Comparing CMAQ Forecasts with a Neural Network Forecast Model for PM2.5 in New York

Quantifying the impacts of Canadian wildfires on regional air pollution networks

Wildfire smoke greatly impacts regional atmospheric systems, causing changes in the behavior of pollution. However, the impacts of wildfire smoke on pollution behavior are not easily quantifiable due to the complex nature of atmospheric systems. Air pollution correlation networks have been used to quantify air pollution behavior during ambient conditions. However, it is unknown how extreme pollution events impact these networks. Therefore, we propose a multidimensional air pollution correlation network framework to quantify the impacts of wildfires on air pollution behavior. The impacts are quantified by comparing two time periods, one during the 2023 Canadian wildfires and one during normal conditions with two complex network types for each period. In this study, the value network represents PM2.5 concentrations and the rate network represents the rate of change of PM2.5 concentrations. Wildfires' impacts on air pollution behavior are captured by structural changes in the networks. The wildfires caused a discontinuous phase transition during percolation in both network types which represents non-random organization of the most significant spatiotemporal correlations. Additionally, wildfires caused changes to the connectivity of stations leading to more interconnected networks with different influential stations. During the wildfire period, highly polluted areas are more likely to form connections in the network, quantified by an 86 % and 19 % increase in the connectivity of the value and rate networks respectively compared to the normal period. In this study, we create novel understandings of the impacts of wildfires on air pollution correlation networks, show how our method can create important insights into air pollution patterns, and discuss potential applications of our methodologies. This study aims to enhance capabilities for wildfire smoke exposure mitigation and response strategies.

24-hour average PM2.5 concentration caused by aircraft in Chinese airports from Jan. 2006 to Dec. 2023

Since 2006, the rapid development of China's aviation industry has been accompanied by a significant increase in one of its emissions, namely, PM2.5, which poses a substantial threat to human health. However, little data is describing the PM2.5 concentration caused by aircraft activities. This study addresses this gap by initially computing the monthly PM2.5 emissions of the landing-take-off (LTO) stage from Jan. 2006 to Dec. 2023 for 175 Chinese airports, employing the modified BFFM2-FOA-FPM method. Subsequently, the study uses the Gaussian diffusion model to measure the 24-hour average PM2.5 concentration resulting from flight activities at each airport. This study mainly draws the following conclusions: Between 2006 and 2023, the highest recorded PM2.5 concentration data at all airports was observed in 2018, reaching 5.7985 micrograms per cubic meter, while the lowest point was recorded in 2022, at 2.0574 micrograms per cubic meter. Moreover, airports with higher emissions are predominantly located in densely populated and economically vibrant regions such as Beijing, Shanghai, Guangzhou, Chengdu, and Shenzhen.

Hourly forecasting on PM2.5 concentrations using a deep neural network with meteorology inputs

The PM2.5 (particulate matter with a diameter of fewer than 2.5 µm) has become a global topic in environmental science. The neural network that based on the non-linear regression algorithm, e.g., deep learning, is now believed to be one of the most facile and advanced approaches in PM2.5 concentration prediction. In this study, we proposed a PM2.5 predictor using deep learning as infrastructure and meteorological data as input, for predicting the next hour PM2.5 concentration in Beijing Aotizhongxin monitor point. We efficiently use the parameter's spatiotemporal correlation by concatenating the dataset with time series. The predicted PM2.5 concentration was based on meteorology changes over a period. Therefore, the accuracy would increase with the period growing. By extracting the intrinsic features between meteorological and PM2.5 concentration, a fast and accurate prediction was carried out. The R square score reached maximum of 0.98 and remained an average of 0.9295 in the whole test. The average bias of the model is 9 μg on the validation set and 1 μg on the training set. Moreover, the differences between the predictions and expectations can be further regarded as the estimation for the emission change. Such results can provide scientific advice to supervisory and policy workers.

Prediction of glaciated area fraction over the Sikkim Himalayan Region, India: a comparative study using response surface method, random forest, and artificial neural network

Glacier area fraction at high altitude mountains is a serious worry in today's time triggered by climate change. The current information on this natural resource is very important for the survival of humanity as it affects the water, food, and energy security of people dependent on it. Due to its problematic accessibility and tough environmental condition, ground monitoring is quite challenging. This study investigates the impact of environmental parameters and pollutants on glacier area fraction over the Eastern Himalaya region and its prediction through random forest (RF), multilayer perceptron (MLP), radial basis function analysis (RBFN), and response surface methodology (RSM) models. The data are obtained from the Goddard Earth Sciences Data and Information Services Center (GES DISC), NASA's data archive portal ( https://giovanni.gsfc.nasa.gov ). The collinearity of independent variables reveals that all selected input parameters are highly correlated with R2 value > 0.9. The RSM and RF model provided valuable insight of the predictor's significance in addition to their capability to predict the response. The model performance was evaluated in terms of R2 value and the error matrices. The model's R2 value was found to be 0.843, 0.839, 0.838, and 0.743 for MLP, RBFN, RF, and RSM respectively. Although, the neural network model R2 values are the highest, but the most reliable and suitable model is RF as the error matrices for this model are much lower than others. This study encourages the investigation of the hybridization of these models for more accurate prediction.

Machine learning-based Decision Tree J48 with grey wolf optimizer for environmental pollution control

ABSTRACTDue to industrialization, activities of human and urbanization, environment is getting polluted. Air pollution has become a main issue in the metropolitan areas of the world. To protect people from diseases, monitoring air quality plays an important thing. This air pollutant may lead to many health issues like respiratory and cardiac problems. The major air pollutants are NO, C₆H₆, CO, etc. Many research works have been done in predicting air pollution-based health issues, predicting air pollution levels, monitoring and controlling the polluted levels. But they are not efficient, cost of maintenance is high and insufficient tool for monitoring it. To overcome these issues, this paper implements hybrid algorithm of Decision Tree J48 and Grey Wolf Optimizer (DT-GWO). This DT-GWO is a better model to addresses the predicting of Air Quality Index (AQI), which minimizes the error rate, accurately and effectively predicting the air quality. The AQI values are categorised as good, moderate, unhealthy, very unhealthy and hazardous. The dataset used in this work is collected from Kaggle website which contains air pollutants details with air quality index values. Accuracy obtained for decision Tree J48 is 93.72%, grey wolf optimizer is 96.83% and our proposed work DT-GWO is 99.78%.

A new hybrid models based on the neural network and discrete wavelet transform to identify the CHIMERE model limitation

A greater understanding of ozone damage to the environment and health led to an increased demand for accurate predictions. This study provides two new accurate hybrid models of ozone prediction. The first one (CHIMERE-NARX) is based on a NARX model as a post-processing of the CHIMERE model. In the second (CHIMERE-NARX-DWT), a discrete wavelet transform (DWT) has been added. Our models were built and validated using ozone measurements from the Mediouna station in Casablanca, Morocco, from February 1^st to March 27^th, 2021. The results highlighted the CHIMERE model limitations, such as wind speed overestimation and insufficient emission data. The first hybrid successfully increased the correlation coefficient from 88 to 93% and reduced RMSE from 23.99 μg/m³ to -3.54 μg/m³, overcoming CHIMERE limitations to some extent, especially during nighttime. A second hybrid addressed the first hybrid limitation, such as using ozone as a single input. This hybrid successfully balanced the weight of NARX at night against the day, increasing the correlation coefficient to 98% and decreasing RMSE to -0.02 μg/m³. This study presents a new generation of post-processing based on deterministic model processes, with the possibility of training them with minimum input data, which can be applied to other models using various pollutants.

Extraction of multi-scale features enhances the deep learning-based daily PM2.5 forecasting in cities

Characterising the daily PM2.5 concentration is crucial for air quality control. To govern the status of the atmospheric environment, a novel hybrid model for PM2.5 forecasting was proposed by introducing a two-stage decomposition technology of complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and variational mode decomposition (VMD); subsequently, a deep learning approach of long short-term memory (LSTM) was proposed. Five cities with unique meteorological and economic characteristics were selected to assess the predictive ability of the proposed model. The results revealed that PM2.5 pollution was generally more severe in inland cities (66.98 ± 0.76 μg m^-3) than in coastal cities (40.46 ± 0.40 μg m-3). The modelling comparison showed that in each city, the secondary decomposition algorithm improved the accuracy and prediction stability of the prediction models. When compared with other prediction models, LSTM effectively extracted featured information and achieved relatively accurate time-series prediction. The hybrid model of CEEMDAN-VMD-LSTM achieved a better prediction in the five cities (R2 = 0.9803 ± 0.01) compared with the benchmark models (R2 = 0.7537 ± 0.03). The results indicate that the proposed approach can identify the inherent correlations and patterns among complex datasets, particularly in time-series analysis.

A novel PM2.5 concentrations probability density prediction model combines the least absolute shrinkage and selection operator with quantile regression

PM2.5 has a significant negative impact on human health and atmospheric quality, and accurate prediction of its concentration is necessary. When using common point prediction models for PM2.5 concentration prediction, the influence of various uncertainties on PM2.5 concentrations makes the prediction results suffer from poor accuracy. To address this issue, this paper proposes the quantile regression neural network (QRNN) model based on the least absolute shrinkage and selection operator (LASSO), combined with kernel density estimation (KDE) for probabilistic density prediction of PM2.5 concentrations. The model uses LASSO regression to select the influential factors, and then the quartiles of daily PM2.5 concentrations calculated by the QRNN model are imported into the KDE model to obtain the probability density predictions of PM2.5 concentrations. In the paper, empirical analyses are carried out with the cities of Beijing and Jinan in China as well as six other datasets, and the prediction performance of the model is assessed by using evaluation criteria in both point prediction and interval prediction. The simulation reveals that the predictive performance of the LASSO-QRNN-KDE model is well, and the model is not only effective in filtering high-dimensional data, but also has a higher accuracy compared to common research models. In addition, the model is able to describe the uncertainty of PM2.5 concentration fluctuations and carry more information on the variation of PM2.5 concentrations, which can provide a novel and excellent PM2.5 concentration prediction tool for relevant policy makers.

An ensemble LSTM-based AQI forecasting model with decomposition-reconstruction technique via CEEMDAN and fuzzy entropy

Air quality affects people's daily life. Air quality index (AQI) is an essential indicator for controlling air pollution and ensuring public health, whose accurate forecasting can provide timely air pollution warnings and remind people to take protective measures against air pollution in advance. To address this issue, this paper developed a new ensemble learning model for AQI forecasting. In this study, (1) the signal decomposition technique complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) is introduced to decompose the nonlinear and nonstationary AQI history data series into several more regular and more stable subseries firstly. (2) Fuzzy entropy (FE) is selected as the feature indicator to recombine the subseries with similar trends to avoid the problem of over-decomposition and reduce the computing time. (3) An ensemble long short-term memory (LSTM) neural network is established to forecast each reconstructed subseries, whose values are superimposed to predict the AQI value eventually. To validate the predicting performance of the proposed model, daily AQI data of Wuhan, China, dating from January 1, 2019, to February 28, 2022, is used as the experiment case. And comparative analysis is made between the proposed model and other common-used forecasting models. Benchmarking results of the numerical study demonstrate that the proposed model is superior to the other forecasting models with better AQI prediction accuracy.

A new combination model using decomposition ensemble framework and error correction technique for forecasting hourly PM2.5 concentration

PM_2.5 pollutants are seriously harmful to human health, which is of great significance for the forecasting of PM_2.5 concentration. To accurately forecast hourly PM_2.5 concentration, a new combination model based on agreement index variational mode decomposition (AIVMD), radial basis function neural network (RBF), induced ordered weighted averaging (IOWA) operator, long short-term memory neural network (LSTM) and error correction (EC), named AIVMD-RBF-IOWA-LSTM-EC, is proposed, which uses decomposition ensemble framework and error correction technique. Taking the reduction of reconstruction error in the process of VMD as the goal, an adaptive method to determine the mode number of VMD by agreement index (AI), named AIVMD, is proposed. Firstly, PM_2.5 concentration data are decomposed into simple intrinsic mode function components (IMFs) by AIVMD to reduce the complexity of the data. Secondly, LSTM and RBF models are established for each IMF component, and the prediction results of each model are combined separately. Thirdly, an error correction model based on RBF is established to correct the prediction results. The predicted values of error are not only used to correct the prediction results, but also can be used as the induced value of IOWA operators to solve the weight allocation problem. Finally, the IOWA operator is used to weight the error correction prediction results, and the final result is obtained. To solve the problem that the forecasting accuracy of the combination model based on IOWA operators is low when the complementarity between single models is poor, a combination forecasting method with complementary disadvantage based on IOWA operators is proposed, which effectively improves the robustness of the model. A formula for calculating the proportion of complementary points is given. By solving the formula, the complementarity of the models can be judged, and the method of calculating the weight of the combined model can be selected accordingly. The proposed model is used to forecast PM_2.5 concentration in Xi'an, and compared with the predicted results of contrast models. The results show that the proposed model has a great advantage in short-term forecasting of PM_2.5 concentration.

A BP Neural Network Algorithm for Multimedia Data Monitoring of Air Particulate Matter

In order to study a BP neural network algorithm for air particulate matter data monitoring, firstly, the monitoring data collected by particle sensor using the light scattering method are proposed. Then, based on the improved BP neural network method, the mapping relationship between the actual measured value of the sensor, weather and other influencing factors, and the standard value of the monitoring station is established, and the calibration model of air particulate matter is realized. Finally, through theoretical analysis and experimental comparison, the results show that the model based on BP neural network algorithm has good accuracy and generalization ability in the evaluation of air particulate index, which makes it possible to scientifically and accurately refine the evaluation and management of urban air particulate index. The experimental results show that the air particle calibration model based on the light scattering method and improved BP neural network algorithm is practical and effective.

Combining Machine Learning and Numerical Simulation for High-Resolution PM2.5 Concentration Forecast

Forecasting ambient PM_2.5 concentrations with spatiotemporal coverage is key to alerting decision makers of pollution episodes and preventing detrimental public exposure, especially in regions with limited ground air monitoring stations. The existing methods rely on either chemical transport models (CTMs) to forecast spatial distribution of PM_2.5 with nontrivial uncertainty or statistical algorithms to forecast PM_2.5 concentration time series at air monitoring locations without continuous spatial coverage. In this study, we developed a PM_2.5 forecast framework by combining the robust Random Forest algorithm with a publicly accessible global CTM forecast product, NASA's Goddard Earth Observing System "Composition Forecasting" (GEOS-CF), providing spatiotemporally continuous PM_2.5 concentration forecasts for the next 5 days at a 1 km spatial resolution. Our forecast experiment was conducted for a region in Central China including the populous and polluted Fenwei Plain. The forecast for the next 2 days had an overall validation R² of 0.76 and 0.64, respectively; the R² was around 0.5 for the following 3 forecast days. Spatial cross-validation showed similar validation metrics. Our forecast model, with a validation normalized mean bias close to 0, substantially reduced the large biases in GEOS-CF. The proposed framework requires minimal computational resources compared to running CTMs at urban scales, enabling near-real-time PM_2.5 forecast in resource-restricted environments.

Development and Assessment of Spatially Continuous Predictive Algorithms for Fine Particulate Matter in New York State

Health risks connected with fine particulate matter (PM2.5) pollutants are well documented; increased risks of asthma, heart attack and heart failure are a few of the effects associated with PM2.5. Accurately forecasting PM2.5 is crucial for state agencies directed to devise State Implementation Plans (SIPS) to deal with National Ambient Air Quality Standards (NAAQS) exceedances. In previous work, we explored the application of multi-temporal data-driven neural networks (NNs) to forecasting PM2.5. Our work showed that under different input conditions, the NN approach achieves higher forecasting scores for local (12 km) resolution when compared to the other Chemical Transport Model forecast models, such as the Community Multi-Scale Air Quality system (CMAQ). Critical to our approach was the inclusion of prior PM2.5 concentrations, retrieved from ground monitoring stations, as part of the input dataset for the NN. The NN approach can provide high-level forecasting accuracy; however, because of the dependency on ground monitoring stations, the forecast coverage is sparse. Here, we extend our previous station-specific efforts by forecasting hourly PM2.5 values that are spatially continuous through the use of a deep neural network (DNN). The DNN approach combines spatial Kriging with additional local source variables to interpolate the measured PM2.5 concentrations across non-station locations. These interpolated PM2.5 values are used as inputs in the original forecasting NN. Cross-validation testing, using all New York State AirNow PM2.5 stations, showed that this forecast approach achieves accurate results, with a regression coefficient (R2) of 0.59, and a root mean square error (RMSE) of 2.22μgm3. Additionally, herein we demonstrate the usefulness of this approach on specific temporal events where significant dynamics of PM2.5 were observed; particularly, we show that even bias-corrected CMAQ forecasts do not track these transients and our NN method.

Wildfire and prescribed burning impacts on air quality in the United States

Air quality impacts from wildfires have been dramatic in recent years, with millions of people exposed to elevated and sometimes hazardous fine particulate matter (PM _2.5 ) concentrations for extended periods. Fires emit particulate matter (PM) and gaseous compounds that can negatively impact human health and reduce visibility. While the overall trend in U.S. air quality has been improving for decades, largely due to implementation of the Clean Air Act, seasonal wildfires threaten to undo this in some regions of the United States. Our understanding of the health effects of smoke is growing with regard to respiratory and cardiovascular consequences and mortality. The costs of these health outcomes can exceed the billions already spent on wildfire suppression. In this critical review, we examine each of the processes that influence wildland fires and the effects of fires, including the natural role of wildland fire, forest management, ignitions, emissions, transport, chemistry, and human health impacts. We highlight key data gaps and examine the complexity and scope and scale of fire occurrence, estimated emissions, and resulting effects on regional air quality across the United States. The goal is to clarify which areas are well understood and which need more study. We conclude with a set of recommendations for future research.

IMPLICATIONS

In the recent decade the area of wildfires in the United States has increased dramatically and the resulting smoke has exposed millions of people to unhealthy air quality. In this critical review we examine the key factors and impacts from fires including natural role of wildland fire, forest management, ignitions, emissions, transport, chemistry and human health.

Air Quality Index and Air Pollutant Concentration Prediction Based on Machine Learning Algorithms

Air pollution has become an important environmental issue in recent decades. Forecasts of air quality play an important role in warning people about and controlling air pollution. We used support vector regression (SVR) and random forest regression (RFR) to build regression models for predicting the Air Quality Index (AQI) in Beijing and the nitrogen oxides (NOX) concentration in an Italian city, based on two publicly available datasets. The root-mean-square error (RMSE), correlation coefficient (r), and coefficient of determination (R2) were used to evaluate the performance of the regression models. Experimental results showed that the SVR-based model performed better in the prediction of the AQI (RMSE = 7.666, R2 = 0.9776, and r = 0.9887), and the RFR-based model performed better in the prediction of the NOX concentration (RMSE = 83.6716, R2 = 0.8401, and r = 0.9180). This work also illustrates that combining machine learning with air quality prediction is an efficient and convenient way to solve some related environment problems.

Machine Learning Approach To Estimate Hourly Exposure to Fine Particulate Matter for Urban, Rural, and Remote Populations during Wildfire Seasons

Exposure to wildfire smoke averaged over 24-hour periods has been associated with a wide range of acute cardiopulmonary events, but little is known about the effects of sub-daily exposures immediately preceding these events. One challenge for studying sub-daily effects is the lack of spatially and temporally resolved estimates of smoke exposures. Inexpensive and globally applicable tools to reliably estimate exposure are needed. Here we describe a Random Forests machine learning approach to estimate 1-hour average population exposure to fine particulate matter during wildfire seasons from 2010 to 2015 in British Columbia, Canada, at a 5 km × 5 km resolution. The model uses remotely sensed fire activity, meteorology assimilated from multiple data sources, and geographic/ecological information. Compared with observations, model predictions had a correlation of 0.93, root mean squared error of 3.2 μg/m³, mean fractional bias of 15.1%, and mean fractional error of 44.7%. Spatial cross-validation indicated an overall correlation of 0.60, with an interquartile range from 0.48 to 0.70 across monitors. This model can be adapted for global use, even in locations without air quality monitoring. It is useful for epidemiologic studies on sub-daily exposure to wildfire smoke and for informing public health actions if operationalized in near-real-time.