Artificial neural network (ANN) modeling for the prediction of odor emission rates from landfill working surface

Municipal solid waste management for low-carbon transition: A systematic review of artificial neural network applications for trend prediction

Improper municipal solid waste (MSW) management contributes to greenhouse gas emissions, necessitating emissions reduction strategies such as waste reduction, recycling, and composting to move towards a more sustainable, low-carbon future. Machine learning models are applied for MSW-related trend prediction to provide insights on future waste generation or carbon emissions trends and assist the formulation of effective low-carbon policies. Yet, the existing machine learning models are diverse and scattered. This inconsistency poses challenges for researchers in the MSW domain who seek to identify and optimize the machine learning techniques and configurations for their applications. This systematic review focuses on MSW-related trend prediction using the most frequently applied machine learning model, artificial neural network (ANN), while addressing potential methodological improvements for reducing prediction uncertainty. Thirty-two papers published from 2013 to 2023 are included in this review, all applying ANN for MSW-related trend prediction. Observing a decrease in the size of data samples used in studies from daily to annual timescales, the summarized statistics suggest that well-performing ANN models can still be developed with approximately 33 annual data samples. This indicates promising opportunities for modeling macroscale greenhouse gas emissions in future works. Existing literature commonly used the grid search (manual) technique for hyperparameter (e.g., learning rate, number of neurons) optimization and should explore more time-efficient automated optimization techniques. Since there are no one-size-fits-all performance indicators, it is crucial to report the model's predictive performance based on more than one performance indicator and examine its uncertainty. The predictive performance of newly-developed integrated models should also be benchmarked to show performance improvement clearly and promote similar applications in future works. The review analyzed the shortcomings, best practices, and prospects of ANNs for MSW-related trend predictions, supporting the realization of practical applications of ANNs to enhance waste management practices and reduce carbon emissions.

Assessing the spatial distribution of odor at an urban waterfront using AERMOD coupled with sensor measurements

Impressions of a place are partly formed by smell. The urban waterfronts often leave a rather poor impression due to odor pollution, resulting in recurring complaints. The nature of such complaints can be subjective and vague, so there is a growing interest in quantitative measurements of emissions to explore the causes of malodorous influence. In the present work, an air quality monitor with an H2S sensor was employed to continuously measure emissions of malodors at 1-min resolution. H2S is often considered to be the predominant odorous substance from sludge and water bodies as it is readily perceptible. The integrated means of concentration from in situ measurements were combined with the AERMOD dispersion model to reveal the spatial distribution of odor concentrations and estimate the extent of odor-prone areas at a daily time step. Year-long observations showed that the diurnal profile exhibits a positively skewed distribution. Meteorology plays a vital role in odor dispersion; the degree of dispersion was explored on a case-by-case basis. There is a greater likelihood of capturing the concentration peaks at night (21:00 to 6:00) as the air is more stable then with less tendency for vertical mixing but favors a horizontal spread. This study indicates that malodors are changeable in time and space and establishes a new approach to using H2S sensor data and resolves a long-standing question about odor in Hong Kong.Implications: this study establishes a new approach combining dispersion model with novel H2S sensor data to understand the characteristics and pattern of odor emanated from the urban waterfront in Hong Kong. The sensor has dynamic concentration range to detect the episodic level of H2S and low level at background conditions. It provides more complete information in relation to odor annoyance, as well as quantitative information useful for odor regulation.

Artificial intelligence for waste management in smart cities: a review

The rising amount of waste generated worldwide is inducing issues of pollution, waste management, and recycling, calling for new strategies to improve the waste ecosystem, such as the use of artificial intelligence. Here, we review the application of artificial intelligence in waste-to-energy, smart bins, waste-sorting robots, waste generation models, waste monitoring and tracking, plastic pyrolysis, distinguishing fossil and modern materials, logistics, disposal, illegal dumping, resource recovery, smart cities, process efficiency, cost savings, and improving public health. Using artificial intelligence in waste logistics can reduce transportation distance by up to 36.8%, cost savings by up to 13.35%, and time savings by up to 28.22%. Artificial intelligence allows for identifying and sorting waste with an accuracy ranging from 72.8 to 99.95%. Artificial intelligence combined with chemical analysis improves waste pyrolysis, carbon emission estimation, and energy conversion. We also explain how efficiency can be increased and costs can be reduced by artificial intelligence in waste management systems for smart cities.

Development of a smart spectral analysis method for the determination of mulberry (Morus alba var. nigra L.) juice quality parameters using FT-IR spectroscopy

Recently, the application of Fourier transform infrared (FT-IR) spectroscopy as a noninvasive technique combined with chemometric methods has been widely noted for quality evaluation of agricultural products. Mulberry (Morus alba var. nigra L.) is a native fruit of Iran and there is limited information about its quality characteristics. The present study aims at assessing a nondestructive optical method for determining the internal quality of mulberry juice. To do so, first, FT-IR spectra were acquired in the spectral range 1000-8333 nm. Then, the principal component analysis (PCA) was used to extract the principal components (PCs) which were given as inputs to three predictive models (support vector regression (SVR), partial least square (PLS), and artificial neural network (ANN)) to predict the internal parameters of the mulberry juice. The performance of predictive models showed that SVR got better results for the prediction of ascorbic acid (R 2 = .84, RMSE = 0.29), acidity (R 2 = .71, RMSE = 0.0004), phenol (R 2 = .35, RMSE = 0.19), total anthocyanin (R 2 = .93, RMSE = 5.85), and browning (R 2 = .89, RMSE = 0.062) compared to PLS and ANN. However, the ANN predicted the parameters TSS (R 2 = .98, RMSE = 0.003) and pH (R 2 = .99, RMSE = 0.0009) better than the other two models. The results indicated that a good prediction performance was obtained using the FT-IR technique along with SVR and this method could be easily adapted to detect the quality parameters of mulberry juice.

Health risk assessment of volatile organic compounds (VOCs) emitted from landfill working surface via dispersion simulation enhanced by probability analysis

The assessment of the health risks of volatile organic compounds (VOCs) emitted from landfills via dispersion model is crucial but also challenging because of remarkable variations in their emissions and meteorological conditions. This study used a probabilistic approach for the assessment of the health risks of typical VOCs by combining artificial neural network models for emission rates and a numerical dispersion model enhanced by probability analysis. A total of 8753 rounds of simulation were performed with distributions of waste compositions and the valid hourly meteorological conditions for 1 year. The concentration distributions and ranges of the typical health-risky VOCs after dispersion were analyzed with 95% probability. The individual and cumulative non-carcinogenic risks of the typical VOCs were acceptable with all values less than 1 in the whole study domain. For individual carcinogenic risks, only ethylbenzene, benzene, chloroform, and 1, 2-dichloroethane at extreme concentrations showed minor or moderate risks with a probability of 0.1%-1% and an impact distance of 650-3000 m at specific directions. The cumulative carcinogenic risks were also acceptable at 95% probability in the whole study domain, but exceeded 1 × 10^-6 or even 1 × 10^-4 at some extreme conditions, especially within the landfill area. The vertical patterns of the health risks with height initially increased, and then decreased rapidly, and the peak values were observed around the height of the emission source. The dispersion simulation and health risk assessment of the typical health-risky VOCs enhanced by Monte Carlo can accurately reflect their probabilistic dispersion patterns and health risks to surrounding residents from both spatial and temporal dimensions. With this approach, this study can provide important scientific basis and technical support for the health risk assessment and management of landfills.

Spatial Distribution of Precise Suitability of Plantation: A Case Study of Main Coniferous Forests in Hubei Province, China

(1) Background. Conifers are the main plantation species in southern China, including Masson Pine (MP), Chinese fir (CF) and Chinese thuja (CT). Clarifying the suitable site conditions for these conifers is helpful for large-area afforestation, so as to manage forests to provide a higher level of ecosystem services. To achieve the research goals, we take the conifers in Hubei Province of southern China as a case study. (2) Methods. The situations of conifers, as well as environmental conditions of 448 sampling plots, were then investigated. The suitable growth environment of conifers in the studied area was determined by the maximum entropy algorithm, and the suitability spatial distribution of coniferous forests at the provincial level was also analyzed. (3) Results. The effect of the conifers suitability prediction model reached an accurate level, where AUC values of MP, CF and CT training set were 0.828, 0.856 and 0.970, respectively. Among multiple environmental factors, such as geography and climate, altitude is the most important factor affecting conifer growth. The contribution of altitude to the growth suitability of MP, CF and CT was 38.1%, 36.2% and 36.1%, respectively. Suitable areas of MP, CF and CT were 97,400 ha, 74,300 ha and 39,900 ha, accounting for 52.45%, 39.97% and 21.46% of the studied area, respectively. We concluded that the suitable site conditions of conifer plantations were 2800-5600 oC annual accumulated temperature, 40-1680 m a.s.l., and < 40° slopes. (4) Conclusions. The study suggests that accurate spatial suitability evaluation should be carried out to provide sufficient support for the large-area afforestation in southern China. However, due to our data and study area limitations, further studies are needed to explore the above findings for a full set of plantation species in an extensive area of southern China.

Genetic algorithm (GA) - Artificial neural network (ANN) modeling for the emission rates of toxic volatile organic compounds (VOCs) emitted from landfill working surface

Volatile organic compounds (VOCs) emitted from the working surface of landfills have received increasing attention due to the potential risks to human health. Quantifying the emission rates of risky VOCs is important to their health risk assessment but is also challenging because of their high variation and complicated relationship between the emission rates and various influencing factors. In this study, a continuous nine-month sampling of VOCs was conducted on a landfill working surface to identify dominant VOCs that are risky to human health and to construct artificial neural network (ANN) models for their emission rates by involving 105 datasets. Among the 63 detected VOCs, ethanol presented the highest emission rate (885.28 ± 1398.10 μg·m^-2·s^-1), and the dominant compounds with high emission rates and detection frequencies were characterized in each category. According to the human toxicity impact scores calculated with USEtox method, carbon tetrachloride, ethanol, tetrachloroethylene, 1, 2-dichloroethane, benzene, ethylbenzene, and chloroform were identified as the dominant carcinogenic VOCs, and acrolein, carbon tetrachloride, and 1, 2-dichloropropane were the dominant noncarcinogenic VOCs. ANN models were established for the emission rates of six typical risky VOCs, with meteorological conditions and waste compositions as input parameters and emission rates as output parameters. With the structure optimization and genetic algorithm, all the ANN models achieved good performance and excellent prediction capability with high R² and low root mean square error (RMSE) values. The emission rates under a 95% probability were predicted for each risky VOCs via the established ANN models, by randomly sampling the input parameters under their data distribution. The approach proposed and results obtained can provide scientific methodology and important information for the monitoring, prediction, and health risk assessment of the VOCs emitted from MSW landfills.

An Intelligent Site Selection Model for Hydrogen Refueling Stations Based on Fuzzy Comprehensive Evaluation and Artificial Neural Network—A Case Study of Shanghai

With the gradual popularization of hydrogen fuel cell vehicles (HFCVs), the construction and planning of hydrogen refueling stations (HRSs) are increasingly important. Taking operational HRSs in China’s coastal and major cities as examples, we consider the main factors affecting the site selection of HRSs in China from the three aspects of economy, technology and society to establish a site selection evaluation system for hydrogen refueling stations and determine the weight of each index through the analytic hierarchy process (AHP). Then, combined with fuzzy comprehensive evaluation (FCE) method and artificial neural network model (ANN), FCE method is used to evaluate HRS in operation in China’s coastal areas and major cities, and we used the resulting data obtained from the comprehensive evaluation as the training data to train the neural network. So, an intelligent site selection model for HRSs based on fuzzy comprehensive evaluation and artificial neural network model (FCE-ANN) is proposed. The planned HRSs in Shanghai are evaluated, and an optimal site selection of the HRS is obtained. The results show that the optimal HRSs site selected by the FCE-ANN model is consistent with the site selection obtained by the FCE method, and the accuracy of the FCE-ANN model is verified. The findings of this study may provide some guidelines for policy makers in planning the hydrogen refueling stations.