Optimization of artificial neural network models through genetic algorithms for surface ozone concentration forecasting

Introduction

This study proposes three methodologies to define artificial neural network models through genetic algorithms (GAs) to predict the next-day hourly average surface ozone (O₃) concentrations. GAs were applied to define the activation function in hidden layer and the number of hidden neurons.

Methods

Two of the methodologies define threshold models, which assume that the behaviour of the dependent variable (O₃ concentrations) changes when it enters in a different regime (two and four regimes were considered in this study). The change from one regime to another depends on a specific value (threshold value) of an explanatory variable (threshold variable), which is also defined by GAs. The predictor variables were the hourly average concentrations of carbon monoxide (CO), nitrogen oxide, nitrogen dioxide (NO₂), and O₃ (recorded in the previous day at an urban site with traffic influence) and also meteorological data (hourly averages of temperature, solar radiation, relative humidity and wind speed). The study was performed for the period from May to August 2004.

Results and discussion

Several models were achieved and only the best model of each methodology was analysed. In threshold models, the variables selected by GAs to define the O₃ regimes were temperature, CO and NO₂ concentrations, due to their importance in O₃ chemistry in an urban atmosphere.

Conclusion

In the prediction of O₃ concentrations, the threshold model that considers two regimes was the one that fitted the data most efficiently.     

PAH air pollution at a Portuguese urban area: carcinogenic risks and sources identification

This study aimed to characterize air pollution and the associated carcinogenic risks of polycyclic aromatic hydrocarbon (PAHs) at an urban site, to identify possible emission sources of PAHs using several statistical methodologies, and to analyze the influence of other air pollutants and meteorological variables on PAH concentrations.The air quality and meteorological data were collected in Oporto, the second largest city of Portugal. Eighteen PAHs (the 16 PAHs considered by United States Environment Protection Agency (USEPA) as priority pollutants, dibenzo[a,l]pyrene, and benzo[j]fluoranthene) were collected daily for 24 h in air (gas phase and in particles) during 40 consecutive days in November and December 2008 by constant low-flow samplers and using polytetrafluoroethylene (PTFE) membrane filters for particulate (PM₁₀ and PM_2.5 bound) PAHs and pre-cleaned polyurethane foam plugs for gaseous compounds. The other monitored air pollutants were SO₂, PM₁₀, NO₂, CO, and O₃; the meteorological variables were temperature, relative humidity, wind speed, total precipitation, and solar radiation. Benzo[a]pyrene reached a mean concentration of 2.02 ng?m^?3, surpassing the EU annual limit value. The target carcinogenic risks were equal than the health-based guideline level set by USEPA (10^?6) at the studied site, with the cancer risks of eight PAHs reaching senior levels of 9.98?×?10^?7 in PM₁₀ and 1.06?×?10^?6 in air. The applied statistical methods, correlation matrix, cluster analysis, and principal component analysis, were in agreement in the grouping of the PAHs. The groups were formed according to their chemical structure (number of rings), phase distribution, and emission sources. PAH diagnostic ratios were also calculated to evaluate the main emission sources. Diesel vehicular emissions were the major source of PAHs at the studied site. Besides that source, emissions from residential heating and oil refinery were identified to contribute to PAH levels at the respective area. Additionally, principal component regression indicated that SO₂, NO₂, PM₁₀, CO, and solar radiation had positive correlation with PAHs concentrations, while O₃, temperature, relative humidity, and wind speed were negatively correlated.     

Predicting hourly air pollutant levels using artificial neural networks coupled with uncertainty analysis by Monte Carlo simulations

Recent progress in developing artificial neural network (ANN) metamodels has paved the way for reliable use of these models in the prediction of air pollutant concentrations in urban atmosphere. However, improvement of prediction performance, proper selection of input parameters and model architecture, and quantification of model uncertainties remain key challenges to their practical use. This study has three main objectives: to select an ensemble of input parameters for ANN metamodels consisting of meteorological variables that are predictable by conventional weather forecast models and variables that properly describe the complex nature of pollutant source conditions in a major city, to optimize the ANN models to achieve the most accurate hourly prediction for a case study (city of Tehran), and to examine a methodology to analyze uncertainties based on ANN and Monte Carlo simulations (MCS). In the current study, the ANNs were constructed to predict criteria pollutants of nitrogen oxides (NO_x), nitrogen dioxide (NO₂), nitrogen monoxide (NO), ozone (O₃), carbon monoxide (CO), and particulate matter with aerodynamic diameter of less than 10 μm (PM₁₀) in Tehran based on the data collected at a monitoring station in the densely populated central area of the city. The best combination of input variables was comprehensively investigated taking into account the predictability of meteorological input variables and the study of model performance, correlation coefficients, and spectral analysis. Among numerous meteorological variables, wind speed, air temperature, relative humidity and wind direction were chosen as input variables for the ANN models. The complex nature of pollutant source conditions was reflected through the use of hour of the day and month of the year as input variables and the development of different models for each day of the week. After that, ANN models were constructed and validated, and a methodology of computing prediction intervals (PI) and probability of exceeding air quality thresholds was developed by combining ANNs and MCSs based on Latin Hypercube Sampling (LHS). The results showed that proper ANN models can be used as reliable metamodels for the prediction of hourly air pollutants in urban environments. High correlations were obtained with R ² of more than 0.82 between modeled and observed hourly pollutant levels for CO, NO_x, NO₂, NO, and PM₁₀. However, predicted O₃ levels were less accurate. The combined use of ANNs and MCSs seems very promising in analyzing air pollution prediction uncertainties. Replacing deterministic predictions with probabilistic PIs can enhance the reliability of ANN models and provide a means of quantifying prediction uncertainties.     

Artificial intelligence modeling to evaluate field performance of photocatalytic asphalt pavement for ambient air purification

In recent years, the application of titanium dioxide (TiO₂) as a photocatalyst in asphalt pavement has received considerable attention for purifying ambient air from traffic-emitted pollutants via photocatalytic processes. In order to control the increasing deterioration of ambient air quality, urgent and proper risk assessment tools are deemed necessary. However, in practice, monitoring all process parameters for various operating conditions is difficult due to the complex and non-linear nature of air pollution-based problems. Therefore, the development of models to predict air pollutant concentrations is very useful because it can provide early warnings to the population and also reduce the number of measuring sites. This study used artificial neural network (ANN) and neuro-fuzzy (NF) models to predict NO_x concentration in the air as a function of traffic count (T_r) and climatic conditions including humidity (H), temperature (T), solar radiation (S), and wind speed (W) before and after the application of TiO₂ on the pavement surface. These models are useful for modeling because of their ability to be trained using historical data and because of their capability for modeling highly non-linear relationships. To build these models, data were collected from a field study where an aqueous nano TiO₂ solution was sprayed on a 0.2-mile of asphalt pavement in Baton Rouge, LA. Results of this study showed that the NF model provided a better fitting to NO_x measurements than the ANN model in the training, validation, and test steps. Results of a parametric study indicated that traffic level, relative humidity, and solar radiation had the most influence on photocatalytic efficiency.     

Ozone and grosswetterlagen

Meteorological conditions have a decisive impact on surface ozone concentrations. In this study, an empirical model is used to explain the interdependence of ozone and grosswetterlagen. Different meteorological parameters such as air temperature, global solar radiation, relative humidity, wind direction and wind speed are used. Additional nitric oxide (NO) was taken as a representative for the emission situation and ozone maximum of the preceding day in order to evaluate the development of the photochemical situation. The dataset includes data collected over a period of three years (1992–1994) from three stations outside of Munich and one in the center of Munich. All values become variables by calculating means, sums or maxima of the basic dataset consisting of half-hour means. Seasonal periodicity of data is detected with Fourier analysis and eliminated by a division method after computing a seasonal index. The dataset is divided into three different grosswetterlagen groups, depending on main wind direction. One mostly cyclonic (westerly winds), onemixed (alternating winds) and one onlyanticyclonic (easterly winds). The last is completed with one summertime group including values from April to August. Factor analysis is performed for each group to obtain independent linear variable combinations. Overall, relative humidity is the dominant parameter, a typical value indicating meteorological conditions during a grosswetterlage. Linear multiple regression analysis is performed using the factors obtained to reveal how the ozone concentrations are explained in terms of meteorological parameters and NO. The results improve from cyclonic to anticyclonic grosswetterlagen in conformance with the increasing significance of photochemistry, indicated by the high solar radiation and high temperature, and the low relative humidity and low wind speed. The explained variance r² reaches its maximum with more than 50 % of the time in Munich center. This empirical model is applicable to the forecasting of local ozone maximum concentrations with a total standard error deviation of 8.5 to 12.8 % and, if ozone concentrations exceed 80 ppb, with a standard error deviation of 5.4 to 9.5 %.     

A measurement campaign in a street canyon in Helsinki and comparison of results with predictions of the OSPM model

In 1997, a measuring campaign was conducted in a street canyon (Runeberg St.) in Helsinki. Hourly mean concentrations of CO, NO_x, NO₂ and O₃ were measured at street and roof levels, the latter in order to determine the urban background concentrations. The relevant hourly meteorological parameters were measured at roof level; these included wind speed and direction, temperature and solar radiation. Hourly street level measurements and on-site electronic traffic counts were conducted throughout the whole of 1997; roof level measurements were conducted for approximately two months, from 3 March to 30 April in 1997. CO and NO_x emissions from traffic were computed using measured hourly traffic volumes and evaluated emission factors. The Operational Street Pollution Model (OSPM) was used to calculate the street concentrations and the results were compared with the measurements. The overall agreement between measured and predicted concentrations was good for CO and NO_x (fractional bias were −4.2 and +4.5%, respectively), but the model overpredicted the measured NO₂ concentrations (fractional bias was +22%). The agreement between the measured and predicted values was also analysed in terms of its dependence on wind speed and direction; the latter analysis was performed separately for two categories of wind velocity. The model qualitatively reproduces the observed behaviour very well. The database, which contains all measured and predicted data, is available for further testing of other street canyon dispersion models. The dataset contains a larger proportion of low wind speed cases, compared with other available street canyon measurement datasets.     

Ozone and grosswetterlagen

Meteorological conditions have a decisive impact on surface ozone concentrations. In this study, an empirical model is used to explain the interdependence of ozone and grosswetterlagen. Different meteorological parameters such as air temperature, global solar radiation, relative humidity, wind direction and wind speed are used. Additional nitric oxide (NO) was taken as a representative for the emission situation and ozone maximum of the preceding day in order to evaluate the development of the photochemical situation. The dataset includes data collected over a period of three years (1992–1994) from three stations outside of Munich and one in the center of Munich. All values become variables by calculating means, sums or maxima of the basic dataset consisting of half-hour means. Seasonal periodicity of data is detected with Fourier analysis and eliminated by a division method after computing a seasonal index. The dataset is divided into three different grosswetterlagen groups, depending on main wind direction. One mostly cyclonic (westerly winds), one mixed (alternating winds) and one only anticyclonic (easterly winds). The last is completed with one summertime group including values from April to August. Factor analysis is performed for each group to obtain independent linear variable combinations. Overall, relative humidity is the dominant parameter, a typical value indicating meteorological conditions during a grosswetterlage. Linear multiple regression analysis is performed using the factors obtained to reveal how the ozone concentrations are explained in terms of meteorological parameters and NO. The results improve from cyclonic to anticyclonic grosswetterlagen in conformance with the increasing significance of photochemistry, indicated by the high solar radiation and high temperature, and the low relative humidity and low wind speed. The explained variance r² reaches its maximum with more than 50 % of the time in Munich center. This empirical model is applicable to the forecasting of local ozone maximum concentrations with a total standard error deviation of 8.5 to 12.8 % and, if ozone concentrations exceed 80 ppb, with a standard error deviation of 5.4 to 9.5 %.     

A hierarchical Bayesian model to estimate and forecast ozone through space and time

A Bayesian hierarchical regime switching model describing the spatial–temporal behavior of ozone (O₃) within a domain covering Lake Michigan during spring–summer 1999 is developed. The model incorporates linkages between ozone and meteorology. It is specifically formulated to identify meteorological regimes conducive of high ozone levels and allow ozone behavior during these periods to be different from typical ozone behavior. The model is used to estimate or forecast spatial fields of O₃ conditional on observed (or forecasted) meteorology including temperature, humidity, pressure, and wind speed and direction. The model is successful at forecasting the onset of periods of high ozone levels, but more work is needed to also accurately identify departures from these periods.     

Air Quality Modeling in the South Coast Air Basin of California: What Do the Numbers Really Mean?

Abstract

This study evaluates air quality model sensitivity to input and to model components. Simulations are performed using the California Institute of Technology (CIT) airshed model. Results show the impacts on ozone (O₃) concentration in the South Coast Air Basin (SCAB) of California because of changes in: (1) input data, including meteorological conditions (temperature, UV radiation, mixing height, and wind speed), boundary conditions, and initial conditions (ICs); and (2) model components, including advection solver and chemical mechanism. O₃ concentrations are strongly affected by meteorological conditions and, in particular, by temperature. ICs also affect O₃ concentrations, especially in the first 2 days of simulation. On the other hand, boundary conditions do not significantly affect the absolute peak O₃ concentration, although they do affect concentrations near the inflow boundaries. Moreover, predicted O₃ concentrations are impacted considerably by the chemical mechanism. In addition, dispersion of pollutants is affected by the advection routine used to calculate its transport. Comparison among CIT, California Photochemical Grid Model (CALGRID), and Urban Airshed Model air quality models suggests that differences in O₃ predictions are mainly caused by the different chemical mechanisms used. Additionally, advection solvers contribute to the differences observed among model predictions. Uncertainty in predicted peak O₃ concentration suggests that air quality evaluation should not be based solely on this single value but also on trends predicted by air quality models using a number of chemical mechanisms and with an advection solver that is mass conservative.     

Air pollution characteristics associated with mesoscale atmospheric patterns in northwest continental Europe

An impact related daily air quality index (DAQx), calculated for 15 air quality monitoring stations (traffic, background, and industry) in Belgium, France, Germany and Luxembourg, was compared to mesoscale atmospheric patterns between 2001 and 2007. Meteorological conditions were described by the Hess and Brezowsky synoptic weather classification system and gridded data of the EU FP6 ENSEMBLES project of total precipitation and mean surface temperature. DAQx values indicate sufficient to poor air quality in the urban area of Brussels and at urban traffic stations, as well as satisfactory air quality at the background stations. The air quality index refers to more than 90% to the presence of high PM₁₀, O₃ and NO₂ concentrations. SO₂ and CO play only a minor role. The investigation of weather regimes indicates that zonal and mixed cyclonic circulation regimes are associated with better air quality than meridional and anticyclonic weather regimes. In general, weather regimes with high daily precipitation lead to better air quality than dryer air masses because of lower contribution of PM₁₀ to the air quality index. A trend analysis of weather regimes from 1978 to 2007 shows significant (α = 0.05) positive trends for weather classes associated with lower PM₁₀ concentrations. The results of a case study at a German station examining the relationship between PM₁₀ concentrations and local meteorological quantities (wind speed and precipitation) confirm the results of the regional analysis.     

Evaluation of the CAR-FMI model against measurements near a major road

A field measurement campaign was conducted near a major road in southern Finland from September 15 to October 30, 1995. The concentrations of NO, NO₂ and O₃ were measured simultaneously at three locations, at three heights (3.5, 6 and 10 m) on both sides of the road. Traffic densities and relevant meteorological parameters were also measured on-site. We have compared measured concentration data with the predictions of the road network dispersion model CAR-FMI, used in combination with a meteorological pre-processing model MPP-FMI. In comparison with corresponding results presented previously in the literature, the agreement of measured and predicted datasets was good, as measured using various statistical parameters. For all data (N=587), the index of agreement (IA) was 0.83, 0.82 and 0.89 for the measurements of NO_x, NO₂ and O₃, respectively. The IA is a statistical measure of the correlation of the predicted and measured time series of concentrations. However, the modelling system overpredicts NO_x concentrations with a fractional bias FB=+13%, and O₃ concentrations with FB=+8%, while for NO₂ concentrations FB=−2%. We also analyzed the difference between model predictions and measured data in terms of meteorological parameters. Model performance clearly deteriorated as the wind direction approached a direction parallel to the road, and for the lowest wind speeds. The range of variability concerning atmospheric stability, ambient temperature and the amount of solar radiation was modest during the measurement campaign. As expected, no clear dependencies of model performance were therefore detected in terms of these parameters. The experimental dataset is available for the evaluation of other roadside dispersion models.     

Measurement of surface ozone and its precursors in an urban area in South Brazil

Weekly and seasonal variations of surface ozone and their precursors – nitrogen oxides, carbon monoxide-associated with meteorological parameters (wind direction, temperature, solar radiation) – are reported. Measurements were performed continuously during 2006 at two sampling stations located in the metropolitan area of Porto Alegre, Brazil. Results have shown that O₃ concentrations remained almost constant between weekdays. Levels of NO_x precursors decreased especially on Sundays, due to lighter traffic. The seasonal variation has shown a maximum O₃ concentration during summer and spring while NO_x and NO₂ have maxima at the colder months. The daily cycle of highest ozone concentrations reveals a lower nightly level and an inverse relation between O₃ and NO_x, evidencing the photochemical formation of O₃. There are seasonal variation and source heterogeneity.     

Variation of ozone concentrations in the ambient air of Baghdad

Measurements of ozone concentration in the ambient air of the city of Baghdad were carried out for the period October 1983 to October 1984. The O₃, probably of local origin, showed a typical diurnal and seasonal variation. Maximum daily 1-h O₃ concentrations higher than the international ambient air quality standards were observed regularly during the summer months. High O₃ concentrations during the night were also observed. Scatter diagrams were used to relate the O₃ concentrations with temperature, solar radiation and humidity.     

Emissions into the Atmosphere Due to the Ceramic Salt Glazing Process

The techniques of Principal Component Analysis (PCA) and subsequent regression analysis were used in an attempt to describe local and upwind chemical and physical factors which affect the variability of SO₄ ^–2 concentrations observed in a rural area of the northeastern U.S. The data used in the analyses included upwind and local O₃ concentrations, temperature, relative humidity and other climatological information, SO₂, and meteorological information associated with backward trajectories. The investigation identified five principal components, three major (eigenvalues >1) and two minor (eigenvalues < one), which accounted for 52% (r = 0.72) of the variability in the SO₄ ^–2 regression model. These components can be described as representing local and upwind photochemistry, droplet growth, SO₂ emissions, and air mass characteristics. The study also indicated that in future studies it will be necessary to a priori select air pollution and meteorological variables for measurement to potentially increase the sensitivity of this type of receptor model.     

An assessment of the impact of climate change on air quality at two UK sites

Possible effects of climate change on air quality are studied for two urban sites in the UK, London and Glasgow. Hourly meteorological data were obtained from climate simulations for two periods representing the current climate and a plausible late 21st century climate. Of the meteorological quantities relevant to air quality, significant changes were found in temperature, specific humidity, wind speed, wind direction, cloud cover, solar radiation, surface sensible heat flux and precipitation. Using these data, dispersion estimates were made for a variety of single sources and some significant changes in environmental impact were found in the future climate. In addition, estimates for future background concentrations of NO_x, NO₂, ozone and PM₁₀ upwind of London and Glasgow were made using the meteorological data in a statistical model. These showed falls in NO_x and increases in ozone for London, while a fall in NO₂ was the largest percentage change for Glasgow. Other changes were small. With these background estimates, annual-average concentrations of NO_x, NO₂, ozone and PM₁₀ were estimated within the two urban areas. For London, results averaged over a number of sites showed a fall in NO_x and a rise in ozone, but only small changes in NO₂ and PM₁₀. For Glasgow, the changes in all four chemical species were small. Large-scale background ozone values from a global chemical transport model are also presented. These show a decrease in background ozone due to climate change. To assess the net impact of both large scale and local processes will require models which treat all relevant scales.     

Some Empirical Models for the Los Angeles Photochemical Smog Data

This paper presents (i) an empirico-mechanistic model which describes the dependence of CO, NO, NO₂, and O₃ on total hydrocarbons, traffic, wind speed, inversion base height, and solar radiation as well as the photochemical reactions associated with these pollutants; (ii) a detailed study of weather conditions when the instantaneous daily maximum O₃ exceeds the L.A. County alert level of 50 pphm; and (iii) regression models for the prediction of daily maximum O₃ values.     

Impacts of Meteorological Factors on PM10: Artificial Neural Networks (ANN) and Multiple Linear Regression (MLR) Approaches

In this study, prediction capacities of multi-linear regression (MLR) and artificial neural networks (ANN) onto coarse particulate matter (PM₁₀) concentrations were investigated. Different meteorological factors on particulate pollution were chosen for operating variables in the model analyses. Two different regions (urban and industrial) were identified in the region of Kocaeli, Turkey. All data sets were obtained from air quality monitoring network of the Ministry of Environment and Urban Planning, and 120 data sets were used in the MLR and ANN models. Regression equations explained the effects of the meteorological factors in MLR analyses. In the ANN model, backpropagation network with two hidden layers has achieved the best prediction efficiency. Determination coefficients and error values were examined for each model. ANN models displayed more accurate results compared to MLR.     

Prediction of Ambient PM10 and Toxic Metals Using Artificial Neural Networks

Abstract

In this study, an artificial neural network is employed to predict the concentration of ambient respirable particu-late matter (PM₁₀) and toxic metals observed in the city of Jaipur, India. A feed-forward network with a back-propagation learning algorithm is used to train the neural network the behavior of the data patterns. The meteorological variables of wind speed, wind direction, relative humidity, temperature, and time are taken as input to the network. The results indicate that the network is able to predict concentrations of PM₁₀ and toxic metals quite accurately.     

Neural Network Model for Predicting Peak Photochemical Pollutant Levels

ABSTRACT

In this paper, an attempt is made for the 24-hr prediction of photochemical pollutant levels using a neural network model. For this purpose, a model is developed that relates peak pollutant concentrations to meteorological and emission variables and indexes. The analysis is based on measurements of O₃ and NO₂ from the city of Athens. The meteorological variables are selected to cover atmospheric processes that determine the fate of the airborne pollutants while special care is taken to ensure the availability of the required input data from routine observations or forecasts. The comparison between model predictions and actual observations shows a good agreement. In addition, a series of sensitivity tests is performed in order to evaluate the sensitivity of the model to the uncertainty in meteorological variables. Model forecasts are generally rather insensitive to small perturbations in most of the input meteorological data, while they are relatively more sensitive in changes in wind speed and direction.     

Application of Artificial Neural Networks to Modeling and Prediction of Ambient Ozone Concentrations

ABSTRACT

The deterministic modeling of ambient O₃ concentrations is difficult because of the complexity of the atmospheric system in terms of the number of chemical species; the availability of accurate, time-resolved emissions data; and the required rate constants. However, other complex systems have been successfully approximated using artificial neural networks (ANNs). In this paper, ANNs are used to model and predict ambient O₃ concentrations based on a limited number of measured hydrocarbon species, NO_x compounds, temperature, and radiant energy. In order to examine the utility of these approaches, data from the Coastal Oxidant Assessment for Southeast Texas (COAST) program in Houston, TX, have been used. In this study, 53 hydrocarbon compounds, along with O₃, nitrogen oxides, and meteorological data were continuously measured during summer 1993. Steady-state ANN models were developed to examine the ability of these models to predict current O₃ concentrations from measured VOC and NO concentrations. To predict the future concentrations of O₃, dynamic models were also explored and were used for extraction of chemical information such as reactivity estimations for the VOC species.

The steady-state model produced an approximation of O₃ data and demonstrated the functional relationship between O₃ and VOC-NO_x concentrations. The dynamic models were able to the adequately predict the O₃ concentration and behavior of VOC-NO_x-O₃ system a number of hourly intervals into the future. For 3 hr into the future, O₃ concentration could be predicted with a root-mean squared error (RMSE) of 8.21 ppb. Extending the models further in time led to an RMSE of 11.46 ppb for 5-hr-ahead values. This prediction capability could be useful in determining when control actions are needed to maintain measured concentrations within acceptable value ranges.