Investigation of Road Transport-Based Greenhouse Gas Prediction Models and the Use of Intelligent Transportation Systems for Emission Reduction

Investigation of Road Transport-Based Greenhouse Gas Prediction Models and the Use of Intelligent Transportation Systems for Emission Reduction | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Investigation of Road Transport-Based Greenhouse Gas Prediction Models and the Use of Intelligent Transportation Systems for Emission Reduction Hande Beba, Zübeyde Öztürk This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7194240/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 18 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted 16 You are reading this latest preprint version Abstract Based on the modeling results, the carbon equivalent values for each pollutant were calculated. In cases where transportation-related emissions exceed regulatory thresholds, an ITS-based traffic management strategy was proposed to redirect vehicles to alternative routes. Transportation sector is among the critical domains where effective public interventions and adaptive strategies are essential to mitigate CO₂ emissions. In this context, artificial intelligence (AI)-based models are increasingly utilized to support emission reduction efforts. This study focuses on developing a predictive model for road transport-related greenhouse gas emissions in Dilovası, a district of Kocaeli Province known for its high levels of air pollution. Two AI-based approaches, Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), were employed to model pollutant emissions. At the same time, the potential contribution of Intelligent Transportation Systems (ITS) to emission mitigation and climate change adaptation was also examined. Initially, NO x and CO emissions from light and heavy vehicles were modeled using ANFIS and ANN, and the results were compared with outputs from the COPERT 4 (Calculations of Emissions from Road Transport) software. The high adaptability of the ANFIS model allowed for a more accurate representation of the influence of environmental variables and vehicle counts on emission levels. Earth and environmental sciences/Climate sciences Earth and environmental sciences/Environmental sciences Physical sciences/Mathematics and computing Artificial intelligence Climate change Greenhouse gas emission Intelligent transportation systems Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Greenhouse gases (GHGs) are components of the Earth's atmosphere that, while allowing sunlight to pass through, trap heat by preventing thermal energy emitted from the Earth's surface from escaping into space. As such, GHGs act as a heat trap, contributing to the rise in global temperatures. The most significant greenhouse gases in the Earth's atmosphere include carbon dioxide (CO₂), water vapor (H₂O), methane (CH₄), nitrous oxide (N₂O), and ozone (O₃) [1]. Global warming and climate change, as adverse by-products of greenhouse gas (GHG) emissions, have recently emerged as two of the most widely debated issues not only due to their detrimental effects on ecosystems but also because of their increasingly harmful consequences for humanity. As the world intensifies its efforts to combat climate change, reducing transportation-related emissions stands out as one of the most critical contributions to this endeavor. Over the past century, the transportation sector has become one of the largest contributors to the global carbon footprint. Today, nearly 15% of energy-related CO₂-equivalent emissions originate from the transportation sector, with road transport being responsible for the majority of these emissions [2]. With rapidly increasing levels worldwide, carbon emissions from the transportation sector are projected to continue growing. In the absence of aggressive and sustainable policy interventions, transportation-related greenhouse gas emissions are estimated to have the potential to double by the year 2050 [3]. Urban road networks that accommodate high volumes of vehicular traffic significantly contribute to the deterioration of air quality and the generation of substantial amounts of greenhouse gases. [4]. This situation is further exacerbated under stop-and-go congested traffic conditions, which increase emissions and local air pollution due to low speeds and pollutant dispersion. Global urbanization is projected to reach 60% by 2030 [5], therefore, urban areas are expected to continue contributing disproportionately to road transport emissions relative to their geographic size. Therefore, efforts to mitigate road traffic emissions are of paramount importance. Achieving this requires a comprehensive understanding and accurate modeling of traffic emission rates. Numerous air quality models exist in the literature for predicting air quality, which is both influenced by and influences climate change. The Calculations of Emissions from Road Transport (COPERT 4) model, developed by the Environmental Protection Agency (EPA), is widely used for estimating road transport emissions in official annual national inventories. [6]. A key distinction of the COPERT 4 model compared to some traditional models is its capability to estimate cold start, hot running, and non-exhaust emissions from all road vehicle categories [7]. Conventional air quality models, which employ numerical and mathematical techniques to simulate the physical and chemical processes influencing air pollutants, have been widely utilized to assist in the design of effective air pollution mitigation strategies. However, in recent years, artificial intelligence methods such as Fuzzy Logic (FL) and Artificial Neural Networks (ANN) have gained prominence due to their greater adaptability and flexibility CITATION Han142 \l 1055 [8] . Previous studies [9], have demonstrated that traditional air quality models commonly used in the literature possess significant drawbacks, including the incorporation of highly complex inputs and extended computational times. In contrast, artificial intelligence methods, particularly the Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), have been increasingly employed for air quality prediction in recent years. Notably, ANFIS has established itself as an effective predictive model not only among neuro-fuzzy systems but also across a range of other machine learning techniques [10]. ANFIS is an adaptive network that integrates neural networks with fuzzy logic principles and is capable of processing both linear and nonlinear parameters [11]. A number of researchers have modeled CO₂ emissions across various domains using ANFIS. Environmental factors such as temperature, emission levels, and air quality have been successfully predicted using ANFIS, which has demonstrated strong performance [12]. Intelligent Transportation Systems (ITS) are an integrated blend of software, hardware, traffic engineering concepts, and communication technologies designed to enhance the efficiency and safety of transportation systems [13], and are widely used for incident detection, ramp metering, traffic signal control, parking management, speed detection, travel planning, passenger information, route guidance, variable message sign display, and vehicle violation detection. In this context, fiber optic communication, radio frequency identification (RFID), dedicated short-range communications (DSRC), cellular communication networks, data systems, and location-map information applications provide drivers with real-time traffic information [14]. The traffic monitoring system is an integral component of Intelligent Transportation Systems (ITS), and one of its core functions is vehicle classification. In vehicle classification, in-road sensors offer high classification accuracy due to their proximity to passing vehicles, allowing them to effectively capture vehicle body characteristics and dynamic responses. Examples of in-road sensors include inductive loop sensors, vibration sensors, magnetic sensors, and piezoelectric cables [15]. There is a strong correlation between Intelligent Transportation Systems (ITS) and the control of vehicle emissions, with road transport being the largest source of transportation-related greenhouse gas emissions. Therefore, this study focuses on estimating the quantities of emissions (NO x and CO) originating from road transport and examining the potential positive impact of ITS on climate change. For this purpose, road transport emissions in the Dilovası district of Kocaeli Province were modeled using Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), and the prediction results were compared with those generated by the COPERT 4 model. Based on the evaluation, and by considering the carbon equivalents of emissions, an ITS-based strategy was proposed for pollution mitigation through the redirection of vehicles to alternative routes in cases where road transport emission limits are exceeded in the study area. The model results were evaluated under two different scenarios, specifically focusing on emission reduction from traffic sources, to assess the applicability of the proposed system and its potential contribution to climate change mitigation. The primary aim of the study is to model transportation-related emissions in the Dilovası district—selected as the case area due to its high air pollution levels—and to investigate the feasibility of using ITS for emission reduction in the region. In doing so, the study seeks to highlight the environmental benefits of ITS by emphasizing its impact on vehicle emission control and its contributions to transportation efficiency, economic improvement, and climate change mitigation. Data and methods Study area and data sources The models were trained using four years of traffic data (2015–2018) from the Dilovası district of Kocaeli Province. Dilovası (40° 50′ 04″ N, 29° 35′ 22″ E) is a district characterized by a high concentration of industrial facilities and severely polluted air (Figure 1). The district's basin-like topographical structure causes emissions released from factory chimneys to have long-term impacts on the local population [16]. Moreover, vehicles traveling on the D-100 Highway and the E-80 TEM Motorway, which pass through the district and accommodate various classes of vehicles, constitute a significant source of greenhouse gas emissions and air pollution in the area CITATION Züb19 \l 1055 [17] . Data collection and classification For model development, both training and testing datasets were constructed using the number of vehicles passing through the district by type, along with environmental variables, including monthly average daily minimum temperature (℃), maximum temperature (℃), and average relative humidity (%), collected from the Kocaeli Cengiz Topel Naval Air Station Meteorological Station (MeteoBlue, 2023). To determine the monthly average daily data for all vehicle types passing through the district, vehicles were initially categorized into two main groups: light and heavy vehicles. For a more detailed analysis, both light and heavy vehicles were further classified into specific vehicle types based on distribution ratios obtained from the Turkish Statistical Institute (TÜİK). Passenger cars were categorized into three subgroups according to fuel type: gasoline, diesel, and LPG. Medium-duty commercial vehicles, including passenger cars, pickup trucks, and minibuses, were classified as "light commercial vehicles," while buses, trucks, and articulated lorries were considered "heavy vehicles." It was assumed that all heavy vehicles operate on diesel fuel. The statistical data necessary to estimate vehicle numbers by type and composition were obtained from the General Directorate of Highways (KGM). After acquiring environmental data, more detailed vehicle data were generated using ratios obtained from relevant institutions. The COPERT 4 emission estimation program was utilized to predict transportation-related emissions, while ANFIS and ANN tools modeled greenhouse gas emissions by employing the necessary environmental and vehicle data to produce prediction results. The workflow for model development is illustrated in Figure 2. Raw data were initially subjected to preprocessing. Except for meteorological variables, data related to vehicle classification inputs underwent processing before modeling. Model definition of the transportation-related air pollution modeling system This section introduces the transportation-related air pollution modeling system and details the specific components of each sub-model. The framework of the modeling system is illustrated in Figure 2. The COPERT 4 model was employed to estimate emissions generated by different vehicle types. The ANFIS and ANN models utilize vehicle and environmental data obtained from COPERT 4 to apply artificial intelligence techniques, producing vehicle emission outputs. Consequently, air pollution emission data are processed by the emission models. This modeling system is used for projecting transportation-related greenhouse gas emissions. COPERT 4 model COPERT 4 has been developed to estimate cold start emissions by incorporating various parameters such as ambient temperature, vehicle usage patterns, and the cold-to-hot start ratio. Additionally, COPERT 4 has established methodologies to adjust the base hot emission factors based on vehicle mileage and fuel characteristics [6]. For improved accuracy and reliability, classified data were captured and used to simulate future transportation-related NO x and CO pollutant parameters in Dilovası. In the study, vehicle counts and environmental information passing through the district were processed appropriately for the model. The COPERT 4 framework utilized in this research is illustrated on the right side of Figure 2. Considering the characteristics of the model outputs, the necessary vehicle and environmental data were disaggregated by month over multiple years and processed for emission estimation. According to the study by (Tulyasuwan & Federici, 2011), the use of Tier 2 in the COPERT 4 model is recommended. Furthermore, the demonstrated effectiveness of time series emission results (Tongwane, et al., 2015) has influenced their implementation in the present study. Artificial intelligence models The input data used in the study were compared against prediction results from different models. Accordingly, various artificial intelligence modeling tools were employed to process vehicle and environmental data and generate emission estimates for the transportation-related air quality model. The first of these is the Adaptive Neuro-Fuzzy Inference System (ANFIS) model. Literature reviews indicate that when emission values are modeled using ANFIS and Artificial Neural Network (ANN), ANFIS predictions tend to be more reliable [18]. The second model employed is ANN, which is capable of analyzing large datasets and modeling complex processes, thereby enabling more accurate emission predictions. This capability makes ANN a critical tool for the development of effective emission reduction strategies [19]. The ANFIS model was implemented using the "ANFIS" toolbox available in MATLAB 2021b. For both pollutant parameters, 80% of the dataset was allocated for training, while the remaining 20% was used for testing. Increasing the number of subclusters resulted in higher test errors; therefore, the network was configured with two subclusters. The optimal number of training epochs was set to 20, with an error tolerance of zero. For the ANN model, a network architecture with 10 inputs and 1 output, similar to the ANFIS model, was designed, utilizing a feedforward backpropagation network. Various combinations of training and learning functions were tested, with the TRAINLM training function yielding the best performance. To prevent overfitting and unnecessary complexity, the number of hidden neurons was not increased during model development. Intelligent transportation system implementation Introduction of the routes to which vehicles will be redirected Within the scope of the study, two alternative routes expected to contribute to the reduction of vehicle-related emissions in the Dilovası district were introduced: the Designed Alternative Route (DAR) and the Northern Marmara Motorway (NMM) (Figure 3). The alternative route options were designed using the Microstation Inroads software, by the project design criteria established by [20]. Among three proposed alternatives, the optimal route was determined to be 9.32 kilometers in length, based on the evaluation of various parameters, including total route length, viaduct length, tunnel length, maximum viaduct pier height, pier edge length, span between viaduct piers, as well as cut and fill volumes. The other route, the 170.2 km long Northern Marmara Motorway, Kurtköy-Akyazı Section (including connection roads) (KMO), provides uninterrupted connectivity from Kurtköy to Akyazı. In this study, a 41.35 km segment of the motorway, located near the district boundaries and marked in Figure 4, was taken into consideration. Based on data from the [20], the construction cost of the alternative route was estimated at $5,172,229, while the total construction cost of the KMO was reported to be $2,200,000,000. Assessment of environmental factors and selection of the most suitable intelligent transportation system In the district, there is a need for a system that provides drivers with information in the event of air pollution limit exceedances to facilitate redirection to alternative routes. It is planned that this will be achieved through Variable Message Signs (VMS). The VMS will be capable of dynamically reflecting the prevailing air pollution conditions during vehicle passage. This system is intended to notify drivers of incidents during air pollution limit exceedance events and to provide guidance on alternative routing options. One of the key criteria in selecting the sensor type for vehicle redirection via VMS in the Dilovası district has been the meteorological conditions. Meteorological assessments for Kocaeli Province indicate that precipitation occurs in nearly half of the months throughout the year CITATION Met \l 1055 [21] . The district’s unfavorable meteorological conditions may limit the standalone deployment of sensors such as cameras, LIDAR, radio, and Wi-Fi, which are highly sensitive to weather variations. Additionally, inductive loop systems were deemed unsuitable for exclusive use due to their inadequate performance on high-traffic volume roads. In the studied district, the objective is to redirect vehicles to alternative routes when air pollution limits are exceeded. Drivers who do not comply with the diversion warnings will be detected through the Electronic Enforcement System (EES) and subjected to penalties. For this purpose, a camera-based system will be required. Given the significant impact of meteorological conditions on this system, it is suggested that its use in combination with another system, rather than standalone, will yield more effective results. The integration of the selected sensor system with both VMS and EES has been identified as the most suitable approach for reducing air pollution. A cost comparison among various sensor technologies can only be made when the specific application is clearly defined [22]. In selecting an appropriate sensor for the intended application, factors such as ease of installation and maintenance, as well as design requirements, must be taken into consideration. Sensor selection for a particular application depends on several criteria, including data parameters, data accuracy, detection range, suitable data transmission medium, location, specific installation requirements, initial cost, and the acceptability of the maintenance burden imposed by the sensor CITATION ANe \l 1055 [23] . Table 1. Cost and Emission Data for Pollutant Parameters Cost Information Cost ($) Sensor Cost 13.430 Construction Cost of the Designed Route 5.172.229 Total Cost 5.185.659 Emission Information Quantity (ton) Annual Emission Quantity for the NO x Pollutant Parameter 48 Daily Average Emission Quantity for the NO x Pollutant Parameter 0,1315 Annual Emission Quantity for the CO Pollutant Parameter 173 Daily Average Emission Quantity for the CO Pollutant Parameter 0,4735 In the long term, excluding the maintenance, repair, and operational costs of Intelligent Transportation Systems (ITS), an evaluation was conducted based on the approximate initial total costs of inductive loop sensors, magnetic sensors, microwave sensors, vibration sensors, LIDAR, video image processors, and Weigh-In-Motion (WIM) sensors [24], along with the daily average threshold emission quantity calculated per sensor cost (Table 1). Considering these factors, the most suitable sensor system determined to contribute to the reduction of transportation-related greenhouse gas emissions in the district was the combined system of Vibration Sensor + Variable Message Signs (VMS) + Electronic Enforcement System (EES). Analysis of the depreciation period for the proposed ıntelligent transportation system In calculating the number of years required for emission cost savings to offset the total cost of the proposed alternative route and Intelligent Transportation System (ITS), the maximum values of daily average emission results for NO x and CO parameters obtained from the COPERT 4 program were used. These values were evaluated in terms of CO₂ equivalent (CO₂eq). In this context, the COPERT 4 output yielded a total of 960 tons of CO₂eq, while the daily average threshold emission quantity calculated based on sensor costs was 235 tons of CO₂eq. According to the evaluations, if ITS is implemented, the payback period in which CO₂eq-related cost savings offset the total investment cost is found to be less than one year. This indicates that the system would be highly efficient in terms of cost-performance. Results The COPERT 4 model output is significantly influenced by vehicle counts, vehicle types, and environmental data inputs. Therefore, in this section, we evaluate the performance of the COPERT 4 and artificial intelligence models by examining their capabilities using vehicle and environmental data, providing an assessment of model applicability. A previous study [25] observed that NO x concentrations begin to increase in September, following the end of the summer months. The COPERT 4 model results indicated the highest daily average values in September 2016, with NO x emissions reaching 0.84 tons and CO emissions 4.77 tons. These findings effectively capture the intense traffic-related air pollution occurring in the district at the end of summer. Based on the performance of the COPERT 4 model, vehicle and environmental data were utilized as primary input variables for artificial intelligence models. The ANFIS and ANN models were then applied to predict future transportation-related NO x and CO pollutant concentrations in the Dilovası district. Input and output data from both models were processed through the COPERT 4 program to ensure model validation. The effectiveness of the model developed and trained within the ANFIS workspace can be evaluated by comparing the model’s predictions on historical data with actual observed values. The relationship between the outputs obtained from the model for both pollutant parameters and the data calculated via the COPERT 4 program is presented in Table 1. The determination coefficients approaching a value of 1 for the training data of both pollutant parameters indicate a strong correlation between the two models. The relationship between the outputs obtained from the ANN model for each pollutant parameter and the data calculated using the COPERT 4 program was examined, and a regression analysis was conducted. First, the training data were simulated, and subsequently, test data were used to obtain predicted emission values for ten input sets previously unseen by the network. Similar evaluations were carried out for the CO pollutant parameter. When comparing the performance of both models, the ANFIS model was observed to yield more accurate results than the ANN model in predicting vehicle-related emissions (Table 2). Table 2. Comparison of R² Values for Model Results MODEL Data NO X (R 2 ) CO (R 2 ) ANFIS Training 0,99 0,99 Testing 0,89 0,94 ANN Training 0,64 0,85 Testing 0,36 0,72 According to the model results, a high R² value indicates a good fit of the regression model, which can be attributed to the adaptive nature of the ANFIS model. Discussion Investigation of the contribution of vehicle emissions to climate change and greenhouse gases Greenhouse gas impact according to different vehicle routing scenarios In this section, the impact on greenhouse gas emissions of scenarios in which light vehicles are separately routed as passenger cars and light commercial vehicles is examined. To this end, the number of vehicles calculated using threshold emission values expressed in carbon dioxide equivalent for CO and NO x pollutant parameters was selected as the emission source for the ANFIS model to evaluate its performance and investigate future greenhouse gas emissions in the Dilovası district. The study specifically focuses on the effect of changes in vehicle types on air pollution under transportation-related greenhouse gas impacts. Air pollution control strategies represent the initial step of this research and do not include other emission sources, such as those from industrial activities, that contribute to climate change. Calculation of the ratio of vehicles to be routed based on threshold emission values of CO and NO x pollutant parameters expressed in CO₂ equivalent This section explains the determination of the number of vehicles to be routed based on CO₂ equivalent according to the model results. Specifically, it investigates at which vehicle count the ITS would be activated when threshold emission values expressed in CO₂ equivalents are exceeded according to the model outputs. The annual average daily NO x emission obtained from the ANFIS model is 0.504 tons, and the CO emission is 0.690 tons. For the selected ITS setup, consisting of the Vibration Sensor + EES + VMS combination, threshold emission values were determined as 0.131 tons for NO x and 0.473 tons for CO. To establish the vehicle routing ratio, the annual average daily emission values obtained from the ANFIS model for each pollutant parameter were divided by the daily average threshold emission amounts calculated based on sensor costs. In the initial step of the assessment, the carbon dioxide equivalent per unit of the NO x pollutant was taken as 900, and for the CO pollutant as 250 [ 26 ]. Accordingly, the model-derived NO x amount corresponded to 453.6 tons CO₂ equivalent, and the CO amount to 172.5 tons CO₂ equivalent, resulting in a total emission carbon dioxide equivalent of 626.1 tons based on the ANFIS model output. In the subsequent step, the daily average threshold emission values calculated based on sensor costs were converted into carbon dioxide equivalents for each pollutant parameter to determine threshold emission values. The threshold emission values obtained were determined as 118.39 tons CO₂ equivalent for NO x and 118.39 tons CO₂ equivalent for CO, yielding a total carbon dioxide equivalent of 236.78 tons. When the ratio of the total emission carbon dioxide equivalent obtained from the ANFIS model to the total threshold emission carbon dioxide equivalent calculated from sensor costs is considered, it is observed that the threshold emission values expressed in carbon dioxide equivalent for pollutant parameters correspond to approximately 38% of the model result. Thus, when the annual average daily vehicle count in the district reaches approximately 38% of this value, directing vehicles to alternative routes is deemed appropriate for air pollution mitigation purposes. Investigation of greenhouse gas emissions under the scenario of light vehicle redirectering This section evaluates the impact on greenhouse gas emissions of scenarios in which light vehicles are separately rerouted to alternative routes, categorized as passenger cars and medium commercial vehicles. The impact of greenhouse gas emissions under the DAR scenario The DAR-PC (Designed Alternative Route - Passenger Car) scenario refers to the redirection of passenger cars, while the DAR-MCV (Designed Alternative Route - Light Commercial Vehicle) scenario denotes the redirection of light commercial vehicles to the designed alternative route. According to the results of these scenarios, the total fuel consumption cost amounted to $ 7,564,509 for passenger cars and $ 6,424,000 for light commercial vehicles. This indicates that the fuel consumption cost in the DAR-PC scenario is approximately 1.15 times higher than that of the DAR-MCV scenario. However, considering the route lengths of 9.327 km for DAR and 41.35 km for NMM, the emission reduction in the DAR-PC scenario was approximately three times greater than that observed in the DAR-MCV scenario (Fig. 4 ). It was concluded that, within the DAR scenario, passenger cars exhibit a greater emission reduction capacity compared to light commercial vehicles, relative to their respective fuel consumption costs. Impact of greenhouse gas emissions under the NMM scenario This section refers to the NMM-PC (Northern Marmara Motorway–Personal Car) scenario, which directs passenger cars, and the NMM-MCV (Northern Marmara Motorway–Medium Commercial Vehicle) scenario, which directs medium commercial vehicles to the Northern Marmara Motorway. According to the scenario results, the total fuel consumption cost—calculated based on the number of vehicles, fuel price, fuel consumption per vehicle, and travel distance—was estimated at $ 33,536,231 for NMM-PC and $ 28,479,941 for NMM-MCV. When examining the emission reduction capacity in Fig. 5 , similar to the DAR scenario, the NMM-PC scenario yielded approximately three times the reduction compared to the NMM-MCV scenario. This outcome confirms that passenger cars have a greater potential for emission reduction. Investigation of the seasonal variation in transportation-related emission reduction The detailed seasonal variation in the capacity for reducing transportation-related greenhouse gas emissions in the district, based on the evaluation of all scenarios, is illustrated in Fig. 5 . Focusing on the analyzed CO 2 values, it is observed that the highest concentration profiles occur during the spring and summer months. Different scenarios exhibit similar seasonal patterns. The transportation-related greenhouse gas effect by season is characterized by a bell-shaped profile, peaking in the summer. The seasonal variability of CO 2 serves as a powerful tool for identifying emission sources in urban areas. Between spring and summer, traffic flow increases by 1.3 to 2 times, with all major citywide roads experiencing elevated traffic, making them primary emission sources [ 27 ]. Considering this, a marked seasonal variation in CO 2 levels is evident. This finding confirms that the long-distance transport of CO 2 primarily occurs during the summer months, as observed in the NMM-PC and NMM-MCV scenarios. In several studies [ 28 ], CO 2 emissions from light vehicles have been identified as the most significant source among the analyzed emissions. In this study, the capacity for reducing CO 2 emissions from passenger cars over long distances was found to be higher compared to that of medium commercial vehicles (Fig. 5 ). It is important to emphasize that this reflects the proportionality of emission production to the distance traveled. Figure 6 illustrates the impact of seasonal variations on CO 2 concentrations across different scenarios. The figure was generated using data collected during the same period for each scenario and subsequently aggregated. As shown, the seasonal emission reduction capacities for vehicles correspond to the DAR-PC, DAR-MCV, NMM-PC, and NMM-MCV scenarios, represented by the blue, orange, yellow, and gray box plots, respectively. For each scenario, the data spread during the winter months was the greatest, while the spring months exhibited the smallest spread. This indicates that the most heterogeneous data sets occur in winter, whereas the most homogeneous are observed in spring. Regardless of the designated routing, a decrease in CO 2 reduction capacity was observed in the MCV scenarios compared to the PC scenarios. Moreover, all scenarios showed a decline in CO 2 reduction capacity from autumn to winter. In the NMM-PC scenario, the transition from summer to autumn (Fig. 6 -c) exhibited a less pronounced effect on CO 2 reduction capacity. This phenomenon can be attributed to the sustained frequency of private car usage following the summer season. In relation to emission reduction capacity, the highest decrease in CO 2 was observed under the NMM-PC scenario across all seasons. This is due to emissions resulting from long-distance travel. Previous studies [ 29 ] have demonstrated a significantly strong correlation between air temperature and various pollutants and greenhouse gas emissions. In this study, the temperature-dependent effect of transportation-related greenhouse gas emissions is explained by the CO 2 reduction capacity observed in the PC scenarios, which results in substantial decreases in CO 2 concentrations. Conversely, scenarios involving medium commercial vehicles (MCV) (Figs. 6 -b, 6 -d) exhibited smaller reductions compared to the PC scenarios. This trend is attributed to the higher emission factors associated with private cars. The contribution of future transportation-related emission reductions to air quality in Dilovası District When comparing CO 2 levels under the DAR and NMM scenarios, directing vehicles to a more distant route has significantly contributed to the reduction of vehicle-related emissions in Dilovası district. In this context, the CO 2 reduction capacity under the NMM emission scenario yielded substantially better results than the DAR scenario. Despite the differences in emission reduction amounts, the fuel consumption costs were higher in the NMM scenario compared to the DAR scenario. Considering the contribution of emission reductions by vehicle type to greenhouse gases in the district, the CO 2 reduction capacity under the NMM-PC scenario is 38.6% higher than that of the NMM-MCV scenario (Fig. 6 ). Within the NMM scenario itself, the NMM-PC scenario demonstrated the lowest fuel consumption cost despite a high emission reduction rate. Similar to the results of the DAR-PC and DAR-MCV scenarios, these findings support the new regulations aiming for zero CO 2 emissions from new passenger cars by 2035 (European Comission, 2023), making it plausible that passenger car emissions are relatively higher than those from light commercial vehicles. According to the scenario outcomes, the average fuel consumption cost is higher for medium commercial vehicles, reflecting recent efforts to improve fuel efficiency in this vehicle category (Uluslararası Enerji Ajansı, 2021). Conclusions This study initially emphasizes the environmental impacts of transportation-related greenhouse gas emissions in Dilovası district of Kocaeli Province, an area characterized by extremely high pollution concentrations. Subsequently, as greenhouse gas levels exceed pollution limits, their CO 2 reduction capacities and seasonal distributions were evaluated. These were then compared with scenarios where vehicles were redirected to alternative routes, investigating the effect of CO 2 on air quality. The influence of various variables—such as vehicle type, season, and travel distance—was analyzed to identify the optimal configuration that achieves maximum CO 2 reduction alongside minimal fuel consumption costs. The presence of localized CO 2 concentrations exceeding pollution limits in all seasons, attributable to transportation, underscores a widespread environmental issue with the potential to trigger climate change. This highlights the critical importance of air pollution control in regions exhibiting high pollution concentrations. Given the active heavy and light vehicle traffic in the district and the proximity of highways to the district center, the study demonstrates how diverting vehicles away from the center toward alternative routes contributes to mitigating CO 2 pollution. Additionally, it was found that average fuel consumption costs are higher for medium commercial vehicles compared to passenger cars, reflecting recent efforts to improve fuel efficiency in this vehicle category. The analysis also revealed that seasonal factors play a significant role in the CO 2 reduction capacity within transportation dynamics, with CO 2 pollutant peaks observed during spring and summer. Such seasonal variations reflect the impact of heavily trafficked urban arterial roads serving as primary emission sources, underscoring the need for region-specific monitoring and pollution reduction strategies. Furthermore, the greatest CO 2 reduction was achieved in scenarios directing passenger cars to the Northern Marmara Motorway across all seasons, attributable to emissions associated with longer travel distances. The study advocates for stricter environmental regulations while targeting reductions in transportation-related greenhouse gas emissions. Research supporting pollution monitoring protocols and predictive models in cities with high pollution concentrations is essential to fully comprehend the environmental and climatic effects of CO 2 . Such research also guides policy development aimed at reducing the global greenhouse gas burden. Declarations Data Availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. Author contribution H.B. carried out the investigation, methodology, statistical analysis, validation, software implementation, and writing tasks. Z.Ö. contributed to the review and editing of the manuscript. The work has been reviewed by all authors. Funding This work was not supported by any funding. Competing interests The authors declare no competing interests. Additional information Correspondence and requests for materials should be addressed to H.B. References Nusa, K. & Kodak, G. Comparison of Maritime and Road Transportations in Emissions Perspective: A Review Article vol. 10pp. 48–60 (International Journal of Environment and Geoinformatics, 2023). Albuquerque, F. D., Maraqa, M. A., Chowdhury, R., Mauga, T. & Alzard, M. Greenhouse gas emissions associated with road transport projects: current status, benchmarking, and assessment tools, Transportation Research Procedia, vol. 48, no. pp. 2018–2030, 26–30 May 2019. (2020). Zhang, R., Fujimori, S., Dai, H. & Hanaoka, T. Contribution of the transport sector to climate change mitigation: Insights from a global passenger transport model coupled with a computable general equilibrium model (International Instute for Applied System Analysis, 2018). Gebre, T. & Nigussa, F. Greenhouse Gas Emission Reduction Measures in the Urban Road Transport Sector of Ethiopia (Environmental Progress & Sustainable Energy, 2019). United Nations. Policies on spatial distribution and urbanization have broad impacts on sustainable development, (2020). Wang, H. & McGlinchy, L. Review of vehicle emission modelling and the issues for New Zealand, (2009). Ntziachristos, L. & Samaras Z EMEP/EEA air pollutant emission inventory guidebook 2023, (2023). Sastry, H. G. & Marriboyina, V. A Novel Model for Air Quality Prediction using Soft Computing Techniques, ResearchGate , (2014). Lu Bai, J. W. X. M. H. L. Air Pollution Forecasts: An Overview. International J. Environ. Res. Public. Health , (2018). Salleh, M. N. M., Talpur, N. & Hussain, K. Adaptive Neuro-Fuzzy Inference System: Overview, Strengths, Limitations, and Solutions, ResearchGate , (2018). Alhindawi, R., Nahleh, Y. A., Kumar, A. & Shiwakoti, N. Application of a Adaptive Neuro-Fuzzy Technique for Projection of the Greenhouse Gas Emissions from Road Transportation, Sustainability , vol. 11, no. 6346, (2019). Khan, M. Z. & Khan, M. F. Application of ANFIS, ANN and fuzzy time series models to CO2 emission from the energy sector and global temperature increase. International J. Clim. Change Strategies Management , pp. 622–642, (2019). Alsabaan, M., Naik, K., Khalifa, T. & Nayak, A. Vehicular networks for reduction of fuel consumption and CO2 emission, in 8th IEEE International Conference on Industrial Informatics , Osaka, (2010). Karayolları, G. & Müdürlüğü Akıllı Ulaşım Sistemleri , Karayolları Genel Müdürlüğü, (2022). Won, M. Intelligent Traffic Monitoring Systems for Vehicle Classification: A Survey, USA, (2020). Türk, T. & Birliği Türk Tabipleri Birliği Dilovası Raporu, Türk Tabipleri Birliği Yayınları, (2012). Öztürk, Z., Arslantaş, O. A., Beba, E., Yılmaz, H. & Toros, H. M. and Air Pollution Reduction with Intelligent Transportation Systems: Dilovası Scenario. Journal Res. Atmospheric Science , (2019). Kunt, F., Ayturan, Z. C. & Dursun, S. Used Some Modelling Applications in Air Pollution Estimates. J. Int. Environ. Application Sci. 11 (4), 418–425 (2016). Cowls, J., Taddeo, M., Tsamados, A. & Floridi, L. The AI Gambit — Leveraging Artificial Intelligence to Combat Climate Change: Opportunities, Challenges, and Recommendations, ResearchGate , (2021). Karayolları, G. & Müdürlüğü Karayolları Tasarım El Kitabı, Karayolları Genel Müdürlüğü, (2016). Meteoroloji Genel & Müdürlüğü [Online]. (2024). Available: https://www.mgm.gov.tr/veridegerlendirme/il-ve-ilceler-istatistik.aspx?k=H FHWA, A. Summary of Vehicle Detection and Surveillance Technologies used in Intelligent Transportation Systems (Federal Highway Administration’s Intelligent Transportation Systems, 2007). United States Department of Transportation Federal Highway Administration,. [Online]. (2007). Available: https://highways.dot.gov/public-roads/novdec-2007/new-look-sensors Klein, L. A. & Roadside Sens. Traffic Manage. 22 , 01, (2022). Tübitak Dilovası Endüstri Bölgesi ve Çevresinde Hava Kirliliğine Neden olan Organik ve İnorganik Kirleticilerin Düzeylerinin ve Kaynaklarının Belirlenmesi, (2017). Avrupa Çevre, A. Costs of air pollution from European industrial facilities 2008–2017, (2021). Krylov, P. M., Volodin, O. N., Zaitsev, G. A., Nekrasova, L. P. & Klyuchnikov, D. A. Estimating Summer Emissions from Land Transportation vol. 16 (Asian Journal of Water, Environment and Pollution, 2019). Congressional Budget Office. Emissions of Carbon Dioxide in the Transportation Sector, U.S., (2022). Hoseinifar, S. E., Ashrafi, K. & Pour-Motlagh, M. S. The Effects of Seasonal Changes of Ambient Temperature and Humidity on Exhaust Pipe Emissions and Greenhouse Gases, 2383 , 4501, (2023). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 18 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 09 Oct, 2025 Reviews received at journal 08 Oct, 2025 Reviewers agreed at journal 08 Oct, 2025 Reviewers agreed at journal 05 Oct, 2025 Reviewers agreed at journal 03 Oct, 2025 Reviews received at journal 11 Sep, 2025 Reviews received at journal 07 Aug, 2025 Reviewers agreed at journal 06 Aug, 2025 Reviewers agreed at journal 06 Aug, 2025 Reviewers agreed at journal 05 Aug, 2025 Reviewers agreed at journal 04 Aug, 2025 Reviewers invited by journal 04 Aug, 2025 Editor assigned by journal 04 Aug, 2025 Editor invited by journal 01 Aug, 2025 Submission checks completed at journal 28 Jul, 2025 First submitted to journal 28 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7194240","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":496525619,"identity":"46525ac4-8ed3-44ad-96f6-88dbfa366cf2","order_by":0,"name":"Hande Beba","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxUlEQVRIiWNgGAWjYPCCA3JsYJqNaB0JB4zZ2KBaeIjVkthAtBZ+scPPHnz8cSe9T77HgOFD2WEGe+kD+LVIzk4zN5yR8Cy3jY3HgHHGucMMPHwJ+LUY3E4wk+ZJOAzWwszbBtRCyGX2t9O/Sf9JOJzOBtLylxgtBtI5ZtIMCYcTwFoYidEicTunTLIn7bBhG1tawcGec+k8PGcIaOGfnb5N4ofNYXn55sMbH/wos5Zj7yGgBQUcYCA6JkfBKBgFo2AU4AUAauQ6a3Ca29MAAAAASUVORK5CYII=","orcid":"","institution":"Istanbul Technical University","correspondingAuthor":true,"prefix":"","firstName":"Hande","middleName":"","lastName":"Beba","suffix":""},{"id":496525620,"identity":"fea4c0d8-2247-49c5-a638-eb74639d774c","order_by":1,"name":"Zübeyde Öztürk","email":"","orcid":"","institution":"Istanbul Technical University","correspondingAuthor":false,"prefix":"","firstName":"Zübeyde","middleName":"","lastName":"Öztürk","suffix":""}],"badges":[],"createdAt":"2025-07-23 08:53:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7194240/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7194240/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-29724-6","type":"published","date":"2025-12-18T15:58:05+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":88652958,"identity":"37379c10-2cd8-4133-a100-ae4242e56108","added_by":"auto","created_at":"2025-08-08 18:03:03","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":7146560,"visible":true,"origin":"","legend":"\u003cp\u003eDilovası District\u003c/p\u003e","description":"","filename":"Figure1TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/a4449a80134a94d5842141cf.png"},{"id":88652201,"identity":"f003ea7e-b01b-4130-bf63-a1861347d6b0","added_by":"auto","created_at":"2025-08-08 17:55:03","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":4944343,"visible":true,"origin":"","legend":"\u003cp\u003eFramework of the future transportation-related greenhouse gas research modeling system (left) and the data classification process (right).\u003c/p\u003e","description":"","filename":"Figure2TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/00de144eec369a10457ef052.png"},{"id":88652960,"identity":"04a0b4ee-c4ec-4d26-837d-fb3cca94ed98","added_by":"auto","created_at":"2025-08-08 18:03:03","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":29337962,"visible":true,"origin":"","legend":"\u003cp\u003eAlternative Routes for Vehicle Diversion in the Event of Air Pollution Limit Exceedance\u003c/p\u003e","description":"","filename":"Figure3TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/70582d41b479f4c687a4c868.png"},{"id":88652961,"identity":"394973a5-950e-455e-ac84-0c5e4097b941","added_by":"auto","created_at":"2025-08-08 18:03:04","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":342387,"visible":true,"origin":"","legend":"\u003cp\u003eEmission Reduction Capacity and Fuel Consumption Cost of Scenario Results\u003c/p\u003e","description":"","filename":"Figure4TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/11409b3c90b2874ebada8c88.png"},{"id":88652203,"identity":"f0ff0428-df62-448b-9eb7-b8b737a13a0b","added_by":"auto","created_at":"2025-08-08 17:55:03","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":694278,"visible":true,"origin":"","legend":"\u003cp\u003eAnalysis of Scenario Results in Terms of Seasonal Variation\u003c/p\u003e","description":"","filename":"Figure5TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/1ce8f043e7192855a97a1a90.png"},{"id":88652216,"identity":"ee62d386-f362-4134-bf91-091865f8c2cc","added_by":"auto","created_at":"2025-08-08 17:55:04","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":882935,"visible":true,"origin":"","legend":"\u003cp\u003eDiagram illustrating the seasonal variations of the DAR and NMM scenario results. Panels (a) and (b) depict the PC and MCV results under the DAR scenario, respectively, while panels (c) and (d) represent the PC and MCV results under the NMM scenario, respectively.\u003c/p\u003e","description":"","filename":"Figure6TIFF.png","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/51ef1445fda937cd56b6dff9.png"},{"id":98813990,"identity":"195c40b1-1a05-4c29-9a14-6d44ef454cca","added_by":"auto","created_at":"2025-12-22 16:09:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":40076733,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7194240/v1/2714ce73-03b2-40a5-bddb-a87a0ebe1a9d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Investigation of Road Transport-Based Greenhouse Gas Prediction Models and the Use of Intelligent Transportation Systems for Emission Reduction","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGreenhouse gases (GHGs) are components of the Earth\u0026apos;s atmosphere that, while allowing sunlight to pass through, trap heat by preventing thermal energy emitted from the Earth\u0026apos;s surface from escaping into space. As such, GHGs act as a heat trap, contributing to the rise in global temperatures. The most significant greenhouse gases in the Earth\u0026apos;s atmosphere include carbon dioxide (CO₂), water vapor (H₂O), methane (CH₄), nitrous oxide (N₂O), and ozone (O₃) [1]. Global warming and climate change, as adverse by-products of greenhouse gas (GHG) emissions, have recently emerged as two of the most widely debated issues not only due to their detrimental effects on ecosystems but also because of their increasingly harmful consequences for humanity. As the world intensifies its efforts to combat climate change, reducing transportation-related emissions stands out as one of the most critical contributions to this endeavor. Over the past century, the transportation sector has become one of the largest contributors to the global carbon footprint. Today, nearly 15% of energy-related CO₂-equivalent emissions originate from the transportation sector, with road transport being responsible for the majority of these emissions [2]. With rapidly increasing levels worldwide, carbon emissions from the transportation sector are projected to continue growing. In the absence of aggressive and sustainable policy interventions, transportation-related greenhouse gas emissions are estimated to have the potential to double by the year 2050 [3].\u003c/p\u003e\n\u003cp\u003eUrban road networks that accommodate high volumes of vehicular traffic significantly contribute to the deterioration of air quality and the generation of substantial amounts of greenhouse gases. [4]. This situation is further exacerbated under stop-and-go congested traffic conditions, which increase emissions and local air pollution due to low speeds and pollutant dispersion. Global urbanization is projected to reach 60% by 2030 [5], therefore, urban areas are expected to continue contributing disproportionately to road transport emissions relative to their geographic size. Therefore, efforts to mitigate road traffic emissions are of paramount importance. Achieving this requires a comprehensive understanding and accurate modeling of traffic emission rates.\u003c/p\u003e\n\u003cp\u003eNumerous air quality models exist in the literature for predicting air quality, which is both influenced by and influences climate change. The Calculations of Emissions from Road Transport (COPERT 4) model, developed by the Environmental Protection Agency (EPA), is widely used for estimating road transport emissions in official annual national inventories. [6]. A key distinction of the COPERT 4 model compared to some traditional models is its capability to estimate cold start, hot running, and non-exhaust emissions from all road vehicle categories [7]. Conventional air quality models, which employ numerical and mathematical techniques to simulate the physical and chemical processes influencing air pollutants, have been widely utilized to assist in the design of effective air pollution mitigation strategies. However, in recent years, artificial intelligence methods such as Fuzzy Logic (FL) and Artificial Neural Networks (ANN) have gained prominence due to their greater adaptability and flexibility\u003c!--[if supportFields]\u003e\u003cspan style='mso-element: field-begin'\u003e\u003c/span\u003eCITATION Han142 \\l 1055 \u003cspan style='mso-element:field-separator'\u003e\u003c/span\u003e\u003c![endif]--\u003e [8]\u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-end'\u003e\u003c/span\u003e\u003c![endif]--\u003e. Previous studies [9], have demonstrated that traditional air quality models commonly used in the literature possess significant drawbacks, including the incorporation of highly complex inputs and extended computational times. In contrast, artificial intelligence methods, particularly the Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), have been increasingly employed for air quality prediction in recent years. Notably, ANFIS has established itself as an effective predictive model not only among neuro-fuzzy systems but also across a range of other machine learning techniques [10]. ANFIS is an adaptive network that integrates neural networks with fuzzy logic principles and is capable of processing both linear and nonlinear parameters [11]. A number of researchers have modeled CO₂ emissions across various domains using ANFIS. Environmental factors such as temperature, emission levels, and air quality have been successfully predicted using ANFIS, which has demonstrated strong performance [12].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIntelligent Transportation Systems (ITS) are an integrated blend of software, hardware, traffic engineering concepts, and communication technologies designed to enhance the efficiency and safety of transportation systems [13], and are widely used for incident detection, ramp metering, traffic signal control, parking management, speed detection, travel planning, passenger information, route guidance, variable message sign display, and vehicle violation detection. In this context, fiber optic communication, radio frequency identification (RFID), dedicated short-range communications (DSRC), cellular communication networks, data systems, and location-map information applications provide drivers with real-time traffic information [14]. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe traffic monitoring system is an integral component of Intelligent Transportation Systems (ITS), and one of its core functions is vehicle classification. In vehicle classification, in-road sensors offer high classification accuracy due to their proximity to passing vehicles, allowing them to effectively capture vehicle body characteristics and dynamic responses. Examples of in-road sensors include inductive loop sensors, vibration sensors, magnetic sensors, and piezoelectric cables [15].\u003c/p\u003e\n\u003cp\u003eThere is a strong correlation between Intelligent Transportation Systems (ITS) and the control of vehicle emissions, with road transport being the largest source of transportation-related greenhouse gas emissions. Therefore, this study focuses on estimating the quantities of emissions (NO\u003csub\u003ex\u003c/sub\u003e and CO) originating from road transport and examining the potential positive impact of ITS on climate change. For this purpose, road transport emissions in the Dilovası district of Kocaeli Province were modeled using Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), and the prediction results were compared with those generated by the COPERT 4 model. Based on the evaluation, and by considering the carbon equivalents of emissions, an ITS-based strategy was proposed for pollution mitigation through the redirection of vehicles to alternative routes in cases where road transport emission limits are exceeded in the study area. The model results were evaluated under two different scenarios, specifically focusing on emission reduction from traffic sources, to assess the applicability of the proposed system and its potential contribution to climate change mitigation. The primary aim of the study is to model transportation-related emissions in the Dilovası district\u0026mdash;selected as the case area due to its high air pollution levels\u0026mdash;and to investigate the feasibility of using ITS for emission reduction in the region. In doing so, the study seeks to highlight the environmental benefits of ITS by emphasizing its impact on vehicle emission control and its contributions to transportation efficiency, economic improvement, and climate change mitigation.\u003c/p\u003e"},{"header":"Data and methods","content":"\u003ch2\u003eStudy area and data sources\u003c/h2\u003e\n\u003cp\u003eThe models were trained using four years of traffic data (2015\u0026ndash;2018) from the Dilovası district of Kocaeli Province. Dilovası (40\u0026deg; 50\u0026prime; 04\u0026Prime; N, 29\u0026deg; 35\u0026prime; 22\u0026Prime; E) is a district characterized by a high concentration of industrial facilities and severely polluted air (Figure 1). The district\u0026apos;s basin-like topographical structure causes emissions released from factory chimneys to have long-term impacts on the local population [16]. Moreover, vehicles traveling on the D-100 Highway and the E-80 TEM Motorway, which pass through the district and accommodate various classes of vehicles, constitute a significant source of greenhouse gas emissions and air pollution in the area \u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-begin'\u003e\u003c/span\u003eCITATION Züb19 \\l 1055 \u003cspan style='mso-element:field-separator'\u003e\u003c/span\u003e\u003c![endif]--\u003e[17]\u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-end'\u003e\u003c/span\u003e\u003c![endif]--\u003e.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eData collection and classification\u003c/h2\u003e\n\u003cp\u003eFor model development, both training and testing datasets were constructed using the number of vehicles passing through the district by type, along with environmental variables, including monthly average daily minimum temperature (℃), maximum temperature (℃), and average relative humidity (%), collected from the Kocaeli Cengiz Topel Naval Air Station Meteorological Station (MeteoBlue, 2023). To determine the monthly average daily data for all vehicle types passing through the district, vehicles were initially categorized into two main groups: light and heavy vehicles. For a more detailed analysis, both light and heavy vehicles were further classified into specific vehicle types based on distribution ratios obtained from the Turkish Statistical Institute (T\u0026Uuml;İK). Passenger cars were categorized into three subgroups according to fuel type: gasoline, diesel, and LPG. Medium-duty commercial vehicles, including passenger cars, pickup trucks, and minibuses, were classified as \u0026quot;light commercial vehicles,\u0026quot; while buses, trucks, and articulated lorries were considered \u0026quot;heavy vehicles.\u0026quot; It was assumed that all heavy vehicles operate on diesel fuel. The statistical data necessary to estimate vehicle numbers by type and composition were obtained from the General Directorate of Highways (KGM). \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAfter acquiring environmental data, more detailed vehicle data were generated using ratios obtained from relevant institutions. The COPERT 4 emission estimation program was utilized to predict transportation-related emissions, while ANFIS and ANN tools modeled greenhouse gas emissions by employing the necessary environmental and vehicle data to produce prediction results. The workflow for model development is illustrated in Figure 2. Raw data were initially subjected to preprocessing. Except for meteorological variables, data related to vehicle classification inputs underwent processing before modeling.\u003c/p\u003e\n\u003ch2\u003eModel definition of the transportation-related air pollution modeling system\u003c/h2\u003e\n\u003cp\u003eThis section introduces the transportation-related air pollution modeling system and details the specific components of each sub-model. The framework of the modeling system is illustrated in Figure 2. The COPERT 4 model was employed to estimate emissions generated by different vehicle types. The ANFIS and ANN models utilize vehicle and environmental data obtained from COPERT 4 to apply artificial intelligence techniques, producing vehicle emission outputs. Consequently, air pollution emission data are processed by the emission models. This modeling system is used for projecting transportation-related greenhouse gas emissions.\u003c/p\u003e\n\u003ch2\u003eCOPERT 4 model\u003c/h2\u003e\n\u003cp\u003eCOPERT 4 has been developed to estimate cold start emissions by incorporating various parameters such as ambient temperature, vehicle usage patterns, and the cold-to-hot start ratio. Additionally, COPERT 4 has established methodologies to adjust the base hot emission factors based on vehicle mileage and fuel characteristics [6].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor improved accuracy and reliability, classified data were captured and used to simulate future transportation-related NO\u003csub\u003ex\u003c/sub\u003e and CO pollutant parameters in Dilovası. In the study, vehicle counts and environmental information passing through the district were processed appropriately for the model. The COPERT 4 framework utilized in this research is illustrated on the right side of Figure 2. Considering the characteristics of the model outputs, the necessary vehicle and environmental data were disaggregated by month over multiple years and processed for emission estimation.\u003c/p\u003e\n\u003cp\u003eAccording to the study by (Tulyasuwan \u0026amp; Federici, 2011), the use of Tier 2 in the COPERT 4 model is recommended. Furthermore, the demonstrated effectiveness of time series emission results (Tongwane, et al., 2015) has influenced their implementation in the present study.\u003c/p\u003e\n\u003ch2\u003eArtificial intelligence models\u003c/h2\u003e\n\u003cp\u003eThe input data used in the study were compared against prediction results from different models. Accordingly, various artificial intelligence modeling tools were employed to process vehicle and environmental data and generate emission estimates for the transportation-related air quality model. The first of these is the Adaptive Neuro-Fuzzy Inference System (ANFIS) model. Literature reviews indicate that when emission values are modeled using ANFIS and Artificial Neural Network (ANN), ANFIS predictions tend to be more reliable [18]. The second model employed is ANN, which is capable of analyzing large datasets and modeling complex processes, thereby enabling more accurate emission predictions. This capability makes ANN a critical tool for the development of effective emission reduction strategies [19]. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe ANFIS model was implemented using the \u0026quot;ANFIS\u0026quot; toolbox available in MATLAB 2021b. For both pollutant parameters, 80% of the dataset was allocated for training, while the remaining 20% was used for testing. Increasing the number of subclusters resulted in higher test errors; therefore, the network was configured with two subclusters. The optimal number of training epochs was set to 20, with an error tolerance of zero. For the ANN model, a network architecture with 10 inputs and 1 output, similar to the ANFIS model, was designed, utilizing a feedforward backpropagation network. Various combinations of training and learning functions were tested, with the TRAINLM training function yielding the best performance. To prevent overfitting and unnecessary complexity, the number of hidden neurons was not increased during model development.\u003c/p\u003e\n\u003ch1\u003e\u003cstrong\u003eIntelligent transportation system implementation\u003c/strong\u003e\u003c/h1\u003e\n\u003ch2\u003eIntroduction of the routes to which vehicles will be redirected\u003c/h2\u003e\n\u003cp\u003eWithin the scope of the study, two alternative routes expected to contribute to the reduction of vehicle-related emissions in the Dilovası district were introduced: the Designed Alternative Route (DAR) and the Northern Marmara Motorway (NMM) (Figure 3). The alternative route options were designed using the Microstation Inroads software, by the project design criteria established by [20]. Among three proposed alternatives, the optimal route was determined to be 9.32 kilometers in length, based on the evaluation of various parameters, including total route length, viaduct length, tunnel length, maximum viaduct pier height, pier edge length, span between viaduct piers, as well as cut and fill volumes.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe other route, the 170.2 km long Northern Marmara Motorway, Kurtk\u0026ouml;y-Akyazı Section (including connection roads) (KMO), provides uninterrupted connectivity from Kurtk\u0026ouml;y to Akyazı. In this study, a 41.35 km segment of the motorway, located near the district boundaries and marked in Figure 4, was taken into consideration. Based on data from the [20], the construction cost of the alternative route was estimated at $5,172,229, while the total construction cost of the KMO was reported to be $2,200,000,000.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eAssessment of environmental factors and selection of the most suitable intelligent transportation system\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eIn the district, there is a need for a system that provides drivers with information in the event of air pollution limit exceedances to facilitate redirection to alternative routes. It is planned that this will be achieved through Variable Message Signs (VMS). The VMS will be capable of dynamically reflecting the prevailing air pollution conditions during vehicle passage. This system is intended to notify drivers of incidents during air pollution limit exceedance events and to provide guidance on alternative routing options. One of the key criteria in selecting the sensor type for vehicle redirection via VMS in the Dilovası district has been the meteorological conditions. Meteorological assessments for Kocaeli Province indicate that precipitation occurs in nearly half of the months throughout the year \u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-begin'\u003e\u003c/span\u003eCITATION Met \\l 1055 \u003cspan style='mso-element:field-separator'\u003e\u003c/span\u003e\u003c![endif]--\u003e[21]\u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-end'\u003e\u003c/span\u003e\u003c![endif]--\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe district\u0026rsquo;s unfavorable meteorological conditions may limit the standalone deployment of sensors such as cameras, LIDAR, radio, and Wi-Fi, which are highly sensitive to weather variations. Additionally, inductive loop systems were deemed unsuitable for exclusive use due to their inadequate performance on high-traffic volume roads.\u003c/p\u003e\n\u003cp\u003eIn the studied district, the objective is to redirect vehicles to alternative routes when air pollution limits are exceeded. Drivers who do not comply with the diversion warnings will be detected through the Electronic Enforcement System (EES) and subjected to penalties. For this purpose, a camera-based system will be required. Given the significant impact of meteorological conditions on this system, it is suggested that its use in combination with another system, rather than standalone, will yield more effective results. The integration of the selected sensor system with both VMS and EES has been identified as the most suitable approach for reducing air pollution.\u003c/p\u003e\n\u003cp\u003eA cost comparison among various sensor technologies can only be made when the specific application is clearly defined [22]. In selecting an appropriate sensor for the intended application, factors such as ease of installation and maintenance, as well as design requirements, must be taken into consideration. Sensor selection for a particular application depends on several criteria, including data parameters, data accuracy, detection range, suitable data transmission medium, location, specific installation requirements, initial cost, and the acceptability of the maintenance burden imposed by the sensor \u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-begin'\u003e\u003c/span\u003eCITATION ANe \\l 1055 \u003cspan style='mso-element:field-separator'\u003e\u003c/span\u003e\u003c![endif]--\u003e[23]\u003c!--[if supportFields]\u003e\u003cspan style='mso-element:field-end'\u003e\u003c/span\u003e\u003c![endif]--\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Cost and Emission Data for Pollutant Parameters\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable style=\"width: 4.3e+2pt;border: none;\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eCost Information\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eCost ($)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eSensor Cost\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e13.430\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eConstruction Cost of the Designed Route\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.172.229\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eTotal Cost\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.185.659\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eEmission Information\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eQuantity (ton)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAnnual Emission Quantity for the NO\u003csub\u003ex\u003c/sub\u003e Pollutant Parameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDaily Average Emission Quantity for the NO\u003csub\u003ex\u0026nbsp;\u003c/sub\u003ePollutant Parameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0,1315\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAnnual Emission Quantity for the CO Pollutant Parameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e173\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDaily Average Emission Quantity for the CO\u003csub\u003e\u0026nbsp;\u003c/sub\u003ePollutant Parameter\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0,4735\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eIn the long term, excluding the maintenance, repair, and operational costs of Intelligent Transportation Systems (ITS), an evaluation was conducted based on the approximate initial total costs of inductive loop sensors, magnetic sensors, microwave sensors, vibration sensors, LIDAR, video image processors, and Weigh-In-Motion (WIM) sensors [24], along with the daily average threshold emission quantity calculated per sensor cost (Table 1). Considering these factors, the most suitable sensor system determined to contribute to the reduction of transportation-related greenhouse gas emissions in the district was the combined system of Vibration Sensor + Variable Message Signs (VMS) + Electronic Enforcement System (EES).\u003c/p\u003e\n\u003ch2\u003eAnalysis of the depreciation period for the proposed ıntelligent transportation system\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eIn calculating the number of years required for emission cost savings to offset the total cost of the proposed alternative route and Intelligent Transportation System (ITS), the maximum values of daily average emission results for NO\u003csub\u003ex\u003c/sub\u003e and CO parameters obtained from the COPERT 4 program were used. These values were evaluated in terms of CO₂ equivalent (CO₂eq). In this context, the COPERT 4 output yielded a total of 960 tons of CO₂eq, while the daily average threshold emission quantity calculated based on sensor costs was 235 tons of CO₂eq. According to the evaluations, if ITS is implemented, the payback period in which CO₂eq-related cost savings offset the total investment cost is found to be less than one year. This indicates that the system would be highly efficient in terms of cost-performance.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe COPERT 4 model output is significantly influenced by vehicle counts, vehicle types, and environmental data inputs. Therefore, in this section, we evaluate the performance of the COPERT 4 and artificial intelligence models by examining their capabilities using vehicle and environmental data, providing an assessment of model applicability.\u003c/p\u003e\n\u003cp\u003eA previous study\u0026nbsp;[25] observed that NO\u003csub\u003ex\u003c/sub\u003e concentrations begin to increase in September, following the end of the summer months. The COPERT 4 model results indicated the highest daily average values in September 2016, with NO\u003csub\u003ex\u003c/sub\u003e emissions reaching 0.84 tons and CO emissions 4.77 tons. These findings effectively capture the intense traffic-related air pollution occurring in the district at the end of summer.\u003c/p\u003e\n\u003cp\u003eBased on the performance of the COPERT 4 model, vehicle and environmental data were utilized as primary input variables for artificial intelligence models. The ANFIS and ANN models were then applied to predict future transportation-related NO\u003csub\u003ex\u003c/sub\u003e and CO pollutant concentrations in the Dilovası district. Input and output data from both models were processed through the COPERT 4 program to ensure model validation.\u003c/p\u003e\n\u003cp\u003eThe effectiveness of the model developed and trained within the ANFIS workspace can be evaluated by comparing the model\u0026rsquo;s predictions on historical data with actual observed values. The relationship between the outputs obtained from the model for both pollutant parameters and the data calculated via the COPERT 4 program is presented in Table 1. The determination coefficients approaching a value of 1 for the training data of both pollutant parameters indicate a strong correlation between the two models.\u003c/p\u003e\n\u003cp\u003eThe relationship between the outputs obtained from the ANN model for each pollutant parameter and the data calculated using the COPERT 4 program was examined, and a regression analysis was conducted. First, the training data were simulated, and subsequently, test data were used to obtain predicted emission values for ten input sets previously unseen by the network. Similar evaluations were carried out for the CO pollutant parameter. When comparing the performance of both models, the ANFIS model was observed to yield more accurate results than the ANN model in predicting vehicle-related emissions (Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2.\u003c/strong\u003e Comparison of R\u0026sup2; Values for Model Results\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"357\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 66px;\"\u003e\n \u003cp\u003eMODEL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eData\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003eNO\u003csub\u003eX \u0026nbsp;\u003c/sub\u003e(R\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003eCO \u003csub\u003e\u0026nbsp;\u003c/sub\u003e(R\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003eANFIS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eTraining\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0,99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0,99\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eTesting\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0,89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0,94\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003eANN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eTraining\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0,64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0,85\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 65px;\"\u003e\n \u003cp\u003eTesting\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 105px;\"\u003e\n \u003cp\u003e0,36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 120px;\"\u003e\n \u003cp\u003e0,72\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eAccording to the model results, a high R\u0026sup2; value indicates a good fit of the regression model, which can be attributed to the adaptive nature of the ANFIS model.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eInvestigation of the contribution of vehicle emissions to climate change and greenhouse gases\u003c/p\u003e\u003cp\u003eGreenhouse gas impact according to different vehicle routing scenarios\u003c/p\u003e\u003cp\u003eIn this section, the impact on greenhouse gas emissions of scenarios in which light vehicles are separately routed as passenger cars and light commercial vehicles is examined. To this end, the number of vehicles calculated using threshold emission values expressed in carbon dioxide equivalent for CO and NO\u003csub\u003ex\u003c/sub\u003e pollutant parameters was selected as the emission source for the ANFIS model to evaluate its performance and investigate future greenhouse gas emissions in the Dilovası district. The study specifically focuses on the effect of changes in vehicle types on air pollution under transportation-related greenhouse gas impacts. Air pollution control strategies represent the initial step of this research and do not include other emission sources, such as those from industrial activities, that contribute to climate change.\u003c/p\u003e\u003cp\u003eCalculation of the ratio of vehicles to be routed based on threshold emission values of CO and NO\u003csub\u003ex\u003c/sub\u003e pollutant parameters expressed in CO₂ equivalent\u003c/p\u003e\u003cp\u003eThis section explains the determination of the number of vehicles to be routed based on CO₂ equivalent according to the model results. Specifically, it investigates at which vehicle count the ITS would be activated when threshold emission values expressed in CO₂ equivalents are exceeded according to the model outputs. The annual average daily NO\u003csub\u003ex\u003c/sub\u003e emission obtained from the ANFIS model is 0.504 tons, and the CO emission is 0.690 tons. For the selected ITS setup, consisting of the Vibration Sensor\u0026thinsp;+\u0026thinsp;EES\u0026thinsp;+\u0026thinsp;VMS combination, threshold emission values were determined as 0.131 tons for NO\u003csub\u003ex\u003c/sub\u003e and 0.473 tons for CO. To establish the vehicle routing ratio, the annual average daily emission values obtained from the ANFIS model for each pollutant parameter were divided by the daily average threshold emission amounts calculated based on sensor costs.\u003c/p\u003e\u003cp\u003eIn the initial step of the assessment, the carbon dioxide equivalent per unit of the NO\u003csub\u003ex\u003c/sub\u003e pollutant was taken as 900, and for the CO pollutant as 250 [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Accordingly, the model-derived NO\u003csub\u003ex\u003c/sub\u003e amount corresponded to 453.6 tons CO₂ equivalent, and the CO amount to 172.5 tons CO₂ equivalent, resulting in a total emission carbon dioxide equivalent of 626.1 tons based on the ANFIS model output. In the subsequent step, the daily average threshold emission values calculated based on sensor costs were converted into carbon dioxide equivalents for each pollutant parameter to determine threshold emission values. The threshold emission values obtained were determined as 118.39 tons CO₂ equivalent for NO\u003csub\u003ex\u003c/sub\u003e and 118.39 tons CO₂ equivalent for CO, yielding a total carbon dioxide equivalent of 236.78 tons. When the ratio of the total emission carbon dioxide equivalent obtained from the ANFIS model to the total threshold emission carbon dioxide equivalent calculated from sensor costs is considered, it is observed that the threshold emission values expressed in carbon dioxide equivalent for pollutant parameters correspond to approximately 38% of the model result. Thus, when the annual average daily vehicle count in the district reaches approximately 38% of this value, directing vehicles to alternative routes is deemed appropriate for air pollution mitigation purposes.\u003c/p\u003e\u003cp\u003eInvestigation of greenhouse gas emissions under the scenario of light vehicle redirectering\u003c/p\u003e\u003cp\u003eThis section evaluates the impact on greenhouse gas emissions of scenarios in which light vehicles are separately rerouted to alternative routes, categorized as passenger cars and medium commercial vehicles.\u003c/p\u003e\u003cp\u003eThe impact of greenhouse gas emissions under the DAR scenario\u003c/p\u003e\u003cp\u003eThe DAR-PC (Designed Alternative Route - Passenger Car) scenario refers to the redirection of passenger cars, while the DAR-MCV (Designed Alternative Route - Light Commercial Vehicle) scenario denotes the redirection of light commercial vehicles to the designed alternative route. According to the results of these scenarios, the total fuel consumption cost amounted to \u003cspan\u003e$\u003c/span\u003e7,564,509 for passenger cars and \u003cspan\u003e$\u003c/span\u003e6,424,000 for light commercial vehicles. This indicates that the fuel consumption cost in the DAR-PC scenario is approximately 1.15 times higher than that of the DAR-MCV scenario. However, considering the route lengths of 9.327 km for DAR and 41.35 km for NMM, the emission reduction in the DAR-PC scenario was approximately three times greater than that observed in the DAR-MCV scenario (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIt was concluded that, within the DAR scenario, passenger cars exhibit a greater emission reduction capacity compared to light commercial vehicles, relative to their respective fuel consumption costs.\u003c/p\u003e\u003cp\u003eImpact of greenhouse gas emissions under the NMM scenario\u003c/p\u003e\u003cp\u003eThis section refers to the NMM-PC (Northern Marmara Motorway\u0026ndash;Personal Car) scenario, which directs passenger cars, and the NMM-MCV (Northern Marmara Motorway\u0026ndash;Medium Commercial Vehicle) scenario, which directs medium commercial vehicles to the Northern Marmara Motorway. According to the scenario results, the total fuel consumption cost\u0026mdash;calculated based on the number of vehicles, fuel price, fuel consumption per vehicle, and travel distance\u0026mdash;was estimated at \u003cspan\u003e$\u003c/span\u003e33,536,231 for NMM-PC and \u003cspan\u003e$\u003c/span\u003e28,479,941 for NMM-MCV. When examining the emission reduction capacity in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, similar to the DAR scenario, the NMM-PC scenario yielded approximately three times the reduction compared to the NMM-MCV scenario. This outcome confirms that passenger cars have a greater potential for emission reduction.\u003c/p\u003e\u003cp\u003eInvestigation of the seasonal variation in transportation-related emission reduction\u003c/p\u003e\u003cp\u003eThe detailed seasonal variation in the capacity for reducing transportation-related greenhouse gas emissions in the district, based on the evaluation of all scenarios, is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. Focusing on the analyzed CO\u003csub\u003e2\u003c/sub\u003e values, it is observed that the highest concentration profiles occur during the spring and summer months. Different scenarios exhibit similar seasonal patterns. The transportation-related greenhouse gas effect by season is characterized by a bell-shaped profile, peaking in the summer. The seasonal variability of CO\u003csub\u003e2\u003c/sub\u003e serves as a powerful tool for identifying emission sources in urban areas. Between spring and summer, traffic flow increases by 1.3 to 2 times, with all major citywide roads experiencing elevated traffic, making them primary emission sources [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Considering this, a marked seasonal variation in CO\u003csub\u003e2\u003c/sub\u003e levels is evident. This finding confirms that the long-distance transport of CO\u003csub\u003e2\u003c/sub\u003e primarily occurs during the summer months, as observed in the NMM-PC and NMM-MCV scenarios.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eIn several studies [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e], CO\u003csub\u003e2\u003c/sub\u003e emissions from light vehicles have been identified as the most significant source among the analyzed emissions. In this study, the capacity for reducing CO\u003csub\u003e2\u003c/sub\u003e emissions from passenger cars over long distances was found to be higher compared to that of medium commercial vehicles (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). It is important to emphasize that this reflects the proportionality of emission production to the distance traveled.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e illustrates the impact of seasonal variations on CO\u003csub\u003e2\u003c/sub\u003e concentrations across different scenarios. The figure was generated using data collected during the same period for each scenario and subsequently aggregated. As shown, the seasonal emission reduction capacities for vehicles correspond to the DAR-PC, DAR-MCV, NMM-PC, and NMM-MCV scenarios, represented by the blue, orange, yellow, and gray box plots, respectively. For each scenario, the data spread during the winter months was the greatest, while the spring months exhibited the smallest spread. This indicates that the most heterogeneous data sets occur in winter, whereas the most homogeneous are observed in spring. Regardless of the designated routing, a decrease in CO\u003csub\u003e2\u003c/sub\u003e reduction capacity was observed in the MCV scenarios compared to the PC scenarios. Moreover, all scenarios showed a decline in CO\u003csub\u003e2\u003c/sub\u003e reduction capacity from autumn to winter. In the NMM-PC scenario, the transition from summer to autumn (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-c) exhibited a less pronounced effect on CO\u003csub\u003e2\u003c/sub\u003e reduction capacity.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThis phenomenon can be attributed to the sustained frequency of private car usage following the summer season. In relation to emission reduction capacity, the highest decrease in CO\u003csub\u003e2\u003c/sub\u003e was observed under the NMM-PC scenario across all seasons. This is due to emissions resulting from long-distance travel. Previous studies [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] have demonstrated a significantly strong correlation between air temperature and various pollutants and greenhouse gas emissions. In this study, the temperature-dependent effect of transportation-related greenhouse gas emissions is explained by the CO\u003csub\u003e2\u003c/sub\u003e reduction capacity observed in the PC scenarios, which results in substantial decreases in CO\u003csub\u003e2\u003c/sub\u003e concentrations. Conversely, scenarios involving medium commercial vehicles (MCV) (Figs.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-b, \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e-d) exhibited smaller reductions compared to the PC scenarios. This trend is attributed to the higher emission factors associated with private cars.\u003c/p\u003e\u003cp\u003eThe contribution of future transportation-related emission reductions to air quality in Dilovası District\u003c/p\u003e\u003cp\u003eWhen comparing CO\u003csub\u003e2\u003c/sub\u003e levels under the DAR and NMM scenarios, directing vehicles to a more distant route has significantly contributed to the reduction of vehicle-related emissions in Dilovası district. In this context, the CO\u003csub\u003e2\u003c/sub\u003e reduction capacity under the NMM emission scenario yielded substantially better results than the DAR scenario. Despite the differences in emission reduction amounts, the fuel consumption costs were higher in the NMM scenario compared to the DAR scenario. Considering the contribution of emission reductions by vehicle type to greenhouse gases in the district, the CO\u003csub\u003e2\u003c/sub\u003e reduction capacity under the NMM-PC scenario is 38.6% higher than that of the NMM-MCV scenario (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Within the NMM scenario itself, the NMM-PC scenario demonstrated the lowest fuel consumption cost despite a high emission reduction rate. Similar to the results of the DAR-PC and DAR-MCV scenarios, these findings support the new regulations aiming for zero CO\u003csub\u003e2\u003c/sub\u003e emissions from new passenger cars by 2035 (European Comission, 2023), making it plausible that passenger car emissions are relatively higher than those from light commercial vehicles. According to the scenario outcomes, the average fuel consumption cost is higher for medium commercial vehicles, reflecting recent efforts to improve fuel efficiency in this vehicle category (Uluslararası Enerji Ajansı, 2021).\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003e\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003eThis study initially emphasizes the environmental impacts of transportation-related greenhouse gas emissions in Dilovası district of Kocaeli Province, an area characterized by extremely high pollution concentrations. Subsequently, as greenhouse gas levels exceed pollution limits, their CO\u003csub\u003e2\u003c/sub\u003e reduction capacities and seasonal distributions were evaluated. These were then compared with scenarios where vehicles were redirected to alternative routes, investigating the effect of CO\u003csub\u003e2\u003c/sub\u003e on air quality. The influence of various variables\u0026mdash;such as vehicle type, season, and travel distance\u0026mdash;was analyzed to identify the optimal configuration that achieves maximum CO\u003csub\u003e2\u003c/sub\u003e reduction alongside minimal fuel consumption costs.\u003c/p\u003e\u003cp\u003eThe presence of localized CO\u003csub\u003e2\u003c/sub\u003e concentrations exceeding pollution limits in all seasons, attributable to transportation, underscores a widespread environmental issue with the potential to trigger climate change. This highlights the critical importance of air pollution control in regions exhibiting high pollution concentrations. Given the active heavy and light vehicle traffic in the district and the proximity of highways to the district center, the study demonstrates how diverting vehicles away from the center toward alternative routes contributes to mitigating CO\u003csub\u003e2\u003c/sub\u003e pollution.\u003c/p\u003e\u003cp\u003eAdditionally, it was found that average fuel consumption costs are higher for medium commercial vehicles compared to passenger cars, reflecting recent efforts to improve fuel efficiency in this vehicle category. The analysis also revealed that seasonal factors play a significant role in the CO\u003csub\u003e2\u003c/sub\u003e reduction capacity within transportation dynamics, with CO\u003csub\u003e2\u003c/sub\u003e pollutant peaks observed during spring and summer. Such seasonal variations reflect the impact of heavily trafficked urban arterial roads serving as primary emission sources, underscoring the need for region-specific monitoring and pollution reduction strategies.\u003c/p\u003e\u003cp\u003eFurthermore, the greatest CO\u003csub\u003e2\u003c/sub\u003e reduction was achieved in scenarios directing passenger cars to the Northern Marmara Motorway across all seasons, attributable to emissions associated with longer travel distances. The study advocates for stricter environmental regulations while targeting reductions in transportation-related greenhouse gas emissions.\u003c/p\u003e\u003cp\u003eResearch supporting pollution monitoring protocols and predictive models in cities with high pollution concentrations is essential to fully comprehend the environmental and climatic effects of CO\u003csub\u003e2\u003c/sub\u003e. Such research also guides policy development aimed at reducing the global greenhouse gas burden.\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contribution\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eH.B. carried out the investigation, methodology, statistical analysis, validation, software implementation, and writing tasks. Z.\u0026Ouml;. contributed to the review and editing of the manuscript. The work has been reviewed by all authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis work was not supported by any funding.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrespondence and requests for materials should be addressed to H.B.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003e\u003cspan\u003eNusa, K. \u0026amp; Kodak, G. \u003cem\u003eComparison of Maritime and Road Transportations in Emissions Perspective: A Review Article\u003c/em\u003evol. 10pp. 48\u0026ndash;60 (International Journal of Environment and Geoinformatics, 2023).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAlbuquerque, F. D., Maraqa, M. A., Chowdhury, R., Mauga, T. \u0026amp; Alzard, M. Greenhouse gas emissions associated with road transport projects: current status, benchmarking, and assessment tools, Transportation Research Procedia, vol. 48, no. pp. 2018\u0026ndash;2030, 26\u0026ndash;30 May 2019. (2020).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eZhang, R., Fujimori, S., Dai, H. \u0026amp; Hanaoka, T. \u003cem\u003eContribution of the transport sector to climate change mitigation: Insights from a global passenger transport model coupled with a computable general equilibrium model\u003c/em\u003e (International Instute for Applied System Analysis, 2018).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eGebre, T. \u0026amp; Nigussa, F. \u003cem\u003eGreenhouse Gas Emission Reduction Measures in the Urban Road Transport Sector of Ethiopia\u003c/em\u003e (Environmental Progress \u0026amp; Sustainable Energy, 2019).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eUnited Nations. Policies on spatial distribution and urbanization have broad impacts on sustainable development, (2020).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWang, H. \u0026amp; McGlinchy, L. Review of vehicle emission modelling and the issues for New Zealand, (2009).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eNtziachristos, L. \u0026amp; Samaras Z EMEP/EEA air pollutant emission inventory guidebook 2023, (2023).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSastry, H. G. \u0026amp; Marriboyina, V. A Novel Model for Air Quality Prediction using Soft Computing Techniques, \u003cem\u003eResearchGate\u003c/em\u003e, (2014).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eLu Bai, J. W. X. M. H. L. Air Pollution Forecasts: An Overview. \u003cem\u003eInternational J. Environ. Res. Public. Health\u003c/em\u003e, (2018).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eSalleh, M. N. M., Talpur, N. \u0026amp; Hussain, K. Adaptive Neuro-Fuzzy Inference System: Overview, Strengths, Limitations, and Solutions, \u003cem\u003eResearchGate\u003c/em\u003e, (2018).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAlhindawi, R., Nahleh, Y. A., Kumar, A. \u0026amp; Shiwakoti, N. Application of a Adaptive Neuro-Fuzzy Technique for Projection of the Greenhouse Gas Emissions from Road Transportation, \u003cem\u003eSustainability\u003c/em\u003e, vol. 11, no. 6346, (2019).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKhan, M. Z. \u0026amp; Khan, M. F. Application of ANFIS, ANN and fuzzy time series models to CO2 emission from the energy sector and global temperature increase. \u003cem\u003eInternational J. Clim. Change Strategies Management\u003c/em\u003e, pp. 622\u0026ndash;642, (2019).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAlsabaan, M., Naik, K., Khalifa, T. \u0026amp; Nayak, A. Vehicular networks for reduction of fuel consumption and CO2 emission, in \u003cem\u003e8th IEEE International Conference on Industrial Informatics\u003c/em\u003e, Osaka, (2010).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKarayolları, G. \u0026amp; M\u0026uuml;d\u0026uuml;rl\u0026uuml;ğ\u0026uuml; \u003cem\u003eAkıllı Ulaşım Sistemleri\u003c/em\u003e, Karayolları Genel M\u0026uuml;d\u0026uuml;rl\u0026uuml;ğ\u0026uuml;, (2022).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eWon, M. Intelligent Traffic Monitoring Systems for Vehicle Classification: A Survey, USA, (2020).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eT\u0026uuml;rk, T. \u0026amp; Birliği T\u0026uuml;rk Tabipleri Birliği Dilovası Raporu, T\u0026uuml;rk Tabipleri Birliği Yayınları, (2012).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003e\u0026Ouml;zt\u0026uuml;rk, Z., Arslantaş, O. A., Beba, E., Yılmaz, H. \u0026amp; Toros, H. M. and Air Pollution Reduction with Intelligent Transportation Systems: Dilovası Scenario. \u003cem\u003eJournal Res. Atmospheric Science\u003c/em\u003e, (2019).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKunt, F., Ayturan, Z. C. \u0026amp; Dursun, S. Used Some Modelling Applications in Air Pollution Estimates. \u003cem\u003eJ. Int. Environ. Application Sci.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e (4), 418\u0026ndash;425 (2016).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCowls, J., Taddeo, M., Tsamados, A. \u0026amp; Floridi, L. The AI Gambit \u0026mdash; Leveraging Artificial Intelligence to Combat Climate Change: Opportunities, Challenges, and Recommendations, \u003cem\u003eResearchGate\u003c/em\u003e, (2021).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKarayolları, G. \u0026amp; M\u0026uuml;d\u0026uuml;rl\u0026uuml;ğ\u0026uuml; Karayolları Tasarım El Kitabı, Karayolları Genel M\u0026uuml;d\u0026uuml;rl\u0026uuml;ğ\u0026uuml;, (2016).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eMeteoroloji Genel \u0026amp; M\u0026uuml;d\u0026uuml;rl\u0026uuml;ğ\u0026uuml; [Online]. (2024). Available: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mgm.gov.tr/veridegerlendirme/il-ve-ilceler-istatistik.aspx?k=H\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eFHWA, A. \u003cem\u003eSummary of Vehicle Detection and Surveillance Technologies used in Intelligent Transportation Systems\u003c/em\u003e (Federal Highway Administration\u0026rsquo;s Intelligent Transportation Systems, 2007).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eUnited States Department of Transportation Federal Highway Administration,. [Online]. (2007). Available: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://highways.dot.gov/public-roads/novdec-2007/new-look-sensors\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKlein, L. A. \u0026amp; Roadside \u003cem\u003eSens. Traffic Manage.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 01, (2022).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eT\u0026uuml;bitak Dilovası End\u0026uuml;stri B\u0026ouml;lgesi ve \u0026Ccedil;evresinde Hava Kirliliğine Neden olan Organik ve İnorganik Kirleticilerin D\u0026uuml;zeylerinin ve Kaynaklarının Belirlenmesi, (2017).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eAvrupa \u0026Ccedil;evre, A. Costs of air pollution from European industrial facilities 2008\u0026ndash;2017, (2021).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eKrylov, P. M., Volodin, O. N., Zaitsev, G. A., Nekrasova, L. P. \u0026amp; Klyuchnikov, D. A. \u003cem\u003eEstimating Summer Emissions from Land Transportation\u003c/em\u003evol. 16 (Asian Journal of Water, Environment and Pollution, 2019).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eCongressional Budget Office. Emissions of Carbon Dioxide in the Transportation Sector, U.S., (2022).\u003c/span\u003e\u003c/li\u003e\n \u003cli\u003e\u003cspan\u003eHoseinifar, S. E., Ashrafi, K. \u0026amp; Pour-Motlagh, M. S. The Effects of Seasonal Changes of Ambient Temperature and Humidity on Exhaust Pipe Emissions and Greenhouse Gases, \u003cstrong\u003e2383\u003c/strong\u003e, 4501, (2023).\u003c/span\u003e\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Artificial intelligence, Climate change, Greenhouse gas emission, Intelligent transportation systems","lastPublishedDoi":"10.21203/rs.3.rs-7194240/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7194240/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBased on the modeling results, the carbon equivalent values for each pollutant were calculated. In cases where transportation-related emissions exceed regulatory thresholds, an ITS-based traffic management strategy was proposed to redirect vehicles to alternative routes.\u003c/p\u003e\n\u003cp\u003eTransportation sector is among the critical domains where effective public interventions and adaptive strategies are essential to mitigate CO₂ emissions. In this context, artificial intelligence (AI)-based models are increasingly utilized to support emission reduction efforts.\u003c/p\u003e\n\u003cp\u003eThis study focuses on developing a predictive model for road transport-related greenhouse gas emissions in Dilovası, a district of Kocaeli Province known for its high levels of air pollution. Two AI-based approaches, Adaptive Neuro-Fuzzy Inference System (ANFIS) and Artificial Neural Networks (ANN), were employed to model pollutant emissions. At the same time, the potential contribution of Intelligent Transportation Systems (ITS) to emission mitigation and climate change adaptation was also examined.\u003c/p\u003e\n\u003cp\u003eInitially, NO\u003csub\u003ex\u003c/sub\u003e and CO emissions from light and heavy vehicles were modeled using ANFIS and ANN, and the results were compared with outputs from the COPERT 4 (Calculations of Emissions from Road Transport) software. The high adaptability of the ANFIS model allowed for a more accurate representation of the influence of environmental variables and vehicle counts on emission levels.\u003c/p\u003e","manuscriptTitle":"Investigation of Road Transport-Based Greenhouse Gas Prediction Models and the Use of Intelligent Transportation Systems for Emission Reduction","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-08 17:54:58","doi":"10.21203/rs.3.rs-7194240/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-09T09:57:00+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-08T14:32:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"321543634479787540082520634804331050648","date":"2025-10-08T13:52:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"271232659251160053950954479456664779815","date":"2025-10-06T03:54:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"206743226207340176706772496758152041062","date":"2025-10-03T04:03:56+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-11T14:20:12+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-07T14:25:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"24749646810279978569341553386102911056","date":"2025-08-06T10:08:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"334338794889409207619922323454157919676","date":"2025-08-06T04:57:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"168113910813429731668196044274783718469","date":"2025-08-05T05:20:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"173092146906818363041597578144819611554","date":"2025-08-04T16:29:29+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-04T16:23:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-04T16:20:23+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-01T09:10:10+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-28T12:16:01+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-07-28T12:12:24+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"25fc6d58-f055-41b2-8163-ae787b9988d1","owner":[],"postedDate":"August 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":52740875,"name":"Earth and environmental sciences/Climate sciences"},{"id":52740876,"name":"Earth and environmental sciences/Environmental sciences"},{"id":52740877,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2025-12-22T16:02:21+00:00","versionOfRecord":{"articleIdentity":"rs-7194240","link":"https://doi.org/10.1038/s41598-025-29724-6","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-12-18 15:58:05","publishedOnDateReadable":"December 18th, 2025"},"versionCreatedAt":"2025-08-08 17:54:58","video":"","vorDoi":"10.1038/s41598-025-29724-6","vorDoiUrl":"https://doi.org/10.1038/s41598-025-29724-6","workflowStages":[]},"version":"v1","identity":"rs-7194240","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7194240","identity":"rs-7194240","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}