Prediction Of PM10 Concentration Research Articles

Prediction of PM10 concentration in Seoul, Korea using Bayesian network

Prediction of PM10 concentration in Seoul, Korea using Bayesian network

Prediction of PM10 Concentration in Malaysia Using K-Means Clustering and LSTM Hybrid Model

Following the rapid development of various industrial sectors, air pollution frequently occurs in every corner of the world. As a dominant pollutant in Malaysia, particulate matter PM10 can cause highly detrimental effects on human health. This study aims to predict the daily average concentration of PM10 based on the data collected from 60 air quality monitoring stations in Malaysia. Building a forecasting model for each station is time-consuming and unrealistic; therefore, a hybrid model that combines the k-means clustering technique and the long short-term memory (LSTM) model is proposed to reduce the number of models and the overall model training time. Based on the training set, the stations were clustered using the k-means algorithm and an LSTM model was built for each cluster. Then, the prediction performance of the hybrid model was compared with the univariate LSTM model built independently for each station. The results show that the hybrid model has a comparable prediction performance to the univariate LSTM model, as it gives the relative percentage difference (RPD) less than or equal to 50% based on at least two accuracy metrics for 43 stations. The hybrid model can also fit the actual data trend well with a much shorter training time. Hence, the hybrid model is more competitive and suitable for real applications to forecast air quality.

Comparative assessment of ground-level air quality in the metropolitan area of Prague using local street canyon modelling

Metropolitan areas have recently been identified as hotspots in terms of air quality and pollution-related diseases caused by gaseous pollutants and particulate matter primarily from road emissions. This study was carried out to investigate the suitability of the air dispersion modelling for predicting the short-term concentration of selected pollutants within the asymmetric street canyon geometries. The geographic projection covered a modelling domain of 0.5 × 1.4 km, with a grid resolution of 2 m mapped to the receptor site of an operational ambient air quality monitoring station in Prague, Czech Republic. Basic street canyon simulation showed the best performance metric for the prediction of PM10 concentration. High spatiotemporal performance was found by the advanced street canyon simulation in predicting hourly averages of NOx and PM10. The underprediction of the modelled NOx concentration corresponded to those of previous studies with FB value of 0.11 and NMSE <4. The slightly overprediction of PM10 concentration with the obtained FB value of −0.19, the NMSE value of 0.04, and the FAC2 of 0.53 were in agreement with previous modelling work. The simulation results provided additional evidence for the underpredicted concentration of PM2.5 in different model configurations. Advanced street canyon simulation improves the accuracy of ADMS-Urban by a maximum of 34%. The underprediction of short-term emission was graphically analysed in time-series profiles. The validation results and significant remarks greatly contribute to the essential optimization of urban street canyon model to better represent observations in the street canyon geometries of historic European building architectures.

Performance of Bayesian Model Averaging (BMA) for Short-Term Prediction of PM10 Concentration in the Peninsular Malaysia

In preparation for the Fourth Industrial Revolution (IR 4.0) in Malaysia, the government envisions a path to environmental sustainability and an improvement in air quality. Air quality measurements were initiated in different backgrounds including urban, suburban, industrial and rural to detect any significant changes in air quality parameters. Due to the dynamic nature of the weather, geographical location and anthropogenic sources, many uncertainties must be considered when dealing with air pollution data. In recent years, the Bayesian approach to fitting statistical models has gained more popularity due to its alternative modelling strategy that accounted for uncertainties for all air quality parameters. Therefore, this study aims to evaluate the performance of Bayesian Model Averaging (BMA) in predicting the next-day PM10 concentration in Peninsular Malaysia. A case study utilized seventeen years’ worth of air quality monitoring data from nine (9) monitoring stations located in Peninsular Malaysia, using eight air quality parameters, i.e., PM10, NO2, SO2, CO, O3, temperature, relative humidity and wind speed. The performances of the next-day PM10 prediction were calculated using five models’ performance evaluators, namely Coefficient of Determination (R2), Index of Agreement (IA), Kling-Gupta efficiency (KGE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). The BMA models indicate that relative humidity, wind speed and PM10 contributed the most to the prediction model for the majority of stations with (R2 = 0.752 at Pasir Gudang monitoring station), (R2 = 0.749 at Larkin monitoring station), (R2 = 0.703 at Kota Bharu monitoring station), (R2 = 0.696 at Kangar monitoring station) and (R2 = 0.692 at Jerantut monitoring station), respectively. Furthermore, the BMA models demonstrated a good prediction model performance, with IA ranging from 0.84 to 0.91, R2 ranging from 0.64 to 0.75 and KGE ranging from 0.61 to 0.74 for all monitoring stations. According to the results of the investigation, BMA should be utilised in research and forecasting operations pertaining to environmental issues such as air pollution. From this study, BMA is recommended as one of the prediction tools for forecasting air pollution concentration, especially particulate matter level.

Enhancing Air Quality Prediction Accuracy Using Hybrid Deep Learning

PM2.5 (Particulate Matter) and PM10 are the most common pollutants, and the increasing of concentration in the air will threaten people’s health. The machine learning method has recently been of particular interest to many researchers due to its effectiveness in air quality prediction models. Many solutions employing deep learning-based techniques such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and hybrid CNN-LSTM models to enhance air quality prediction accuracy have been developed. This paper proposes a hybrid Encoder STM model for predicting the next day to the next five days’ PM2.5 and PM10 concentrations in Hanoi. Additionally, we proposed five extended features to increase the accuracy of prediction. Then other models, namely the LSTM model and the Bidirectional LSTM model, are also considered for PM2.5 and PM10 concentration prediction. Our results show that the proposed approaches outperform other state-of-the-art deep learning-based methods on both Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) due to low error and the small number of features.

Effect of Monsoonal Clustering for PM10 Concentration Prediction in Keningau, Sabah using Principal Component Analysis

Particulate matter (PM) has caught scientific attention in scientific research due to its harmful effect on human health. While prediction is essential for future development in Keningau, temporal clustering in Keningau has yet to be studied. Thus, this research aims to determine whether monsoonal clustering is required for meteorological and pollutant concentration data collected in Keningau. Missing data is first imputed using Nearest Neighbour Method (NNM). Then, wind direction and wind speed are converted into northern (Wy ) and eastern (Wx ) component of wind speed. Data is then temporal clustered based on monsoonal season (NEM, IM4, SWM, IM10). Both clustered and unclustered data are analysed using principal component (PC) analysis (PCA). The findings revealed that humidity in PC1 with average EV (explained variation) of 93.92 ± 0.52 contribute the most variation of PM10, followed by Wx in PC2 with average EV of 3.51 ± 0.48. Regression analysis shows that humidity and PM10 are negatively moderate to strongly correlated except for IM4 (intermonsoon April), which may be due to dry climate during the season. As for Wx, it has weak correlation with PM10. This may be due to location of Keningau at western part of Crocker range. However, the spread of PM10 due to eastern wind causes weak to zero correlation. Due to consideration of dry climate as revealed by the findings from IM4 cluster, there is need for data collected by Keningau to be clustered by monsoon.

Prediction of Daily Air Pollutants Concentration and Air Pollutant Index Using Machine Learning Approach

The major air pollutants in Malaysia that contribute to air pollution are carbon monoxide, sulfur dioxide, nitrogen dioxide, ozone, and particulate matter. Predicting the air pollutants concentration can help the government to monitor air quality and provide awareness to the public. Therefore, this study aims to overcome the problem by predicting the air pollutants concentration for the next day. This study focuses on an industrial, the Petaling Jaya monitoring station in Selangor. The data is obtained from the Department of Environment, which contains the dataset from 2004 to 2018. Subsequently, this study is conducted to construct predictive modeling that can predict the air pollutants concentrations for the next day using a tree-based approach. From the comparison of the three models, a random forest is a best-proposed model. The results of PM10 concentration prediction for the random forest is the best performance which is shown by RMSE (15.7611–19.0153), NAE (0.6508–0.8216), and R2 (0.346–0.5911). For SO2, the RMSE was 0.0016–0.0017, the NAE was 0.7056–0.8052, and the R2 was 0.3219–0.4676. The RMSE (0.0062–0.0075), the NAE (0.7892–0.9591), and the R2 (0.0814–0.3609) for NO2. The RMSE (0.3438–0.3975), NAE (0.7387–0.9015), and R2 (0.2005–0.4399) for CO were all within acceptable limits. For O3, the RMSE was 0.0051–0.0057, the NAE was 0.8386–0.9263, and the R2 was 0.1379–0.2953. The API calculation results indicate that PM10 is a significant pollutant in representing the API.

A Case Analysis of Dust Weather and Prediction of PM10 Concentration Based on Machine Learning at the Tibetan Plateau

Dust weather is common and disastrous at the Tibetan Plateau. This study selected a typical case of dust weather and analyzed its main development mechanism in the northeast of the Tibetan Plateau, then applied six machine learning methods and a time series regression model to predict PM10 concentration in this area. The results showed that: (1) The 24-h pressure change was positive when the front intruded on the surface; convergence of vector winds with a sudden drop in temperature and humidity led by a trough on 700 hPa; a “two troughs and one ridge” weather situation appeared on 500 hPa while the cold advection behind the trough was strong and a cyclone vorticity was formed in the east of Inner Mongolia. (2) The trajectory of air mass from the Hexi Corridor was the main air mass path influencing Xining City, in this case, since a significant lag in the peak of PM10 concentration appeared in Xining City when compared with Zhangye City. (3) The Multiple Linear Regression was not only timely and effective in predicting the PM10 concentration but had great abilities for anticipating the transition period of particle concentration and the appearance date of maximum values in such dust weather. (4) The MA and MP in the clean period were much lower than that in the dust period; the PM10 of Zhangye City as an eigenvalue played an important role in predicting the PM10 of Xining City even in clean periods. Different from dust periods, the prediction effect of Random Forest Optimized by Bayesian hyperparameter was superior to Multiple Linear Regression in clean periods.

Classification Prediction of PM10 Concentration Using a Tree-Based Machine Learning Approach

The PM10 prediction has received considerable attention due to its harmful effects on human health. Machine learning approaches have the potential to predict and classify future PM10 concentrations accurately. Therefore, in this study, three machine learning algorithms—namely, decision tree (DT), boosted regression tree (BRT), and random forest (RF)—were applied for the prediction of PM10 in Kota Bharu, Kelantan. The results from these three methods were compared to find the best method to predict PM10 concentration for the next day by using the maximum daily data from January 2002 to December 2017. To this end, 80% of the data were used for training and 20% for validation of the models. The performance measure of the PM10 concentration was based on accuracy, sensitivity, specificity, and precision for RF, BRT, and DT, respectively, which indicates that these three models were developed effectively, and they are applicable in the prediction of other atmospheric environmental data. The best model to use in predicting the next day’s PM10 concentration classification was the random forest classifier, with an accuracy of 98.37, sensitivity of 97.19, specificity of 99.55, and precision of 99.54, but the result of the boosted regression tree was substantially different from the RF model, with an accuracy of 98.12, sensitivity of 97.51, specificity of 98.72, and precision of 98.71. The best model can assist local governments in providing early warnings to people who are at risk of acute and chronic health consequences from air pollution.

Machine learning method using camera image patterns for predictions of particulate matter concentrations

A machine learning technique using image patterns has been proposed to predict the concentrations of particulate matter with an aerodynamic diameter of <10 μm (PM10) originating from on-road vehicles. The analysis process included measurements of PM10 concentrations, image data collection on vehicle transportation patterns and convolutional neural network (CNN). Preprocessing procedures involving high-pass filtering were used to remove the outliers from the measured concentrations of PM10, and multi-neuron components were modeled by connecting the estimated neuron functions. Finally, PM10 concentrations generated from vehicles were predicted through decisions from neural network modeling made in correspondence with the imagery data on estimated vehicles. Machine learning was performed by the classification of four types of images and the results showed good agreement with measurements for the cases of usage in the image patterns of vehicles. Machine learning applied to images of camera photography in this study was designed specifically for the prediction of PM10 concentrations. The findings showed that the proposed method can be used to estimate the real-time contributions of PM10 by vehicles near a road through connection with traffic information from a closed circuit television (CCTV).

Prediction of PM10 concentrations in the city of Agadir (Morocco) using non-linear autoregressive artificial neural networks with exogenous inputs (NARX)

The PM10 (particulate matter with diameter that is less than or equal to 10 μm) is qualified as one of the most dangerous air pollutants, affecting negatively the human health by causing lung, throat and heart diseases. These facts accentuate the need of forecasting the aforementioned air pollutant in order to provide predictions that will help reducing the exposure to the PM10. In this present study, we attempt to forecast PM10 concentrations in Agadir City by employing non-linear autoregressive artificial neural networks with exogenous multi-variable inputs (NARX-ANN) 1-h and 24-h ahead. We have used different NARX-ANN architectures depending on the structure and the input parameters in order to find the most optimal ones. The performance and the quality of the models were evaluated using root mean square error (RMSE), coefficient of correlation (CC) and mean absolute error (MAE).

Analysis PM10 and Visibility During Forest Fire in Palangka Raya

Forest fires have an impact on air quality and visibility. Visibility can be associated with a highly visual indicator of air pollution. This research aims to analyze the relationship between the PM10 concentration and visibility during the forest firest events and normal conditions in Palangkaraya from 2000 to 2014 by using a regression method. The relative humidity data was used to filter the PM10 and visibility. Furthermore, the equation resulted from the regression analysis was used to predict PM10 concentration in Palangka Raya. The result showed that the regression pattern tends to form a logarithmic function. Specifically, without filtering data, the coefficient correlation (r-value) during the forest fire events and normal conditions are 0.69 and 0.5, respectively. Meanwhile, a data filtering method gives a higher relationship between PM10 and visibility, with the r-value of 0.7 for the forest fire events and 0.68 for the normal condition. On the other hand, the prediction of PM10 concentration indicates a high bias value due to the other influenced factors that have not been included in this study.

Coupling of quantile regression into boosted regression trees (BRT) technique in forecasting emission model of PM10 concentration

Air pollution is currently becoming a significant global environmental issue. The sources of air pollution in Malaysia are mobile or stationary. Motor vehicles are one of the mobile sources. Stationary sources originated from emissions caused by urban development, quarrying and power plants and petrochemical. The most noticeable contaminant in the Peninsular of Malaysia is the particulate matter (PM10), the highest contributor of Air Pollution Index (API) compared to other pollution parameters. The aim of this study is to determine the best loss function between quantile regression (QR) and ordinary least squares (OLS) using boosted regression tree (BRT) for the prediction of PM10 concentration in Alor Setar, Klang and Kota Bharu, Malaysia. Model comparison statistics using coefficient of determination (R2), prediction accuracy (PA), index of agreement (IA), normalized absolute error (NAE) and root mean square error (RMSE) show that QR is slightly better than OLS with the performance of R2 (0.60–0.73), PA (0.78–0.85), IA (0.86–0.92), NAE (0.15–0.17) and RMSE (9.52–22.15) for next-day predictions in BRT model.

Assessment and prediction of PM10 concentration using ARIMA

Urban air pollution is rapidly becoming an environmental issue of global concern, affecting public health, regional weather, and climate. The air quality crisis in cities is mainly caused by vehicle emissions. A city's air quality is assessed using the Air Quality Index (AQI), which analyses concentrations of air pollutants such as PM10, sulphur dioxide SO2, and nitrogen dioxide NO2. The current study on air pollution was carried out in the SG Halli area of Bangaluru and required air pollutant data was collected from the Karnataka State Pollution Control Board (CPCB). With the exception of PM10, the pollutant concentration was within the permissible limits set by the Central Pollution Control Authority (CPCB). Pollutant concentrations were analyzed using meteorological variables such as relative humidity (RH), temperature (T), wind direction (WD), wind speed (WS), and solar radiation (SR). PM10 concentrations are predicted using an auto-regressive integrated moving average (ARIMA) model. PM10 concentrations were well predicted by ARIMA model. On the AQI scale, it can be seen that the atmospheric environment of SG Halli is moderately polluted.

A comparative study of different LSTM neural networks in predicting air pollutant concentrations

Aim/Objective: This study aims to identify the key trends among different types of LSTM networks and their performance and usage for air pollutants(PM2:5 and PM10) concentrations prediction. Methods: In this study, the extensive research efforts were made for Particulate Matters (i.e., PM10 and PM2:5) prediction using several LSTM networks, namely Vanilla, Stacked, and Bidirectional. These are trained and tested using air quality data, retrieved from the Central Pollution Control Board (CPCB) of the town Bawana, Delhi. Realtime hourly a data from 2018 to 2020 with nine air pollutants are considered for experimental analysis. We conducted data preparation strategy to select the best features, which improve the quality of the data. An adequate number of experiments are conducted to choose the best hyperparameters using Python package TensorFlow. Findings: MSE, MAE, RMSE, and R2 parameters are used as the statistical criteria for evaluating the model’s performances.The numerical experiments revealed that deep neural networks could predict the Particulate Matters (mg/m3) with high accuracy. We found that Stacked LSTM with minimum MSE, MAE, RMSE, and maximum R2 works better than the other two methods, i.e., Vanilla LSTM and Bidirectional LSTM for PM2:5 and PM10 concentrations prediction. The empirical, experimental analysis also shows that Vanilla, Stacked, and Bidirectional LSTM models have comparatively minimum MSE, MAE, RMSE, and maximum R2 for PM2:5 than PM10 concentration prediction. Applications: With the help of a predictive model, one can find reliable fine concentration prediction information for a particular area. The resultant information on relative performance can help researchers in the selection of an appropriate LSTM algorithm for their studies. Keywords: Air pollutants; air quality index; PM concentrations; LSTM;TensorFlow; air pollution

Spatial prediction of PM10 concentration using machine learning algorithms in Ankara, Turkey

With the increase in population and industrialization, air pollution has become one of the global problems nowadays. Therefore, air pollutant parameters should be measured at regular intervals, and the necessary measures should be taken by evaluating the results of measurements. In order to prevent air pollution, pollutant parameters must be evaluated within the framework of a model. Recently, in order to obtain objective and more sensitive results with regard to air pollution nowadays, studies, which use machine learning algorithms in artificial intelligence technologies, have been carried out. In this study, PM10 concentrations, which are obtained from 7 stations in Ankara province in Turkey, were trained with machine learning algorithms (LASSO, SVR, RF, kNN, xGBoost, ANN). The PM10 concentrations of the years 2009–2017 of 6 stations in Ankara were given as input, and the PM10 concentrations of the seventh station for the year 2018 were predicted. The model development stage was repeated for each station, and the performance and error rates of the algorithms were determined by comparing the results produced by the algorithms with the actual results. The best results were provided with ANN (R2 = 0.58, RMSE = 20.8, MAE = 14.4). The spatial distribution of the estimated concentration results was provided through Geographic Information System (GIS), and spatial strategies for improving air pollution over land use were established.

Modeling air pollution levels in volcanic geological regional properties and microclimatic conditions

Air pollution was predicted in this study by using multiple linear regression and 42 different artificial neural network models in Igdir/Turkey. Daily air quality data for the years 2016–2018 have been used in the modeling. In the prediction of the particulate matter which has 10 μm or less in diameter (PM10) concentration, sulfur dioxide, nitrogen oxides, nitrogen monoxide, ozone, nitrogen dioxide, relative humidity, air pressure, wind direction and wind speed data were used as input parameters. In the artificial neural network structures, two different learning functions, three different transfer functions and seven different neuron numbers were examined in the MATLAB software. According to results, multiple linear regression did not predict the PM10 concentration in the atmosphere. The R2 value was determined as 0.543 for the multiple linear regression. In this model, the RMSE, MAE and R2 were determined as 0.0488, 0.0248 and 0.9826, respectively. Since the R2 value in this model was quite high, it was concluded that the model is suitable for the prediction of PM10 concentration.

Time Series Analysis of PM10 Concentration in Parit Raja Residential Area

Parit Raja is one of the sub-urban area that rapidly grow due to its location containing industrial and education hub. Pollution from factories and the increasing number of vehicles are the main contributors of PM10. Since PM10 can give the adverse effect to human health such as asthma, cardiovascular disease and lung problem, appropriate action mainly involve short-term prediction maybe required as a precaution. This research was conducted to predict the PM10 concentration using the best time series model in Parit Raja, Batu Pahat, Johor. Primary data was obtained using E-Sampler at three monitoring stations; Sekolah Menengah Kebangsaan (SMK) Tun Ismail, Kolej Kediaman Melewar and Sekolah Rendah Kebangsaan Pintas Raya. ARIMA time series model was used to predict the PM10 concentration and the most suitable model is identify using by Akaike Information Criterion (AIC). Prediction of PM10 concentration for for the next 48 hours at all monitoring locations was verified using three error measures which are mean absolute error (MAE), normalized absolute error (NAE) and root mean square error (RMSE). After comparing the time series model, the short term prediction model for station 1 is AR(1), station 2 is ARMA(1,1) and station 3 is ARMA(2,1) based on the smallest AIC value and the best time series model that used for prediction at Parit Raja residential area is AR(1). Since the best model was identified for Parit Raja residential area, PM10 concentration can be predicted using AR(1) model to identify the value of PM10 concentration in the next day.

A hybrid Grey-Markov/ LUR model for PM10 concentration prediction under future urban scenarios

Exploring the spatial distribution of air pollutants under future urban planning scenarios is essential as urban sprawl increases in China. However, existing published prediction models usually forecast pollutant concentrations at the station level or estimate spatial distribution of pollutant in a historical perspective. This study has developed a hybrid Grey-Markov/land use regression (LUR) model (GMLUR) for PM10 concentration prediction under future urban scenarios by employing the forecast of Grey-Markov model as surrogate measurements to calibrate the spatial estimations of LUR model. Taking the agglomeration of Changsha-Zhuzhou-Xiangtan (CZT) in China as a case, the superiority of GMLUR was tested and spatial distribution of PM10 concentrations based on four potential land use scenarios for the year 2020 were predicted. Results show that GMLUR modelling outperforms LUR modelling with clear lower average relative percentage error (5.13% vs. 24.09%) and root-mean-square error (5.50 μg/m3 vs. 21.31 μg/m3). The economic interest scenario identifies the largest demands of future built-up (2 306.50 km2) and bare (34.88 km2) areas. Built-up area demands for the business as usual scenario, resource-conserving scenario, and ecological interest scenario are 362.67, 1 042.22, and 1 014.70 km2, respectively. Correspondingly, the economic interest scenario identifies the severest PM10 pollution with the highest mean predicted concentration of 53.78 μg/m3 and the largest percent (19.43%) of area exceeding the Level 2 value (70 μg/m3) of Chinese National Ambient Air Quality Standard (CNAAQS); these are significantly higher than those of the business as usual scenario (49.63 μg/m3, 6.28%). The resource-conserving scenario (46.79 μg/m3) and ecological interest scenario (46.76 μg/m3) are cleaner with no area exceeding the Level 2 value of CNAAQS. It can be concluded that GMLUR modelling provides a feasible way to evaluate the potential outcome of future urban planning strategies in the perspective of air pollution.

Real-time PM10 Concentration Prediction LSTM Model based on IoT Streaming Sensor data

Real-time PM10 Concentration Prediction LSTM Model based on IoT Streaming Sensor data