Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM2.5 Using a Random Forest Approach

Handschuh, Jana; Erbertseder, Thilo; Baier, F. W.

doi:10.3390/rs15082064

Cited by 9 publications

(3 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, such data is not yet available to our knowledge for the entire BW and the time period under investigation. A possible addition to the variables considered could be wind speed, wind direction and boundary layer height since these parameters have shown a large impact on the variability of particulate matter 59 , 60 . Furthermore, boundary layer height and showed the strongest positive correlation among all the analyzed variables in 57 .…”

Section: Discussionmentioning

confidence: 99%

Investigating the spatiotemporal associations between meteorological conditions and air pollution in the federal state Baden-Württemberg (Germany)

Hoffmann,

Gilardi,

Schmitz

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

When analyzing health data in relation to environmental stressors, it is crucial to identify which variables to include in the statistical model to exclude dependencies among the variables. Four meteorological parameters: temperature, ultraviolet radiation, precipitation, and vapor pressure and four outdoor air pollution parameters: ozone ($$\text{O}_3$$ O 3 ), nitrogen dioxide ($$\text{NO}_2$$ NO 2 ), particulate matter ($$PM_{2.5}$$ P M 2.5 , $$PM_{10}$$ P M 10 ) were studied on a daily basis for Baden-Württemberg (Germany). This federal state covers urban and rural compartments including mountainous and river areas. A temporal and spatial analysis of the internal relationships was performed among the variables using (a) cross-correlations, both on the grand ensemble of data as well as within subsets, and (b) the Local Indications of Spatial Association (LISA) method. Meteorological and air pollution variables were strongly correlated within and among themselves in time and space. We found a strong interaction between nitrogen dioxide and ozone, with correlation coefficients varying over time. The coefficients ranged from negative correlations in January (−0.84), April (−0.47), and October (−0.54) to a positive correlation in July (0.45). The cross-correlation plot showed a noticeable change in the correlation direction for $$\text{O}_3$$ O 3 and $$\text{NO}_2$$ NO 2 . Spatially, $$\text{NO}_2$$ NO 2 , $$PM_{2.5}$$ P M 2.5 , and $$PM_{10}$$ P M 10 concentrations were significantly higher in urban than rural regions. For $$\text{O}_3$$ O 3 , this effect was reversed. A LISA analysis confirmed distinct hot and cold spots of environmental stressors. This work examined and quantified the spatio-temporal relationship between air pollution and meteorological conditions and recommended which variables to prioritize for future health impact analyses. The results found are in line with the underlying physico-chemical atmospheric processes. It also identified postal code areas with dominant environmental stressors for further studies.

show abstract

Section: Discussionmentioning

confidence: 99%

Investigating the spatiotemporal associations between meteorological conditions and air pollution in the federal state Baden-Württemberg (Germany)

Hoffmann,

Gilardi,

Schmitz

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…On the other hand, the use of satellite data and atmospheric reanalysis models provides estimates on a global scale and continuous coverage, including remote and poorly monitored areas (Handschuh;Erbertseder;Baier, 2023). In this context, by combining satellite data, in situ measurements and atmospheric models, the Copernicus Atmosphere Monitoring Service (CAMS) plays a crucial role in providing accurate estimates of AOD on a global scale (Garrigues et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

Evaluation of Aerosol Optical Depth (Aod) Estimated by Copernicus Atmosphere Monitoring Service (Cams) in Brazil

Júnior,

Curado,

Palácios

et al. 2024

Preprint

View full text Add to dashboard Cite

The optical depth of aerosols plays a crucial role in scientific research and environmental policies, making it possible to understand the distribution and extent of aerosols in various regions of Brazil. The objective of this study is to evaluate the estimates of Aerosol Optical Depth (AOD) from the Copernicus Atmosphere Monitoring Service (CAMS) product in Brazil. The study covered the sites of Alta Floresta, Ji-Paraná, Rio Branco, Manaus, ATTO, São Paulo-EACH, São Paulo, Itajubá, Cuiabá, São Martinho, Petrolina and Campo Grande. Measured and estimated values were evaluated using Pearson correlation index "r", accuracy using Willmott index "d", Mean Squared Error, Mean Absolute Error and Percentage Bias. Results from the CAMS product showed good agreement with AOD measurements from the Aerosol Robotic Network. There was a strong correlation between the data, with Willmott index "d" values close to 1 and relatively low errors. However, significant differences were observed in some sites, such as Ji-Paraná, Rio Branco, Manaus and ATTO, where the CAMS tended to overestimate the AOD, while in Petrolina there was an underestimation. Variations in AOD occurred in various regions of Brazil over the years analyzed, with an increase during the dry season due to fires and human activities, and a reduction during the rainy months. The areas most affected were those close to the arc of deforestation in the Amazon. Aerosol concentrations have also been influenced by climatic factors, agricultural, industrial and urban activities in different regions of the country. This variability highlights the complexity of the natural and anthropogenic factors that affect air quality and emphasizes the importance of control and mitigation strategies for aerosol emissions. Therefore, the CAMS has demonstrated satisfactory performance in estimating the AOD in Brazil, providing valuable information on aerosol concentrations.

show abstract

“…Unlike the works in Refs. 20 to 22 that mostly exploit single-platformed satellite AOD products, a systematic evaluation is performed using RF approach 23 for estimating PM 2.5 using four different satellite AOD datasets. Another deep-learning-based forecasting technique, long short-term memory (LSTM), demonstrated its capability to determine the wind forecasts accurately for predicting solar energy using satellite data.…”

Section: Introductionmentioning

confidence: 99%

M2-APNet: A multimodal deep learning network to predict major air pollutants from temporal satellite images

Swetha,

Datla,

Vishnu

et al. 2023

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

Air quality monitoring plays a vital role in the sustainable development of any country. Continuous monitoring of the major air pollutants and forecasting their variations would be helpful in saving the environment and improving the quality of public health. However, this task becomes challenging with the available observations of air pollutants from the on-ground instruments with their limited spatial coverage. We propose a multimodal deep learning network (M 2 -APNet) to predict major air pollutants at a global scale from multimodal temporal satellite images. The inputs to the proposed M 2 -APNet include satellite image, digital elevation model (DEM), and other key attributes. The proposed M 2 -APNet employs a convolutional neural network to extract local features and a bidirectional long short-term memory to obtain longitudinal features from multimodal temporal data. These features are fused to uncover common patterns helpful for regression in predicting the major air pollutants and categorization of air quality index (AQI). We have conducted exhaustive experiments to predict air pollutants and AQI across important regions in India by employing multiple temporal modalities. Further, the experimental results demonstrate the effectiveness of DEM modality over others in learning to predict major air pollutants and determining the AQI.

show abstract

Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM2.5 Using a Random Forest Approach

Cited by 9 publications

References 72 publications

Investigating the spatiotemporal associations between meteorological conditions and air pollution in the federal state Baden-Württemberg (Germany)

Investigating the spatiotemporal associations between meteorological conditions and air pollution in the federal state Baden-Württemberg (Germany)

Evaluation of Aerosol Optical Depth (Aod) Estimated by Copernicus Atmosphere Monitoring Service (Cams) in Brazil

M2-APNet: A multimodal deep learning network to predict major air pollutants from temporal satellite images

Contact Info

Product

Resources

About