Predict Carbon Monoxide Concentrations Research Articles

Machine learning predictions for carbon monoxide levels in urban environments

The increasing carbon emissions in Malaysia necessitate accurate methods to track and control pollution levels. This study focuses on predicting carbon monoxide (CO) concentrations in Petaling Jaya using various machine learning models, and two important parameters, CO concentration and time, were considered in the analysis. Six distinct machine learning models were assessed to gauge their predictive capabilities. These models encompassed a range of regression modeling techniques namely; Linear, Tree, Gaussian Process, Ensemble of Trees, Support Vector, and Artificial Neural Networks (ANN) modelling. The Matern 5/2 Gaussian Process Regression (GPR) model consistently outperformed other models across all scenarios, demonstrating high R2 values and low RMSE, MSE, and MAE values. Specifically, in scenarios 1, 2, 3, and 4, the Matern 5/2 model exhibited the lowest RMSE (0.084–0.088) and highest R2 (0.97), highlighting its reliability and robustness in predicting CO concentrations. Additionally, the Rational Quadratic model achieved an R2 of 0.97 with an RMSE of 0.088 in scenario 1, while the Quadratic SVM excelled in scenario 3 with an R2 of 0.965 and low RMSE, MSE, and MAE values (0.085, 0.007, and 0.066). These findings provide valuable insights for environmental protection, renewable energy transition, energy efficiency, sustainable land use planning, and public awareness. However, acknowledging the study's single-area focus and potential limitations in representing diverse regions, further research is warranted to explore carbon emissions across varied areas and enhance the generalizability of the findings.

PERBANDINGAN MODEL PREDIKSI DATA MINING DALAM MEMPREDIKSI KONSENTRASI POLUTAN KARBON MONOKSIDA (CO) DI JAKARTA

DKI Jakarta, as the capital of Indonesia, faces serious challenges in terms of air quality. Carbon monoxide (CO) is one of the main air pollutants in Jakarta that is harmful to human health and the environment. Data mining is a method that can be used to predict situations based on a model. The study aims to compare data mining models with the best-performing methods to predict carbon monoxide pollutants in Jakarta. The predictive data mining model of the python library is tested and evaluated based on the evaluation metrics of MASE, RMSSE, MAE, RMSE, MAPE and SMAPE values. The model test results showed that K Neighbors with the Conditional Deseasonalize & Detrending model had the best metric evaluation value to predict CO concentration with the value evaluation metrics of MASE 0.2942, RMSSE 0.2483, MAE 2.7362, RMSE 3.3863, MAPE 0.1975 and SMAPE 0.01993. Overall, K Neighbors with the Conditional Deseasonalize & Detrending model shows good performance to predict CO concentrations in Jakarta, but further adjustments are needed to improve accuracy.

Prediction of atmospheric carbon monoxide concentration utilizing different machine learning algorithms: A case study in Kuala Lumpur, Malaysia

Insidious toxin carbon monoxide (CO) can imitate a wide range of different disease states. Clinicians have, and will continue to have, serious concerns about the impact of CO imbalances on the human body. Carbon monoxide concentration has been exceeding the allowable levels in Malaysia. Owing to this, the main objective of this research is to propose a carbon monoxide (CO) prediction model based on machine learning techniques. Three years of historical data were used as input to develop the proposed models to predict carbon monoxide concentrations on a 12-hour and 24-hour basis. Four different machine learning technique models were used for the prediction which are Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), and Automated Neural Network – Multi-Layer Perceptron (ANN-MLP). The input parameters used are wind speed, humidity, Ozone (O3), Nitric oxide (NOx), Sulfur dioxide (SO2), and Nitrogen Dioxide (NO2). For each location, in this study, the uncertainty of the models utilized has been implemented to ensure the robustness of the performance. Furthermore, Taylor Diagram has been constructed to distinguish the performance of each model. The results indicate that ANN-MLP outperformed the all-other models involved in this study and showed efficiency in predicting Carbone monoxide concentration. By using ANN-MLP, the highest determination coefficient R2 were achieved which are 0.7190, 0.8914 and 0.7441 for the first station (S1), second station (S2) and the third station (S3) respectively by using 24-hour dataset. Meanwhile, by using a 12-hour dataset, 0.7490 for S1, 0.8942 for S2 and 0.8127 for S3. The uncertainty analysis of the ANN-MLP has 0.99 of confidence level and the lowest d-factor achieved, at S2 by using 12-hour dataset, is 0.000250455. These results ensure the effectiveness and robustness of ANN-MLP to predict carbon monoxide in the tropospheric layer. Code availabilityNot applicable.

A comparative analysis for a deep learning model (hyDL-CO v1.0) and Kalman filter to predict CO concentrations in China

Abstract. The applications of novel deep learning (DL) techniques in atmospheric science are rising quickly. Here we build a hybrid DL model (hyDL-CO), based on convolutional neural networks (CNNs) and long short-term memory (LSTM) neural networks, to provide a comparative analysis between DL and Kalman filter (KF) to predict carbon monoxide (CO) concentrations in China in 2015–2020. We find the performance of DL model is better than KF in the training period (2015–2018): the mean bias and correlation coefficients are 9.6 ppb and 0.98 over eastern China and are −12.5 ppb and 0.96 over grids with independent observations (i.e., grids with CO observations that are not used in DL training and KF assimilation). By contrast, the assimilated CO concentrations by KF exhibit comparable correlation coefficients but larger negative biases. Furthermore, the DL model demonstrates good temporal extensibility in the test period (2019–2020): the mean bias and correlation coefficients are 95.7 ppb and 0.93 over eastern China and 81.0 ppb and 0.91 over grids with independent observations, while CO observations are not fed into the DL model as an input variable. Despite these advantages, we find a weaker prediction capability of the DL model than KF in the test period, and a noticeable underestimation of CO concentrations at extreme pollution events in the DL model. This work demonstrates the advantages and disadvantages of DL models to predict atmospheric compositions with respect to traditional data assimilation, which is helpful for better applications of this novel technique in future studies.

Impact of Traffic on Carbon Monoxide Concentrations Near Urban Road Mid-Blocks

An increase in population, number of vehicles and rate of urbanization have happened drastically in the past decades, resulting in deteriorating air quality and a threat to the environment. Vehicular emissions are among the primary and noted sources of air pollution in urban areas. Carbon monoxide (CO) is one of the major pollutants emitted from vehicles, affecting the environment adversely. The present study attempts to develop models to predict CO concentrations at different mid-block sections of urban roads using multiple linear regression (MLR) and artificial neural networks (ANN) methods. The proportional share of vehicles and average traffic speed are considered as inputs to the models. The traffic volume, speed and CO concentrations collected at different mid-block sections have been analyzed. A good correlation between average traffic speed, traffic volume and CO concentrations was observed. The study also shows that the classified traffic volume and average traffic speed in a mid-block could help explain the variance in CO levels significantly, with R2 value of 0.91 and 0.97 for MLR and ANN models, respectively. It has been observed that the CO level would be as high as 14 ppm at or below the average speed of 25 km/h for moving vehicles. The linear increase in CO concentration was found due to decrease in average speed of traffic stream. The model validation illustrates that the estimated CO concentrations match well the observed CO concentrations under the same set of traffic and roadway conditions with an MAPE value of 0.18% and 0.11% for MLR and ANN models, respectively. It seems interesting to see that the ANN model had better predicted the CO values than MLR model. This study considers developing cities to determine the CO levels. The results of the study would help Pollution Control Board officials and traffic control authorities to implement necessary measures for improving air quality in the cities under consideration.

Predicting Carbon Monoxide Concentrations in the Air of Pardis City, Iran, Using an Artificial Neural Network

To date, several methods have been proposed to explain the complex process of air pollution prediction. One of these methods uses neural networks. Artificial neural networks (ANN) are a branch of artificial intelligence, and because of their nonlinear mathematical structures and ability to provide acceptable forecasts, they have gained popularity among researchers. The goal of our study as documented in this article was to compare the abilities of two different ANNs, the multilayer perceptron (MLP) and radial basis function (RBF) neural networks, to predict carbon monoxide (CO) concentrations in the air of Pardis City, Iran. For the study, we used data collected hourly on temperature, wind speed, and humidity as inputs to train the networks. The MLP neural network had two hidden layers that contained 13 neurons in the first layer and 25 neurons in the second layer and reached a mean bias error (MBE) of 0.06. The coefficient of determination (R2), index of agreement (IA), and the Nash–Scutcliffe efficiency (E) between the observed and predicted data using the MLP neural network were 0.96, 0.9057, and 0.957, respectively. The RBF neural network with a hidden layer containing 130 neurons reached an MBE of 0.04. The R2, IA, and E between the observed and predicted data using the RBF neural network were 0.981, 0.954, and 0.979, respectively. The results provided by the RBF neural network had greater acceptable accuracy than was the case with the MLP neural network. Finally, the results of a sensitivity analysis using the MLP neural network indicated that temperature is the primary factor in the prediction of CO concentrations and that wind speed and humidity are factors of second and third importance when forecasting CO levels.

Model validation study of carbon monoxide transport due to portable electric generator operation in an attached garage

A series of tests was conducted to characterize the indoor carbon monoxide (CO) concentrations resulting from portable electric generators operating in the attached garage of a test house under various use and environmental conditions. An extensive model validation effort using the multizone airflow and indoor air quality (IAQ) model CONTAM was carried out using the data from seven tests with a generator operating in the attached garage to compare predicted CO concentrations with measured values. The agreement between the measurements and predictions of the O2 concentrations in the garage and the average CO concentration for the house zones was excellent for the data set as a whole. The agreement was somewhat worse for the garage CO concentrations. Overall, the house zone average and garage CO concentration predictions and measurements were within about 20% and 30%, respectively, when averaged over all cases.

Develop dynamic model for predicting traffic CO emissions in urban areas.

The greater the use of energy in the transportation sectors, the higher the emission of carbon monoxide (CO), and hence inevitable harm to environment and human health. In this concern, measuring and predicting of CO emission from transportation sector-especially large cities-is important as it constitute 90% of all CO emission. Many urban cities in developing world have not properly experienced such measurements or predictions. In this paper, for the first time, field measurements of traffic characteristics data and corresponding CO concentration have been performed for developing a model for predicting CO emissions from transportation sector for New Borg El Arab (NBC), Egypt. The performance of Swiss-German Handbook Emission Factors for Road Transport (HBEFA v3.1) model has been assessed for predicting the CO concentration at roadside in the study area. Results indicated that HBEFA v3.1 underestimate emission figures. The developed CO dynamic emission model involves the traffic flow characteristics with roadside CO concentrations. Acceptable representation of measured CO concentration has been shown by the developed dynamic CO emission model which introduces R (2) = 0.77, mean biases and frictional biases of -0.27mgm(-3) and 0.09, respectively. A comparison between predicted CO concentrations using HBEFA v3.1 and the promoted dynamic model indicate that HBEFA v3.1 estimates CO emission concentrations in the study area with a mean error and frictional biases 159.26 and 233.33%, respectively, higher than those of the developed model.

Performance evaluation of ISCST3, adms-urban and aermod for urban air quality management in a mega city of India

Urban air quality has deteriorated in last few decades in the mega cities of both developed and developing countries. Many mathematical models have been widely used as prediction tool for urban air quality management in developed countries. However, applications of these models are limited in developing countries including India due to lack of suffi cient validation studies. In this paper, three state-of-the-art air quality models namely AERMOD, ADMS-Urban and ISCST3 have been used to predict the air quality at an intersection in Delhi city, India, followed by their performance evaluation and sensitive analysis under different meteorological conditions. The models have been run for different climatic conditions, i.e. summer and winter season to predict the concentration of carbon monoxide (CO), nitrogen dioxide (NO 2 ) and PM 2.5 (diameter size less than 2.5 µm). The ISCST3 has performed satisfactorily (d = 0.69) for predicting CO concentrations when compared with AERMOD (d = 0.50) and ADMS-Urban (d = 0.45) for winter period. The ADMS-Urban (d = 0.49) has performed satisfactorily for predicting NO 2 concentration when compared with ISCST3 (d = 0.36) and AERMOD (d = 0.32). The AERMOD, ISCST3 and ADMS-Urban have performed satisfactorily for predicting PM 2.5 concentrations having d values as 0.46, 0.45 and 0.43 respectively. All three models have performed satisfactorily for predicting CO concentrations when wind speed was in the range of 0.5–3 m/s and wind direction in the range 90–180 degrees, i.e. downwind direction. The difference in model’s performance may be due to differences in model formulation and the treatment of terrain features. The causal nature of these Gaussian based models may be one of the reasons for difference in performance of the models, because these are sensitive to

Evaluation of Dispersion Models for Predicting Carbon Monoxide Concentrations from Motor Vehicles in a Metropolitan Area

The objective of this study is to provide a simulation of pollutant atmospheric dispersion from vehicles exhausts. This study evaluates two atmospheric Gaussian dispersion models against experimental results obtained near a roadway intersection located in an urban area; the agreement of measured and predicted CO concentrations is very acceptable. The paper shows also the computed spatial distribution of CO concentrations at the ground level emitted from motor vehicles in the urban area around the monitoring site. The data that have a direct impact on emissions of all motor vehicles were determined through integration of information provided by agencies and institutions, and employing the COPERT methodology. The models were also run to determine when the highest levels of CO would occur in a worst-case meteorological conditions scenario. The results indicate that the levels of CO are below the ambient air quality standard.

Study of the impact of operation distance of outdoor portable generators under different weather conditions

The US Centres for Disease Control and Prevention has reported that up to half of non-fatal carbon monoxide (CO) poisoning incidents during the hurricane seasons in 2004 and 2005 involved generators operated outdoors but within 7 ft of the home. Current guidance for safe operating distances of generators is often neither specific nor consistent. A study was conducted to examine the impact of generator distance on indoor CO exposure. The study was based on computer simulations of CO transport outdoors and subsequently into a generic two-storey house. This paper presents the simulation results when using an indoor air quality model coupled with a computational fluid dynamics model to predict CO concentrations near and within the home. Fire Dynamics Simulator was validated against the measured contaminant dispersion data in a wind tunnel. A parametric study was then conducted for the two-storey house to consider the effects on indoor CO levels of generator location, distance, exhaust temperature and speed and weather conditions. It was found that in most cases, to reduce CO levels for the conditions modelled, it was more effective to point the generator exhaust away from the house and position the generator at a distance of >4.6 m.

An Application and Evaluation of the CAL3QHC Model for Predicting Carbon Monoxide Concentrations from Motor Vehicles Near a Roadway Intersection in Muscat, Oman

The CAL3QHC model was used to predict carbon monoxide (CO) concentrations from motor vehicles at an existing urban intersection (Star Cinema in Muscat area, Oman). The CO concentrations predicted from the model were compared with those measured in the field. Predicted average CO concentrations were found to compare favorably with measured values obtained at all eight receptors considered within the modeled intersection. In general, the comparison indicates good agreement with some underprediction for CO. For receptor 6, the model overpredicts the average CO concentration. This overprediction is associated with the presence of trees and green area in the location of receptor 6. In general, the measurements and the model results indicated that the highest CO concentrations were found to occur close to the intersection and, hence, a decrease in the concentration levels was seen as the distance from the road increased. The results indicated that the levels of CO were well below the ambient air quality standard and that probably no health risk was present in areas adjacent to the star cinema intersection. However, the predicted worst-case 1-h CO concentrations assuming inversion atmospheric stability conditions (class F) and wind speed of 1 m/s indicated that the levels of CO were close to or higher than the Oman's National Ambient Air Quality Standards (NAAQS) value of 35 ppm at all receptors considered. The results of this study are useful in transport development and traffic management planning.

Sensitivity Analysis of CAL3QHC Roadway Intersection Model

Emissions from vehicular sources are a major contribution to air pollution in every country. Air quality models are being used to study the impact of such emissions. The CAL3QHC emissions model was developed by the U.S. Environmental Protection Agency based on California’s CAL3Q model. One area of concern with modeling results is uncertainty in input data, in model calculations, and in natural atmospheric processes. The focus of the study is uncertainty in input data for the CAL3QHC roadway model. Two practical methods—the ASTM approach and the least rigid approach—are used to perform the sensitivity analysis on the input parameters. The two important areas in which CAL3QHC requires sensitivity discussions are the emissions source strength of the vehicles in the queue and the link length (which represents the number of vehicles in a queue). The variability of these two parameters results in a nonlinear relationship between the source strengths and the predicted concentrations. Sensitivity analysis of the CAL3QHC model is carried out for a simple roadway intersection with two traffic lanes and two receptor locations (at the corner of the intersection and midblock). With results from the ASTM method, sensitivity indices are calculated for signal timing, traffic volume, number of traffic lanes, and wind speed. Among all indices, wind speed shows the maximum sensitivity on predicted carbon monoxide concentrations. The model needs no calibration because none of the parameters studied show Type IV sensitivity using ASTM results.

New Statistical Framework for Estimating Carbon Monoxide Impacts

The EPA specifies that level-of-service D intersections must undergo detailed microscale carbon monoxide modeling using the recommended software such as CAL3QHCr, which requires an in-depth knowledge of microscale dispersion modeling of pollutants. In this study, a new statistical framework is introduced that is intended to replicate the microscale modeling results achieved with CAL3QHCr. The new framework provides an important means for assessing air quality impacts during the design phase of a transportation project, without having to acquire the modeling skills and/or dispersion modeling knowledge necessary to execute CAL3QHCr. To develop the framework, approximately 23,000 CAL3QHCr simulation scenarios were generated by varying the input values of the major design and dispersion modeling factors. The input values, together with the corresponding output values of CAL3QHCr simulations, were used to specify a statistical model for predicting carbon monoxide concentrations. The statistical model was then validated using field meteorological conditions for five cities in California. The study results show that the proposed model can be easily and reliably used by traffic engineers to predict potential carbon monoxide exceedances at the planning stages for transportation projects.

A New Screening Measure to Identify Potential Carbon Monoxide Hotspots

ABSTRACT To demonstrate conformity of transportation projects to National Ambient Air Quality Standards in accordance with State Implementation Plans, the U.S. Environmental Protection Agency (EPA) uses intersection level of service (LOS) as one of its major criteria for screening for potential carbon monoxide (CO) hotspots. Although intersection LOS is a measure of traffic volume, signal timing, and related congestion and delay, the assigned level reflects only the computed averaged stopped delay (ASD) per vehicle at the intersection. Thus, intersections can often operate at the same LOS but produce vastly different levels of predicted CO concentrations. For example, a two-lane approach operating at LOS D will produce very different levels of CO than a five-lane approach also operating at LOS D. This study explores the effectiveness of the LOS D criterion as a screen for identifying potential CO hotspots. The study results indicate that LOS is a poor predictor of potential CO hotspots when compared to results generated with the EPA-recommended micro-scale model CAL3QHCr. To more consistently screen out those intersections that will not be identified as CO hotspots using the micro-scale models, a new criterion, equivalent red-time vehicles (ERTV), is introduced. The modeling results using ERTV suggest that it is a more robust measure for identifying potential CO hotspots, and conversely, screening out those intersections that are not likely to be identified as hotspots using micro-scale simulation results.

Line source model for vehicular pollution prediction near roadways and model evaluation through statistical analysis

Vehicular traffic has become a major source of air pollution in urban areas. Mathematical models can be used to determine the environmental impacts of the existing or developing transportation projects which combine the effects of source strength and meteorology to describe the resulting ambient air concentrations. The objective of the present study was to identify the mathematical model that could be most useful to predict the pollutant concentration near roadways. The Gaussian-based Finite Line Source Model was employed to predict pollutant concentration due to automobiles from roadways. The model was applied to four sections of major highway in Madras, the capital city of Tamilnadu, to predict carbon monoxide concentrations. Comparisons between model outputs and measurements were performed using both quantitative data analysis techniques and statistical methods to evaluate model performance. The analysis indicates that the model’s performance was good. ©

Queueing Algorithm for Calculating Idling Emissions in FLINT—the FLorida INTersection Air Quality Model

The theoretical development of the queueing model used in the FLINT (FLorida INTersection) air quality model is described. FLINT is an area source model used to predict carbon monoxide concentrations for under-saturated and oversaturated traffic conditions at signalized intersections. In the FLINT model, deterministic queueing is used to estimate the queue length for cases of undersaturated conditions. In oversaturated cases, a cycle failure method has been developed to estimate queue length. In addition, a new concept is presented for calculating idling time for each vehicle’s position in the queue during both the red and the green phases of the traffic signal cycle. A limited set of undersaturated cases from monitoring data in Melrose Park, Illinois, was used to compare the predicted queue lengths with the measured queue lengths for several air quality models. It was found that FLINT predicted the queue length within one vehicle of the observed queue length. The same cases were tested using CAL3QHC, TEXIN2 intersection air quality models, and the American Automobile Manufacturers Association (AAMA) simulation model. It was found that predictions of the AAMA and the FLINT models were very close to the measured queue lengths in cases of undersaturated conditions. Moreover, although the FLINT and the AAMA models use a different approach to estimate queue length, their predicted queue lengths were very close in oversaturated cases. However, the predicted queue lengths of CAL3QHC were too long for oversaturated cases.

火災時の一酸化炭素生成予測モデル : 火災時における建物内の一酸化炭素濃度予測に関する研究 その1

火災時の一酸化炭素生成予測モデル : 火災時における建物内の一酸化炭素濃度予測に関する研究 その1

TEXIN2: A Versatile Model for Predicting Carbon Monoxide Concentrations near Intersections

The original Texas Intersection Model (TEXIN) for air quality near street intersections has been widely adopted across the nation. At the request of transportation agencies from several states, the TEXIN model was revised to improve its performance and flexibility. The new capabilities include T-intersections, one-way streets, four-way stops, inspection/maintenance capabilities, anti-tampering programs and a short-cut emissions algorithm. TEXIN2 uses the CMA procedures for estimating traffic flow parameters, MOBILE3 to determine free flowing traffic cruise emissions, and CALINE3 to model the pollutant distribution downwind of an intersection. The new model, TEXIN2, offers the user more accurate simulations with enhanced versatility while still requiring a minimal amount of input data.

Modeling of Carbon Monoxide Hot Spots

This paper describes the results of a measurement and modeling study of carbon monoxide (CO) concentrations In the proximity of intersections. Analysis for model performance of paired observed and predicted CO concentrations are presented. Two methodologies of pollutant prediction were used: the Intersection Midblock Model (IMM) and a statistical multiple linear regression. The results showed that both methods underpredicted frequently and dispensed results that were site specific. In addition, correlations of IMM predicted concentrations to observed concentrations were poor (typically r2 values <0.25). Various explanations for this observation are proposed. The statistical approach exhibited an improved accuracy over that of IMM. However, some of the independent variables used might be difficult to obtain as a routine measurement, and use of a one or two independent parameter model yielded adjusted R2 values comparable to the r2 values observed with IMM. Based on these results, an Intersection model applicable under a wide range of conditions of traffic, meteorology, and geometry is not available. Research Is needed to develop one, since its use would often be called on in the development of air quality sections of Environmental Assessments or Environmental Impact Statements.