,"journal":{"name":"Atmospheric Environment: X","volume":"23 ","pages":"Article 100289"},"PeriodicalIF":3.8000,"publicationDate":"2024-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S259016212400056X/pdfft?md5=7cda52bb136ceb925a6c17a741e1f895&pid=1-s2.0-S259016212400056X-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Atmospheric Environment: X","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S259016212400056X","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
In India, scarcity of ground-based measurements of nitrogen dioxide (NO2) is a major challenge for estimating long-term exposure and associated health impacts. This study aimed to develop and validate a national-scale annual NO2 exposure model for India for 2019 and determine if model cross-validation predictive ability was improved by including non-continuous (manual) measurements along with reference-grade, continuous measurements.
We used a supervised forward-addition linear regression method to fit land use regression (LUR) models developed with up to 804 Central Pollution Control Board ground monitoring stations (n = 157 continuous, n = 647 manual) and 209 spatial predictor variables, including satellite-based tropospheric NO2 columns. Two models were developed: one using continuous sites only and one using continuous and manual sites, with standard diagnostics and cross-validation (CV) methods. We also assessed if the kriging of final model residuals reduced spatial autocorrelation and improved model CV results. LUR coefficients for the best-performing model were applied to predictors for 2015–2021 and gridded at 100 m to estimate population-weighted exposure.
The continuous sites-only model and combined continuous and manual sites models had CV-R2 values of 0.59 (root-mean-square error [RMSE]: 9.4 μg/m3) and 0.54 (RMSE: 8.3 μg/m3), respectively, and both included the satellite NO2 predictor. Kriging residuals increased the CV-R2 of the combined model to 0.70 (RMSE: 7.2 μg/m3) but offered no improvement for the continuous site model. National population-weighted average NO2 was 22.1 μg/m3 in 2019. We estimated over 92% of the Indian population was exposed to annual NO2 exceeding the WHO air quality guideline (10 μg/m3). In Delhi, Mumbai, and Kolkata, an estimated 45%, 100%, and 100% of the population, respectively, experienced annual NO2 levels that surpassed Indian standards (40 μg/m3). To our knowledge, this is the first such long-term NO2 LUR model specific to India, and predictions are available to interested researchers.