A review of machine learning for modeling air quality: Overlooked but important issues

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

Published: Jan. 21, 2024

Language: Английский

---

Rasterizing CO2 emissions and characterizing their trends via an enhanced population-light index at multiple scales in China during 2013–2019

The Science of The Total Environment, Journal Year: 2023, Volume and Issue: 905, P. 167309 - 167309

Published: Sept. 22, 2023

Language: Английский

---

A comprehensive evaluation of deep learning approaches for ground-level ozone prediction across different regions

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103024 - 103024

Published: Jan. 1, 2025

Language: Английский

---

Identifying the spatiotemporal patterns and natural and socioeconomic influencing factors of PM2.5 and O3 pollution in China

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(2), P. e0317691 - e0317691

Published: Feb. 13, 2025

To promote collaborative governance of PM 2.5 and O 3 pollution, understanding their spatiotemporal patterns determining factors is crucial to control air pollution in China. Using the ground-monitored data encompassing concentrations 2019 across 337 Chinese cities, this study explores concentrations, then employed Multi-scale Geographically Weighted Regression (MGWR) model examine socioeconomic natural affecting or concentrations. The results show that exhibit distinct monthly U-shaped inverted temporal fluctuation respectively. Spatially, both pollutants manifest spatial clustering characteristic a certain degree bivariate correlation. Elevated are predominantly concentrated on north central China, as well Xinjiang Autonomous Region, whereas higher distributed widely north, east, northwest MGWR outperforms traditional OLS global regression models, evidenced by its enhanced goodness-of-fit metrics. Specifically, R 2 values for models notably high, at 0.842 0.861, Socioeconomic found have multi-scale effects On average, positively correlations with population density, proportion added value secondary industry GDP, wind speed, relative humidity, atmospheric pressure, but negatively relationship per capita road urban greening, temperature, precipitation, sunshine duration. In contrast, also associated energy consumption, duration, correlated temperature. Our findings offer valuable insights inform development comprehensive management policies developing countries.

Language: Английский

---

Innovation-Driven Cities: Reconciling economic growth and ecological sustainability

Sustainable Cities and Society, Journal Year: 2025, Volume and Issue: unknown, P. 106230 - 106230

Published: Feb. 1, 2025

Language: Английский

---

Health benefits from the rapid reduction in ambient exposure to air pollutants after China's clean air actions: progress in efficacy and geographic equality

National Science Review, Journal Year: 2023, Volume and Issue: 11(2)

Published: Oct. 9, 2023

ABSTRACT Clean air actions (CAAs) in China have been linked to considerable benefits public health. However, whether the beneficial effects of CAAs are equally distributed geographically is unknown. Using high-resolution maps distributions major pollutants (fine particulate matter [PM2.5] and ozone [O3]) population, we aimed track spatiotemporal changes health impacts from, geographic inequality embedded in, reduced exposures PM2.5 O3 from 2013 2020. We used a method established by Global Burden Diseases Study. By analyzing loss life expectancy (LLE) attributable O3, calculated gain (GLE) quantify air-quality improvement. Finally, assessed GLE using Gini index (GI). Based on risk assessments during first stage (2013 2017), mean was 1.87 months. Half sum disproportionally about one quarter population exposed (GI 0.44). During second (2017 2020), increased 3.94 months decreased 0.18). According our assessments, were enhanced, stages, terms not only preventing premature mortality but also ameliorating inequalities. The enhancements related sensitivity pollution synergic control levels. Our findings will contribute optimizing future CAAs.

Language: Английский

---

Full-coverage spatiotemporal estimation of surface ozone over China based on a high-efficiency deep learning model

International Journal of Applied Earth Observation and Geoinformation, Journal Year: 2023, Volume and Issue: 118, P. 103284 - 103284

Published: April 1, 2023

Ozone concentration Monitoring is essential to atmospheric pollution prevention and control. Against the background of severe ozone over China in recent years, a spatiotemporal contiguous mapping method was developed. We imputed significant data gaps Instrument's tropospheric NO2 content by using an efficient machine learning model named LightGBM. Then, we developed deep based on three-dimensional Convolutional Neural Network architecture for daily maximum 8 h average estimation China. With support satellite-retrieved precursor, meteorological other ancillary data, our achieved excellent performance with sample-based 10-fold cross-validation R2 = 0.88. Furthermore, generated datasets covering whole from 2016 2020. This study presents novel surface modeling, which can provide fundamental ecological changes caused pollution, such as crop loss, or harmful effects humans, increased incidence respiratory diseases.

Language: Английский

---

Quantitative Model Construction for Sustainable Security Patterns in Social–Ecological Links Using Remote Sensing and Machine Learning

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(15), P. 3837 - 3837

Published: Aug. 1, 2023

With the global issues of extreme climate and urbanization, ecological security patterns (ESPs) in Qinling Mountains are facing prominent challenges. As a crucial barrier China, understanding characteristics ESPs is vital for achieving sustainable development. This study focuses on Yangxian employs methods such as machine learning (ML), remote sensing (RS), geographic information systems (GISs), analytic hierarchy process principal component analysis (AHP–PCA), minimum cumulative resistance (MCR) model to construct an network based multi-factor sensitivity (ES) conduct quantitative spatial analysis. The results demonstrate that AHP–PCA method ML overcomes limitations single-weighting method. were established, consisting 21 main secondary sources with area 592.81 km2 (18.55%), 41 corridors length 738.85 km, 33 nodes. A coupling relationship among three dimensions was observed: comprehensive sensitivity, ESPs, administrative districts (ADs). Huangjinxia Town (1.43 C5) Huayang (7.28 C4) likely have significant areas vulnerability, while Machang Maoping important ESPs. ADs focus protection management. second corridor indicated high-quality construction, necessitating implementation strict policies area. innovation lies utilization methods, RS technologies, pattern planning propose new perspective space. provides foundation urban rural will help facilitate development region.

Language: Английский

---

Determination of major drive of ozone formation and improvement of O3 prediction in typical North China Plain based on interpretable random forest model

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 934, P. 173193 - 173193

Published: May 12, 2024

Language: Английский

---

Global systematical and comprehensive overview of mountainous flood risk under climate change and human activities

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 941, P. 173672 - 173672

Published: May 31, 2024

Language: Английский

---