US COVID-19 Shutdown Demonstrates Importance of Background NO2 in Inferring NOx Emissions From Satellite NO2 Observations.

Satellite nitrogen dioxide (NO2) measurements are used extensively to infer nitrogen oxide emissions and their trends, but interpretation can be complicated by background contributions to the NO2 column sensed from space. We use the step decrease of US anthropogenic emissions from the COVID-19 shutdown to compare the responses of NO2 concentrations observed at surface network sites and from satellites (Ozone Monitoring Instrument [OMI], Tropospheric Ozone Monitoring Instrument [TROPOMI]). After correcting for differences in meteorology, surface NO2 measurements for 2020 show decreases of 20% in March-April and 10% in May-August compared to 2019. The satellites show much weaker responses in March-June and no decrease in July-August, consistent with a large background contribution to the NO2 column. Inspection of the long-term OMI trend over remote US regions shows a rising summertime NO2 background from 2010 to 2019 potentially attributable to wildfires.

An Ensemble Learning Approach for Estimating High Spatiotemporal Resolution of Ground-Level Ozone in the Contiguous United States.

In this paper, we integrated multiple types of predictor variables and three types of machine learners (neural network, random forest, and gradient boosting) into a geographically weighted ensemble model to estimate the daily maximum 8 h O3 with high resolution over both space (at 1 km × 1 km grid cells covering the contiguous United States) and time (daily estimates between 2000 and 2016). We further quantify monthly model uncertainty for our 1 km × 1 km gridded domain. The results demonstrate high overall model performance with an average cross-validated R2 (coefficient of determination) against observations of 0.90 and 0.86 for annual averages. Overall, the model performance of the three machine learning algorithms was quite similar. The overall model performance from the ensemble model outperformed those from any single algorithm. The East North Central region of the United States had the highest R2, 0.93, and performance was weakest for the western mountainous regions (R2 of 0.86) and New England (R2 of 0.87). For the cross validation by season, our model had the best performance during summer with an R2 of 0.88. This study can be useful for the environmental health community to more accurately estimate the health impacts of O3 over space and time, especially in health studies at an intra-urban scale.

Assessing NO2 Concentration and Model Uncertainty with High Spatiotemporal Resolution across the Contiguous United States Using Ensemble Model Averaging.

NO2 is a combustion byproduct that has been associated with multiple adverse health outcomes. To assess NO2 levels with high accuracy, we propose the use of an ensemble model to integrate multiple machine learning algorithms, including neural network, random forest, and gradient boosting, with a variety of predictor variables, including chemical transport models. This NO2 model covers the entire contiguous U.S. with daily predictions on 1-km-level grid cells from 2000 to 2016. The ensemble produced a cross-validated R2 of 0.788 overall, a spatial R2 of 0.844, and a temporal R2 of 0.729. The relationship between daily monitored and predicted NO2 is almost linear. We also estimated the associated monthly uncertainty level for the predictions and address-specific NO2 levels. This NO2 estimation has a very high spatiotemporal resolution and allows the examination of the health effects of NO2 in unmonitored areas. We found the highest NO2 levels along highways and in cities. We also observed that nationwide NO2 levels declined in early years and stagnated after 2007, in contrast to the trend at monitoring sites in urban areas, where the decline continued. Our research indicates that the integration of different predictor variables and fitting algorithms can achieve an improved air pollution modeling framework.

Air Quality and Health Impact of Future Fossil Fuel Use for Electricity Generation and Transport in Africa.

Africa has ambitious plans to address energy deficits and sustain economic growth with fossil fueled power plants. The continent is also experiencing faster population growth than anywhere else in the world that will lead to proliferation of vehicles. Here, we estimate air pollutant emissions in Africa from future (2030) electricity generation and transport. We find that annual emissions of two precursors of fine particles (PM2.5) hazardous to health, sulfur dioxide (SO2) and nitrogen oxides (NOx), approximately double by 2030 relative to 2012, increasing from 2.5 to 5.5 Tg SO2 and 1.5 to 2.8 Tg NOx. We embed these emissions in the GEOS-Chem model nested over the African continent to simulate ambient concentrations of PM2.5 and determine the burden of disease (excess deaths) attributable to exposure to future fossil fuel use. We calculate 48000 avoidable deaths in 2030 (95% confidence interval: 6000-88000), mostly in South Africa (10400), Nigeria (7500), and Malawi (2400), with 3-times higher mortality rates from power plants than transport. Sensitivity of the burden of disease to either population growth or air quality varies regionally and suggests that emission mitigation strategies would be most effective in Southern Africa, whereas population growth is the main driver everywhere else.

An ensemble-based model of PM2.5 concentration across the contiguous United States with high spatiotemporal resolution.

Various approaches have been proposed to model PM2.5 in the recent decade, with satellite-derived aerosol optical depth, land-use variables, chemical transport model predictions, and several meteorological variables as major predictor variables. Our study used an ensemble model that integrated multiple machine learning algorithms and predictor variables to estimate daily PM2.5 at a resolution of 1 km × 1 km across the contiguous United States. We used a generalized additive model that accounted for geographic difference to combine PM2.5 estimates from neural network, random forest, and gradient boosting. The three machine learning algorithms were based on multiple predictor variables, including satellite data, meteorological variables, land-use variables, elevation, chemical transport model predictions, several reanalysis datasets, and others. The model training results from 2000 to 2015 indicated good model performance with a 10-fold cross-validated R2 of 0.86 for daily PM2.5 predictions. For annual PM2.5 estimates, the cross-validated R2 was 0.89. Our model demonstrated good performance up to 60 µg/m3. Using trained PM2.5 model and predictor variables, we predicted daily PM2.5 from 2000 to 2015 at every 1 km × 1 km grid cell in the contiguous United States. We also used localized land-use variables within 1 km × 1 km grids to downscale PM2.5 predictions to 100 m × 100 m grid cells. To characterize uncertainty, we used meteorological variables, land-use variables, and elevation to model the monthly standard deviation of the difference between daily monitored and predicted PM2.5 for every 1 km × 1 km grid cell. This PM2.5 prediction dataset, including the downscaled and uncertainty predictions, allows epidemiologists to accurately estimate the adverse health effect of PM2.5. Compared with model performance of individual base learners, an ensemble model would achieve a better overall estimation. It is worth exploring other ensemble model formats to synthesize estimations from different models or from different groups to improve overall performance.