A Novel Flexible Geographically Weighted Neural Network for High-Precision PM2.5 Mapping across the Contiguous United States DOI Creative Commons
Dongchao Wang, Jianfei Cao, Baolei Zhang

et al.

ISPRS International Journal of Geo-Information, Journal Year: 2024, Volume and Issue: 13(7), P. 217 - 217

Published: June 22, 2024

Air quality degradation has triggered a large-scale public health crisis globally. Existing machine learning techniques have been used to attempt the remote sensing estimates of PM2.5. However, many models ignore spatial non-stationarity predictive variables. To address this issue, study introduces Flexible Geographically Weighted Neural Network (FGWNN) estimate PM2.5 based on multi-source data. FGWNN incorporates Geographical Neuron (FGN) and Activation Function (GWAF) within framework Artificial (ANN) capture intricate non-stationary relationships among A robust air estimation model was constructed using data Aerosol Optical Depth (AOD), Normalized Difference Vegetation Index (NDVI), Temperature (TMP), Specific Humidity (SPFH), Wind Speed (WIND), Terrain Elevation (HGT) as inputs, Ground-Based observation. The results indicated that successfully generates with 2.5 km resolution for contiguous United States (CONUS) in 2022. It exhibits higher regression accuracy compared traditional ANN Regression (GWR) models. holds potential applications high-precision high-resolution scenarios.

Language: Английский

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 DOI Creative Commons
Xuying Ma, Bin Zou, Jun Deng

et al.

Environment International, Journal Year: 2024, Volume and Issue: 183, P. 108430 - 108430

Published: Jan. 1, 2024

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

Language: Английский

Citations

48
A review of machine learning for modeling air quality: Overlooked but important issues DOI
Dié Tang, Yu Zhan, Fumo Yang

et al.

Atmospheric Research, Journal Year: 2024, Volume and Issue: 300, P. 107261 - 107261

Published: Jan. 21, 2024

Language: Английский

Citations

37
Estimation of PM2.5 Concentration across China Based on Multi-Source Remote Sensing Data and Machine Learning Methods DOI Creative Commons
Yujie Yang, Zhige Wang, Chunxiang Cao

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(3), P. 467 - 467

Published: Jan. 25, 2024

Long-term exposure to high concentrations of fine particles can cause irreversible damage people’s health. Therefore, it is extreme significance conduct large-scale continuous spatial particulate matter (PM2.5) concentration prediction for air pollution prevention and control in China. The distribution PM2.5 ground monitoring stations China uneven with a larger number southeastern China, while the sites also insufficient quality control. Remote sensing technology obtain information quickly macroscopically. possible predict based on multi-source remote data. Our study took as research area, using Pearson correlation coefficient GeoDetector select auxiliary variables. In addition, long short-term memory neural network random forest regression model were established estimation. We finally selected (R2 = 0.93, RMSE 4.59 μg m−3) our by evaluation index. across 2021 was estimated, then influence factors high-value regions explored. It clear that not only related local geographical meteorological conditions, but closely economic social development.

Language: Английский

Citations

11
Interaction influence characteristics of air quality and aerosol properties between Beijing-Tianjing-Hebei (BTH) and Yangtze River Delta (YRD), China DOI
Guo‐Lin Chen, Yong Han, Xin Wang

et al.

Urban Climate, Journal Year: 2025, Volume and Issue: 61, P. 102395 - 102395

Published: April 4, 2025

Language: Английский

Citations

1
Enhancing PM2.5 Air Pollution Prediction Performance by Optimizing the Echo State Network (ESN) Deep Learning Model Using New Metaheuristic Algorithms DOI Creative Commons
Iman Zandi,

Ali Jafari,

Aynaz Lotfata

et al.

Urban Science, Journal Year: 2025, Volume and Issue: 9(5), P. 138 - 138

Published: April 23, 2025

Air pollution presents significant risks to both human health and the environment. This study uses air meteorological data develop an effective deep learning model for hourly PM2.5 concentration predictions in Tehran, Iran. evaluates efficient metaheuristic algorithms optimizing hyperparameters improve accuracy of predictions. The optimal feature set was selected using Variance Inflation Factor (VIF) Boruta-XGBoost methods, which indicated elimination NO, NO2, NOx. highlighted PM10 as most important feature. Wavelet transform then applied extract 40 features enhance prediction accuracy. Hyperparameters weights matrices Echo State Network (ESN) were determined algorithms, with Salp Swarm Algorithm (SSA) demonstrating superior performance. evaluation different criteria revealed that ESN-SSA outperformed other hybrids original ESN, LSTM, GRU models.

Language: Английский

Citations

1
Effects of street plants on atmospheric particulate dispersion in urban streets: A review DOI
Xiaoshuang Wang, Zhixiang Zhou, Yang Xiang

et al.

Environmental Reviews, Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 26, 2024

Numerous empirical studies have demonstrated that street trees not only reduce dust pollution and absorb particulate matter (PM) but also improve microclimates, providing both ecological functions aesthetic value. However, recent research has revealed tree canopy cover can impede the dispersion of atmospheric PM within canyons, leading to accumulation pollutants. Although many investigated impact on air pollutant extent their influence remains unclear uncertain. Pollutant corresponds specific characteristics individual coupled with meteorological factors source strength. Notably, exert a significant influence. There is still quantitative gap impacts respect reduction control measures spaces. To urban traffic environments, policymakers mainly focused scientifically based vegetation deployment initiatives in building garden cities improving living environment. address uncertainties regarding streets, this study reviews mechanisms key summarizes approaches used conceptualize examines plant efficiency reducing PM. Furthermore, we current challenges future directions field provide more comprehensive understanding streets role play mitigating pollution.

Language: Английский

Citations

8
LGHAP v2: a global gap-free aerosol optical depth and PM2.5 concentration dataset since 2000 derived via big Earth data analytics DOI Creative Commons
Kaixu Bai, Ke Li,

Liuqing Shao

et al.

Earth system science data, Journal Year: 2024, Volume and Issue: 16(5), P. 2425 - 2448

Published: May 22, 2024

Abstract. The Long-term Gap-free High-resolution Air Pollutants (LGHAP) concentration dataset generated in our previous study has provided spatially contiguous daily aerosol optical depth (AOD) and fine particulate matter (PM2.5) concentrations at a 1 km grid resolution China since 2000. This advancement empowered unprecedented assessments of regional variations their influence on the environment, health, climate over past 20 years. However, there is need to enhance such high-quality AOD PM2.5 with new robust features extended spatial coverage. In this study, we present version 2 global-scale LGHAP (LGHAP v2), which was using improved big Earth data analytics via seamless integration versatile science, pattern recognition, machine learning methods. Specifically, multimodal AODs air quality measurements acquired from relevant satellites, ground monitoring stations, numerical models were harmonized by harnessing capability random-forest-based data-driven models. Subsequently, an tensor-flow-based reconstruction algorithm developed weave multisource products together for filling gaps Multi-Angle Implementation Atmospheric Correction (MAIAC) retrievals Terra. results ablation experiments demonstrated better performance gap-filling method terms both convergence speed accuracy. Ground-based validation indicated good accuracy global gap-free dataset, correlation coefficient (R) 0.85 root mean square error (RMSE) 0.14 compared worldwide observations AErosol RObotic NETwork (AERONET), outperforming purely reconstructed (R = 0.83, RMSE 0.15), but they slightly worse than raw MAIAC 0.88, 0.11). For mapping, novel deep-learning approach, termed SCene-Aware ensemble Graph ATtention network (SCAGAT), hereby applied. While accounting scene representativeness across regions, SCAGAT performed during extrapolation, largely reducing modeling biases regions limited and/or even absent situ measurements. that estimates exhibit higher prediction accuracies, R 0.95 5.7 µg m−3, obtained former holdout sites worldwide. Overall, while leveraging state-of-the-art methods science artificial intelligence, quality-enhanced v2 through cohesively weaving diverse sources. gap-free, high-resolution, coverage merits render invaluable database advancing aerosol- haze-related studies as well triggering multidisciplinary applications environmental management, health-risk assessment, change attribution. All grids user guide visualization codes, are publicly accessible https://zenodo.org/communities/ecnu_lghap (last access: 3 April 2024, Bai Li, 2023a).

Language: Английский

Citations

8
Evaluating urban and nonurban PM 2.5 variability under clean air actions in China during 2010–2022 based on a new high-quality dataset DOI Creative Commons
Boya Liu, Lili Wang, Lei Zhang

et al.

International Journal of Digital Earth, Journal Year: 2024, Volume and Issue: 17(1)

Published: Feb. 2, 2024

The air quality in China has changed due to the implementation of clean actions since 2013. Evaluating spatial pattern PM2.5 and effectiveness reducing anthropogenic emissions urban nonurban areas is crucial. Therefore, Long-term Air Pollutant dataset for (CLAP_PM2.5) was generated from 2010 2022 with a daily 0.1° resolution using random forest model integrating multiple data sources, including extensive in-situ measurements, visibility, satellite retrievals, surface upper-level meteorological other ancillary data. CLAP_PM2.5 more reliable accurate than public datasets. Analysis reveals decrease positive urban-nonurban differences higher decreasing rates most city clusters eastern China. Furthermore, separating emission contributions variability by normalization approach indicates that contribution gradually unfavorable reduction during 2013–2017 favorable decline enhancement 2018–2022, regions, areas. Overall, deweathered concentrations highlights China's significant achievements terms comprehensive actions.

Language: Английский

Citations

7
Towards long-term, high-accuracy, and continuous satellite total and fine-mode aerosol records: Enhanced Land General Aerosol (e-LaGA) retrieval algorithm for VIIRS DOI
Lunche Wang, Xin Su, Yi Wang

et al.

ISPRS Journal of Photogrammetry and Remote Sensing, Journal Year: 2024, Volume and Issue: 214, P. 261 - 281

Published: July 1, 2024

Language: Английский

Citations

5
Spatiotemporal patterns and quantitative analysis of influencing factors of PM2.5 and O3 pollution in the North China Plain DOI Open Access
Mingliang Ma, Mengnan Liu,

Song Xue-yan

et al.

Atmospheric Pollution Research, Journal Year: 2023, Volume and Issue: 15(1), P. 101950 - 101950

Published: Oct. 4, 2023

Language: Английский

Citations

12