A Knowledge Transfer Approach to Map Long-Term Concentrations of Hyperlocal Air Pollution from Short-Term Mobile Measurements

High-Resolution Geospatial Database: National Criteria-Air-Pollutant Concentrations in the Contiguous U.S., 2016-2020

Concentration estimates for ambient air pollution are used widely in fields such as environmental epidemiology, health impact assessment, urban planning, environmental equity and sustainability. This study builds on previous efforts by developing an updated high-resolution geospatial database of population-weighted annual-average concentrations for six criteria air pollutants (PM2.5, PM10, CO, NO2, SO2, O3) across the contiguous U.S. during a five-year period (2016-2020). We developed Land Use Regression (LUR) models within a partial-least-squares-universal kriging framework by incorporating several land use, geospatial and satellite-based predictor variables. The LUR models were validated using conventional and clustered cross-validation, with the former consistently showing superior performance in capturing the variability of air quality. Most models demonstrated reliable performance (e.g., mean squared error-based R 2 > 0.8, standardised root mean squared error < 0.1). We used the best modelling approach to develop estimates by Census Block, which were then population-weighted averaged at Census Block Group, Census Tract and County geographies. Our database provides valuable insights into the dynamics of air pollution, with utility for environmental risk assessment, public health, policy and urban planning.

Assessing the role of spatial aggregation schemes with varying campaign durations of mobile measurements on land use regression models for estimating nitrogen dioxide

Mobile air pollution measurements are typically aggregated by varying road segment lengths, grid cell sizes, and time intervals. How these spatiotemporal aggregation schemas affect the modeling performance of land use regression models has seldom been assessed. We used 5.7 million mobile nitrogen dioxide (NO2) measurements collected over 160 days in Amsterdam (The Netherlands) and subsampled them into five campaign durations (10-70 days). We aggregated the measurements from each campaign duration onto road segments and grid cells with five spatial scales (25-200 m). A stepwise linear regression (SLRs) and random forests (RFs) were trained for each aggregated dataset to predict NO2 concentrations. The model accuracies were validated using a 30% hold-out sample of mobile measurements and external Palmes long-term stationary measurements (n = 105). At increased spatial scales, the prediction accuracy decreased for RFs but increased for SLRs when validated against mobile measurements. Using long-term stationary measurements, prediction accuracy varied across scales without any clear pattern. Regardless of cells or road segments, the models performed similarly at small scales (i.e., 25 m and 50 m). Models based on road segments were less sensitive to spatial scales than those based on cells in mobile and long-term external validations. Longer campaign durations increased the prediction accuracies of long-term NO2 concentrations, though the gain in accuracy diminished after 50 days. In conclusion, our results suggest that road segments are preferred when the aggregation scale gets larger as this approach likely reduces scale-dependent influences. The campaign duration plays a more important role in long-term NO2 prediction than spatial scales.

High-resolution spatio-temporal estimation of street-level air pollution using mobile monitoring and machine learning

High spatio-temporal resolution street-level air pollution (SLAP) estimation is essential for urban air quality management, yet traditional methods face significant challenges in capturing the detailed spatial and temporal variability of pollution. Methods relying on fixed monitoring networks provide limited spatial coverage, while those utilizing mobile monitoring campaigns, despite their flexibility, often suffer from data sparsity and temporal incompleteness. To address these limitations, we propose a Two-Step Machine Learning Gap-Filling Framework employing a Multi-task Graph-based XGBoost (MTGXGB) model to enhance SLAP resolution. This framework expands high-resolution pollution estimation from a purely spatial perspective to a spatio-temporal view and effectively addresses data gaps. Our approach achieves spatial resolutions of 30-200 m and hourly temporal resolutions, capturing both short- and long-term variations in PM2.5 concentrations. Applying this framework to London's urban environment, we identify critical pollution hotspots and uncover correlations between SLAP, traffic speed, and urban environmental features. Additionally, the derived uncertainty maps provide actionable insights for optimizing mobile monitoring strategies. This study advances machine learning methodologies for spatio-temporal SLAP estimation and highlights the potential of high-resolution spatio-temporal SLAP data to inform policy-making, such as Low Emission Zones (LEZs), thereby demonstrating its practicality and scalability for urban air quality management.

Dual-Stage Stacking Machine Learning Method Considering Virtual Sample Generation for the Prediction of ZIF-8' BET Specific Surface Area with Experimental Validation

The widespread application of metal-organic frameworks (MOFs) in wastewater and gas treatment has created an increasing demand for accurate and rapid assessment of their BET specific surface area. However, experimental methods for acquiring sufficient statistical data are often costly and time-consuming. Therefore, this study proposes a dual-stage stacking model with Gaussian mixture model-virtual sample generation (GMM-VSG) technology for the BET specific surface area prediction. In this study, 90 real samples were selected from the MOF database and 300 virtual samples were generated. The performance on both real and virtual samples was evaluated by using four machine learning models, including Bayesian regression (Bayes), adaptive boosting (AdaBoost), random forest (RF), and extreme gradient boosting (XGBoost). Subsequently, three best-performing models and a linear regression model were selected for constructing a two-stage stacking model, with R2 value of 0.974. Finally, experimental conditions were adjusted based on feature importance analysis during the validation process, and the result shows that the prediction accuracy of the BET specific surface area is 0.943. This study contributes to the development of more efficient and accurate evaluation methods.

Electric buses as an air pollution and meteorological observation network: Methodology and preliminary results

Many local agencies in the United States and other countries are tasked to install air pollution monitoring systems of highly accurate sensors that have high acquisition, operating, and maintenance costs. The need for expanded coverage of air quality measurements across Salt Lake County (SLCO), Utah is being met by mounting air quality and temperature sensors on an expanding fleet of battery electric buses (BEBs). Monitoring air quality from a mobile sensor network provides real-time insights into air pollution patterns at high temporal and spatial resolution. Mobile measurements contribute to assessing residents' exposure to air pollution, facilitating the implementation of cost-effective public health policies and highlighting disparities. The Electric Bus Air Quality Observation Project was launched in SLCO during July 2021 and has collected millions of observations to date. A BEB traveling at typical traffic speeds (~10 m s-1) can provide multiple measurements along city block lengths of up to ~200 m. With careful analysis that factors in the time response of the differing sensors, variability from block-to-block may be attributed to fine-scale factors (e.g., pollution and heat sources, tree shading and urban vegetation, etc.). Preliminary findings showcase the value of increased coverage and resolution. During an extreme heat event in July 2023, both the morning and afternoon temperature readings showed differences of over 6.5 °C (12 °F), primarily as an east-west gradient with similar gradients in ozone. We conclude that temperature and pollutant concentration readings, at fine spatial and temporal resolutions, will facilitate future health studies and equitable policy and mitigation strategies. Our study demonstrates that our partnerships established with governmental, non-profit, and transit agencies facilitate the successful transfer of research and development to operational real-time mobile air quality monitoring.

Integrated Transfer Learning and Multitask Learning Strategies to Construct Graph Neural Network Models for Predicting Bioaccumulation Parameters of Chemicals

Accurate prediction of parameters related to the environmental exposure of chemicals is crucial for the sound management of chemicals. However, the lack of large data sets for training models may result in poor prediction accuracy and robustness. Herein, integrated transfer learning (TL) and multitask learning (MTL) was proposed for constructing a graph neural network (GNN) model (abbreviated as TL-MTL-GNN model) using n-octanol/water partition coefficients as a source domain. The TL-MTL-GNN model was trained to predict three bioaccumulation parameters based on enlarged data sets that cover 2496 compounds with at least one bioaccumulation parameter. Results show that the TL-MTL-GNN model outperformed single-task GNN models with and without the TL, as well as conventional machine learning models trained with molecular descriptors or fingerprints. Applicability domains were characterized by a state-of-the-art structure-activity landscape-based (abbreviated as ADSAL) methodology. The TL-MTL-GNN model coupled with the optimal ADSAL was employed to predict bioaccumulation parameters for around 60,000 chemicals, with more than 13,000 compounds identified as bioaccumulative chemicals. The high predictive accuracy and robustness of the TL-MTL-GNN model demonstrate the feasibility of integrating the TL and MTL strategy in modeling small-sized data sets. The strategy holds significant potential for addressing small data challenges in modeling environmental chemicals.

Hyperlocal Air Pollution Mapping: A Scalable Transfer Learning LUR Approach for Mobile Monitoring

Addressing the challenge of mapping hyperlocal air pollution in areas without local monitoring, we evaluated unsupervised transfer learning-based land-use regression (LUR) models developed using mobile monitoring data from other cities: CORrelation ALignment (Coral) and its inverse distance-weighted modification (IDW_Coral). These models mitigated domain shifts and transferred patterns learned from mobile air quality monitoring campaigns in Copenhagen and Rotterdam to estimate annual average air pollution levels in Amsterdam (50m road segments) without involving any Amsterdam measurements in model development. For nitrogen dioxide (NO2), IDW_Coral outperformed Copenhagen and Rotterdam LUR models directly applied to Amsterdam, achieving MAE (4.47 μg/m3) and RMSE (5.36 μg/m3) comparable to a locally fitted LUR model (AMS_SLR) developed using Amsterdam mobile measurements collected for 160 days. IDW_Coral yielded an R2 of 0.35, similar to that of the AMS_SLR based on 20 collection days, suggesting a minimum requirement of 20-day mobile monitoring to capture city-specific insights. For ultrafine particles (UFP), IDW_Coral's citywide predictions strongly correlated with previously published mixed-effect models fitted with 160-day Amsterdam measurements (Pearson correlation of 0.71 for UFP and 0.72 for NO2). IDW_Coral demands no direct measurements in the target area, showcasing its potential for large-scale applications and offering significant economic efficiencies in executing mobile monitoring campaigns.

LUR modeling of long-term average hourly concentrations of NO2 using hyperlocal mobile monitoring data

Mobile monitoring campaigns have effectively captured spatial hyperlocal variations in long-term average concentrations of regulated and unregulated air pollutants. However, their application in estimating spatiotemporally varying maps has rarely been investigated. Tackling this gap, we investigated whether mobile measurements can assess long-term average nitrogen dioxide (NO2) concentrations for each hour of the day. Using mobile NO2 data monitored for 10 months in Amsterdam, we examined the performance of two spatiotemporal land use regression (LUR) methods, Spatiotemporal-Kriging and GTWR (Geographical and Temporal Weighted Regression), alongside two classical spatial LUR models developed separately for each hour. We found that mobile measurements follow the general pattern of fixed-site measurements, but with considerable deviations (indicating collection uncertainty). Leveraging heterogeneous spatiotemporal autocorrelations, GTWR smoothed these deviations and achieved an overall performance of an R2 of 0.49 and a Mean Absolute Error of 6.33 μg/m3, validated by long-term fixed-site measurements (out-of-sample). The other models tested were more affected by the collection uncertainty. We highlighted that the spatiotemporal variations captured in mobile measurements can be used to reconstruct long-term average hourly air pollution maps. These maps facilitate dynamic exposure assessments considering spatiotemporal human activity patterns.

A comprehensive review of the development of land use regression approaches for modeling spatiotemporal variations of ambient air pollution: A perspective from 2011 to 2023

Land use regression (LUR) models are widely used in epidemiological and environmental studies to estimate humans' exposure to air pollution within urban areas. However, the early models, developed using linear regressions and data from fixed monitoring stations and passive sampling, were primarily designed to model traditional and criteria air pollutants and had limitations in capturing high-resolution spatiotemporal variations of air pollution. Over the past decade, there has been a notable development of multi-source observations from low-cost monitors, mobile monitoring, and satellites, in conjunction with the integration of advanced statistical methods and spatially and temporally dynamic predictors, which have facilitated significant expansion and advancement of LUR approaches. This paper reviews and synthesizes the recent advances in LUR approaches from the perspectives of the changes in air quality data acquisition, novel predictor variables, advances in model-developing approaches, improvements in validation methods, model transferability, and modeling software as reported in 155 LUR studies published between 2011 and 2023. We demonstrate that these developments have enabled LUR models to be developed for larger study areas and encompass a wider range of criteria and unregulated air pollutants. LUR models in the conventional spatial structure have been complemented by more complex spatiotemporal structures. Compared with linear models, advanced statistical methods yield better predictions when handling data with complex relationships and interactions. Finally, this study explores new developments, identifies potential pathways for further breakthroughs in LUR methodologies, and proposes future research directions. In this context, LUR approaches have the potential to make a significant contribution to future efforts to model the patterns of long- and short-term exposure of urban populations to air pollution.

Integrating large-scale stationary and local mobile measurements to estimate hyperlocal long-term air pollution using transfer learning methods

Mobile air quality measurements are collected typically for several seconds per road segment and in specific timeslots (e.g., working hours). These short-term and on-road characteristics of mobile measurements become the ubiquitous shortcomings of applying land use regression (LUR) models to estimate long-term concentrations at residential addresses. This issue was previously found to be mitigated by transferring LUR models to the long-term residential domain using routine long-term measurements in the studied region as the transfer target (local scale). However, long-term measurements are generally sparse in individual cities. For this scenario, we propose an alternative by taking long-term measurements collected over a larger geographical area (global scale) as the transfer target and local mobile measurements as the source (Global2Local model). We empirically tested national, airshed countries (i.e., national plus neighboring countries) and Europe as the global scale in developing Global2Local models to map nitrogen dioxide (NO₂) concentrations in Amsterdam. The airshed countries scale provided the lowest absolute errors, and the Europe-wide scale had the highest R². Compared to a "global" LUR model (trained exclusively with European-wide long-term measurements), and a local mobile LUR model (using mobile data from Amsterdam only), the Global2Local model significantly reduced the absolute error of the local mobile LUR model (root-mean-square error, 6.9 vs 12.6 μg/m³) and improved the percentage explained variances compared to the global model (R², 0.43 vs 0.28, assessed by independent long-term NO₂ measurements in Amsterdam, n = 90). The Global2Local method improves the generalizability of mobile measurements in mapping long-term residential concentrations with a fine spatial resolution, which is preferred in environmental epidemiological studies.