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

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

---

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

---

The Trade-offs between Wildfires and Prescribed Fires: A Case Study for 2016 Gatlinburg Wildfires

ACS ES&T Air, Journal Year: 2025, Volume and Issue: 2(2), P. 236 - 248

Published: Jan. 9, 2025

Prescribed burning is an effective land management tool that provides a range of benefits, including ecosystem restoration and wildfire risk reduction. However, prescribed fires, just like wildfires, introduce smoke degrades air quality. Furthermore, while fires help manage risk, they do not eliminate the possibility wildfires. It therefore important to also evaluate fire impacts from wildfires may occur after burn. In this study, we developed framework for understanding quality health related trade-offs between by simulating set counterfactual scenarios postprescribed burn We applied case Gatlinburg found emissions burns subsequent were slightly lower than those itself. This reduction resulted in daily average concentrations exposures PM2.5, O3, NO2. Even considering wildfire, reduced population-weighted maximum 8-h 1-h NO2 concentrations. Sevier County, Tennessee where occurred, these reductions reached 5.28 μg/m3, 0.18 ppb, 1.68 respectively. The person-days wildfire. Our results suggest although cannot can greatly reduce exposure downwind areas distant sites.

Language: Английский

---

Wildland Fires Worsened Population Exposure to PM_2.5 Pollution in the Contiguous United States

Environmental Science & Technology, Journal Year: 2023, Volume and Issue: 57(48), P. 19990 - 19998

Published: Nov. 9, 2023

As wildland fires become more frequent and intense, fire smoke has significantly worsened the ambient air quality, posing greater health risks. To better understand impact of wildfire on we developed a modeling system to estimate daily PM2.5 concentrations attributed both nonsmoke sources across contiguous U.S. We found that most significant quality in West Coast, followed by Southeastern Between 2007 2018, contributed over 25% at ∼40% all regulatory monitors EPA's (AQS) for than one month per year. People residing outside vicinity an EPA AQS monitor (defined 5 km radius) were subject 36% days compared with those nearby. Lowering national standard (NAAQS) annual mean between 9 10 μg/m3 would result approximately 35–49% falling nonattainment areas, taking into account smoke. If contribution is excluded, this percentage be reduced 6 9%, demonstrating negative quality.

Language: Английский

---

High spatio-temporal resolution predictions of PM2.5 using low-cost sensor data

Atmospheric Environment, Journal Year: 2024, Volume and Issue: 326, P. 120486 - 120486

Published: March 26, 2024

We generated PM2.5 predictions at a high spatio-temporal resolution in the Columbus, OH, Denver, CO, and Pittsburgh, PA metropolitan areas using low-cost PurpleAir sensor data. used multiple modeling approaches, namely random forest (RF), spatial interpolation (RFSI), space-time regression kriging (STRK), (RFK). trained separate models for each combination of hour, month, city to predict concentrations 8 AM 6 PM on any specific day 100m. In most cases, that account relationships (e.g., STRK, RFK, RFSI) show better performance than non-spatio-temporal machine learning RF). On average, considering all cities, RFSI (mean MAE = 1.75, R2 0.67) STRK 1.74, 0.63) perform RFK 2.11, 0.59), has clearest patterns. found models, especially are superior capturing resemble generic land use pattern city, while effective when dealing with very large datasets missing cases. Our study demonstrates multi-model approach could inform deployment facilitate air quality modeling. high-resolution also studies short-term, traffic-based exposure assessment.

Language: Английский

---

PM2.5 Estimation using Machine Learning Models and Satellite Data: A Literature Review

International Journal of Advanced Computer Science and Applications, Journal Year: 2023, Volume and Issue: 14(5)

Published: Jan. 1, 2023

Most researchers are beginning to appreciate the use of remote sensing satellites assess PM2.5 levels and machine learning algorithms automate collection, make sense data, extract previously unseen data patterns. This study reviews delicate particulate matter (PM2.5) predictions from satellite aerosol optical depth (AOD) learning. Specifically, we review characteristics gap-filling methods satellite-based AOD products, sources components PM2.5, observable mining, application in publications past two years. The also included functional considerations recommendations covariate selection, addressing spatiotemporal heterogeneity -AOD relationship, cross-validation, aid determining final model. A total 79 articles were out 112 retrieved records consisting published 2022 totaling 43 articles, as 2023 (until February) 19 other years 18 articles. Finally, latest method works well for monthly estimates, while daily hourly can be achieved. is due increased availability computing power large datasets awareness potential benefits predictors working together achieve higher estimation accuracy. Some key findings presented conclusion section this article.

Language: Английский

---

A synchronized estimation of hourly surface concentrations of six criteria air pollutants with GEMS data

npj Climate and Atmospheric Science, Journal Year: 2023, Volume and Issue: 6(1)

Published: July 18, 2023

Abstract Machine learning is widely used to infer ground-level concentrations of air pollutants from satellite observations. However, a single pollutant commonly targeted in previous explorations, which would lead duplication efforts and ignoration interactions considering the interactive nature their common influencing factors. We aim build unified model offer synchronized estimation pollution levels. constructed multi-output random forest (MORF) achieved simultaneous hourly PM 2.5 , 10 O 3 NO 2 CO, SO China, benefiting world’s first geostationary air-quality monitoring instrument Geostationary Environment Monitoring Spectrometer. MORF yielded high accuracy with cross-validated R reaching 0.94. Meanwhile, efficiency was significantly improved compared single-output models. Based on retrieved results, spatial distributions, seasonality, diurnal variations six were analyzed two typical events tracked.

Language: Английский

---

Public Health Benefits From Improved Identification of Severe Air Pollution Events With Geostationary Satellite Data

GeoHealth, Journal Year: 2024, Volume and Issue: 8(1)

Published: Jan. 1, 2024

Abstract Despite improvements in ambient air quality the US recent decades, many people still experience unhealthy levels of pollution. At present, national‐level alert‐day identification relies predominately on surface monitor networks and forecasters. Satellite‐based estimates have rapidly advanced capability to inform exposure‐reducing actions protect public health. we lack a robust framework quantify health benefits these advances applications satellite‐based atmospheric composition data. Here, assess possible using geostationary satellite data, over polar orbiting for identifying particulate alert days (24hr PM 2.5 > 35 μg m −3 ) 2020. We find more extensive spatiotemporal coverage data leads 60% increase person‐alerts (alert × population) 2020 polar‐orbiting apply pre‐existing exposure reduction by individual behavior modification additional may lead 1,200 (800–1,500) or 54% averted ‐attributable premature deaths per year, if geostationary, instead orbiting, alone are used identify days. These an associated economic value 13 (8.8–17) billion dollars ($2019) year. Our results highlight one potential from satellites improving Identifying has important implications guiding use current planning future missions.

Language: Английский

---

A hybrid PM2.5 interval concentration prediction framework based on multi-factor interval decomposition reconstruction strategy and attention mechanism

Atmospheric Environment, Journal Year: 2024, Volume and Issue: 335, P. 120730 - 120730

Published: Aug. 7, 2024

Language: Английский

---

Quantitative Remote Sensing Supporting Deep Learning Target Identification: A Case Study of Wind Turbines

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(5), P. 733 - 733

Published: Feb. 20, 2025

Small Target Detection and Identification (TDI) methods for Remote Sensing (RS) images are mostly inherited from the deep learning models of Computer Vision (CV) field. Compared with natural images, RS not only have common features such as shape texture but also contain unique quantitative information spectral features. Therefore, TDI in CV field, which does use Quantitative (QRS) information, has potential to be explored. With rapid development high-resolution satellites, wind turbine detection become a key research topic power intelligent inspection. To test effectiveness integrating QRS models, case satellite was studied. The YOLOv5 model selected because its stability high real-time performance. following were proposed: (1) Surface reflectance (SR) obtained using Atmospheric Correction (AC) used make samples, SR data input into (YOLOv5_AC). (2) A Convolutional Block Attention Module (CBAM) added network focus on (YOLOv5_AC_CBAM). (3) Based identification results YOLOv5_AC_CBAM, spectral, geometric, textural expert knowledge extracted conduct threshold re-identification (YOLOv5_AC_CBAM_Exp). Accuracy increased 90.5% 92.7%, then 93.2%, finally 97.4%. integration showed tremendous achieve accuracy, should neglected TDI.

Language: Английский

---