Air Quality Prediction Research Articles

Air-Pollution Monitoring using IoT

Abstract: The level of pollution has increased due to a lot of factors like the increase in population, commercial vehicles, industrialization and stubble burning which has caused harmful effects on human wellbeing by directly affecting health of population exposed to it. The worst affected by this are the individuals, like traffic police officers, who are required to stay outside breathing the toxic gases and bearing the strong heat to maintain a constant traffic flow and order in society. Studies and surveys show that they suffer from respiratory diseases, photo sensitivity, skin infections, nausea, high blood pressure and many other problems. This research paper presents an innovative way to monitor real-time air quality and collect data to provide insights to local authorities. The ‘Air Pollution Monitoring System’ using Internet of Things will monitor the air quality over a web server using cloud technology and will trigger a alarm when the air quality goes beyond a certain level, meaning there are dangerous amount of harmful gases present in the air like CO2, CO, particulate matter, alcohol, benzene and NH3.Governments and citizens are looking for scientificintellect to challenge the common threat of pollution in its many procedures. Currently mobile apps are capable of accomplishing functions like reporting status of air quality, air quality forecasts, air quality monitoring in a particular area, and risks highlighting. But these applications do not provide the user with data of a local area, where the user is currently located, which will be more beneficial for an individual.

ENVIRONMENTAL POLLUTION ANALYSIS AND PREDICTION OF INFLUENTIAL FACTORS: A DATA-DRIVEN INVESTIGATION

Environmental pollution is a pressing concern affecting humanity and other Earth species today. The escalation of air pollution due to heavy traffic density and industrial effluents in urban areas poses a significant threat to human life and the environment. The Air Quality Index (AQI) is employed as a crucial metric to gauge the extent of air pollution in a city. Apart from industrial and traffic emissions, meteorological factors, including wind speed, wind direction, temperature, humidity, and total precipitation, play a pivotal role in shaping a citys pollution levels. Mitigating the harm caused by air pollution is contingent on the ability to predict the factors influencing air pollutant levels. This predictive capability enables issuing advanced warnings to citizens or implementing precautionary measures for their protection. This paper focuses on forecasting temperature, humidity, wind speed, wind direction, and total precipitation in the town of Basel, Switzerland, on specific future dates, utilizing historical data maintained by local authorities. The methodology leverages Hive tools and MapReduce frameworks and employs machine learning techniques to predict future values of desired parameters. Various regression techniques, including Lasso, Linear Regression, Ridge Regression, and Polynomial Regression, were evaluated for their performance in predicting air quality factors. Polynomial Regression emerged as the preferred choice due to its superior performance. To enhance user accessibility, a user-friendly graphical interface (GUI) has been developed to input data and visualize predicted results. This study presents a comprehensive approach to address the critical issue of air pollution by providing accurate predictions of meteorological factors, thus enabling proactive measures to safeguard public health and the environment.

The Michigan–Ontario Ozone Source Experiment (MOOSE): An Overview

The Michigan–Ontario Ozone Source Experiment (MOOSE) is an international air quality field study that took place at the US–Canada Border region in the ozone seasons of 2021 and 2022. MOOSE addressed binational air quality issues stemming from lake breeze phenomena and transboundary transport, as well as local emissions in southeast Michigan and southern Ontario. State-of-the-art scientific techniques applied during MOOSE included the use of multiple advanced mobile laboratories equipped with real-time instrumentation; high-resolution meteorological and air quality models at regional, urban, and neighborhood scales; daily real-time meteorological and air quality forecasts; ground-based and airborne remote sensing; instrumented Unmanned Aerial Vehicles (UAVs); isotopic measurements of reactive nitrogen species; chemical fingerprinting; and fine-scale inverse modeling of emission sources. Major results include characterization of southeast Michigan as VOC-limited for local ozone formation; discovery of significant and unaccounted formaldehyde emissions from industrial sources; quantification of methane emissions from landfills and leaking natural gas pipelines; evaluation of solvent emission impacts on local and regional ozone; characterization of the sources of reactive nitrogen and PM2.5; and improvements to modeling practices for meteorological, receptor, and chemical transport models.

A Random Forest Regressor Model for Forecasting Air Quality Index from Particulate Matters

Air pollution is a growing concern in urban areas, including Port Harcourt City, Nigeria. Air quality index (AQI) is a numerical scale that is used for accessing and analyzing day to day air quality pattern of any place. It is projected to express the probable severity of the adverse health impacts on living organisms, most especially humans. It is however particulate matters (PM) that pose the greatest risks human health and the immediate environment. In this dissertation, a Random Forest Regressor model was used to forecast Air Quality Index. The data of the average diurnal values of PM10 and PM2.5 from Rivers State Ministry of Environment, Port Harcourt City over a long period (2017 to 2022) was evaluated. Additional meteorological and environmental variables, such as temperature, humidity, wind speed, and precipitation, were also collected as features for the prediction model. Data preprocessing techniques (MinMax Scaler) were applied to ensure data quality and consistency. The Random Forest Regressor was trained using 80% of the dataset, and the remaining 20% was used for validation. The model was evaluated using mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), and R-squared (R^2). The results obtained showed that the Random Forest Regressor performed well in forecasting Air Quality lndex with a high R^2 value of 97% and low errors (MSE: 5.2, RMSE: 7.25, MAE: 3.54). The Air Quality Index outcome shows that PM10 and PM2.5 across Port Harcourt metropolis are observed to be more noticeable during the periods of dawn than the afternoon. The Air Quality Index results reveals areas with Hazardous, moderate, good, bad and very good air quality.

Enhancing Air Quality Forecasts Across the Contiguous United States (CONUS) During Wildfires Using Analog-Based Post-Processing Methods

With our growing understanding of the risks of air pollution to human health, air quality forecasting has become a very important tool to enable decision-makers to take preventive and corrective measures for current and future policies. In addition, accurate predictions of air quality can help predict and mitigate the impacts of wildfires on human health, which have an increased risk due to anthropogenic climate change. However, errors in air quality forecasts limit their value in decision-making processes. Thus, increasing the accuracy of air quality forecasts is of significant importance. In this study, we have utilized the Community Multiscale Air Quality (CMAQ) modeling system with a 12 km horizontal grid resolution to generate daily 48-h fine particulate matter (PM2.5) forecasts for the Contiguous United States (CONUS) domain for June 1st through September 29th (major wildfire season) during 2015–2021. We conduct CMAQ offline simulations using meteorological inputs generated by the NOAA’s Unified Forecast System (UFS) numerical weather prediction model. We analyze the performance of the CMAQ PM2.5 and UFS meteorological forecasts over seven years of simulations for Environmental Protection Agency (EPA)-defined ten regions using the Air Quality System (AQS) ambient air pollution data from over a thousand monitoring sites across the CONUS. We have found that on average, the CMAQ model performs better in the eastern CONUS with the lowest Root Mean Square Error (RMSE) (10–30 μg/m3) while in the west, where wildfires are prevalent, the model has the highest RMSE of up to 120 μg/m3. To address these substantial model errors, we advance the field by applying Analog Ensemble (AnEn), a state-of-the-art statistical-dynamical method, to enhance PM2.5 forecast accuracy, with a particular focus on wildfire events, when producing an accurate forecast is highly challenging. We further refine this approach by incorporating a Carbon Monoxide-FIRE (CO-FIRE) tracer to precisely monitor fire smoke. In parallel to our objective and for the first time, we introduce two novel predictors, i.e., a CO−FIRE tracer in the model, and the maximum observed PM2.5 concentrations from the previous day (PM2.5−preday−max), and evaluate their potential significance in improving forecast accuracy, which aligns with the primary goal of our study. Despite the challenges of using AnEn for wildfires, we demonstrate that it has the potential to improve the CMAQ model forecast over the CONUS. We find that AnEn decreases the model RMSE by up to 25%, including additional 7% and 15% reduction by CO−FIRE and PM2.5−preday−max predictors, respectively, at different forecast lead times. In addition, the correlation between AnEn forecasts and observations is 20%–40% higher than that between CMAQ and observations. The Mean Bias Error (MBE) for AnEn forecasts is consistent and approximately −0.5 μg/m3 whereas CMAQ MBE varies between −1 and +1 μg/m3 between 0–48 forecast hours. AnEn significantly improves the PM2.5 forecast results during its highest episodes. During the initial days of wildfires, AnEn performs similarly to CMAQ. However, it soon catches up and decreases the error significantly.

A standardized methodology for the validation of air quality forecast applications (F-MQO): lessons learnt from its application across Europe

Abstract. A standardized methodology for the validation of short-term air quality forecast applications was developed in the framework of the Forum for Air quality Modeling (FAIRMODE) activities. The proposed approach, focusing on specific features to be checked when evaluating a forecasting application, investigates the model's capability to detect sudden changes in pollutant concentration levels, predict threshold exceedances and reproduce air quality indices. The proposed formulation relies on the definition of specific forecast modelling quality objectives and performance criteria, defining the minimum level of quality to be achieved by a forecasting application when it is used for policy purposes. The persistence model, which uses the most recent observed value as the predicted value, is used as a benchmark for the forecast evaluation. The validation protocol has been applied to several forecasting applications across Europe, using different modelling paradigms and covering a range of geographical contexts and spatial scales. The method is successful, with room for improvement, in highlighting shortcomings and strengths of forecasting applications. This provides a useful basis for using short-term air quality forecasts as a supporting tool for providing correct information to citizens and regulators.

Randomised controlled trial of a behavioural intervention to reduce exposure to PM2.5 in patients with COPD

BackgroundFine particulate matter (PM2.5) is a well-known risk factor for worse outcomes of chronic obstructive pulmonary disease (COPD). However, evidence-based guidance on effective personal behavioural strategies to minimise the effects of PM2.5 is limited. This study aimed to assess the effectiveness of a behavioural intervention in reducing PM2.5 exposure and improving clinical outcomes in patients with COPD. Materials and MethodsParticipants were 1:1 randomised, and the intervention group received a behavioural intervention consisting of five activities, while the control group received usual care. The participants were followed up for 9 months. The primary outcomes were differences in the score of St. George’s Respiratory Questionnaire for patients with COPD (SGRQ-C) and COPD assessment test (CAT) from baseline. ResultsA total of 106 participants were enrolled and 102 completed the study. At the end of the study, the intervention group showed significant improvements in the primary outcomes compared to the control group, with a group difference of −5.9 in the reduction of total SGRQ-C (−3.4 vs. 2.5; p = 0.049) and −3.8 in the CAT score (−1.2 vs. 2.7; p = 0.001). Participants with good adherence to the intervention demonstrated a greater extent of improvement in CAT score and lower PM2.5 levels compared to those who had poor adherence or were in the control group. Regular checking of air quality forecasts was significantly associated with a reduction in CAT scores among all the intervention activities. ConclusionIndividual-level behavioural interventions can be an effective strategy for mitigating the health hazards associated with PM2.5.ClinicalTrials.gov Identifier: NCT04878367.

An integrated framework for predicting air quality index using pollutant concentration and meteorological data

An integrated framework for predicting air quality index using pollutant concentration and meteorological data

An ensembled method for air quality monitoring and control using machine learning

Air quality monitoring is a significant job in our everyday life and affects the vitality of a human. It paves a better way of understanding the sources, effects and levels of pollutants in the air inhaled by the population of the region. This concept helps us in improving and developing pollution control measures to diminish the effect of air contamination. Air Quality Index (AQI) helps in understanding the level of pollution. This study briefs various state of art methods such as SVM, RF,ANN, RNN and FL in predicting air quality using machine learning. The results shows prominent outcomes which can be deployed in various cost effective hardware platforms for household and commercial purposes. The proposed model used eight parameters like NO2CO, O3, PM2.5, PM10, SO2, TEMP, PRES, DEWP, RAIN,WD, WSPM of Beijing dataset in order to predict AQI and it is tested with regional dataset. The Proposed model works based on process monitoring model. The efficient and flexible way to solve environmental issues is to combine machine learning algorithm with air quality prediction.

Building an IoT temperature and humidity forecasting model based on long short-term memory (LSTM) with improved whale optimization algorithm

In particular, predicting the temperature and humidity information plays a crucial role in plantation, estimating rainfalls and climate change, and predicting air quality via specified geographical regions. The temperature and humidity forecasting information is occasionally presented with low accuracy due to uncertain techniques and vast methods that employ different sensors and models. For this reason, this work proposes an Internet of Things (IoT) temperature and humidity forecasting model based on an improved whale optimization algorithm with long short-term memory (IWOA-LSTM) technique. To increase the convergence speed processing time and overcome the local optimization problem, the IWOA is introduced. The number of hidden layers, learning rate momentum, and weight decay of the LSTM optimized using the IWOA. The actual temperature and humidity data are collected using DHT11 and ESP8266 NodeMCU practical model and processed using the ThingSpeak platform. The processing data stage depends on filling the missing data gaps using the rolling average technique (RAT). The performance evaluation of the proposed IWOA-LSTM forecasting model is assessed using some statistical functions, namely known as mean square error, mean absolute error, root mean square error, and mean absolute percentage error. The IWOA-LSTM techniques were also assessed using throughput, latency, and power consumption. The developed IWOA-LSTM model shows high accuracy, leading to better forecasting information than other forecasting models.

Volatile organic compound fluxes in the agricultural San Joaquin Valley – spatial distribution, source attribution, and inventory comparison

Abstract. The San Joaquin Valley is an agricultural region in California that suffers from poor air quality. Since traffic emissions are decreasing, other sources of volatile organic compounds (VOCs) are gaining importance in the formation of secondary air pollutants. Using airborne eddy covariance, we conducted direct, spatially resolved flux observations of a wide range of VOCs in the San Joaquin Valley during June 2021 at 23–36 ∘C. Through land-cover-informed footprint disaggregation, we were able to attribute emissions to sources and identify tracers for distinct source types. VOC mass fluxes were dominated by alcohols, mainly from dairy farms, while oak isoprene and citrus monoterpenes were important sources of reactivity. Comparisons with two commonly used inventories showed that isoprene emissions in the croplands were overestimated, while dairy and highway VOC emissions were generally underestimated in the inventories, and important citrus and biofuel VOC point sources were missing from the inventories. This study thus presents unprecedented insights into the VOC sources in an intensive agricultural region and provides much needed information for the improvement of inventories, air quality predictions, and regulations.

A two-way coupled regional urban–street network air quality model system for Beijing, China

Abstract. Owing to the substantial traffic emissions in urban areas, especially near road areas, the concentrations of pollutants, such as ozone (O3) and its precursors, have a large difference compared to regional averages, and their distributions cannot be captured accurately by traditional single-scale air quality models. In this study, a new version of a regional urban–street network model (an Integrated Air Quality Modeling System coupling regional urban–street: IAQMS-street v2.0) is presented. An upscaling module is implemented in IAQMS-street v2.0 to calculate the impact of mass transfer to regional scale from street network. The influence of pollutants in the street network is considered in the concentration calculation on the regional scale, which is not considered in a previous version (IAQMS-street v1.0). In this study, the simulated results in Beijing during August 2021, using IAQMS-street v2.0, IAQMS-street v1.0, and the regional model (Nested Air Quality Prediction Modeling System, NAQPMS), are compared. On-road traffic emissions in Beijing, as the key model input data, were established using intelligent image-recognition technology and real-time traffic big data from navigation applications. The simulated results showed that the O3 and nitrogen oxide (NOx) concentrations in Beijing were reproduced by using IAQMS-street v2.0 on both the regional scale and street scale. The prediction fractions within a factor of 2 (FAC2s) between simulations and observations of NO and NO2 increased from 0.11 and 0.34 in NAQPMS to 0.78 and 1.00 in IAQMS-street v2.0, respectively. The normalized mean biases (NMBs) of NO and NO2 decreased from 2.67 and 1.33 to −0.25 and 0.08. In the coupled model, the concentration of NOx at the street scale is higher than that at the regional scale, and the simulated distribution of pollutants on a regional scale was improved in IAQMS-street v2.0 when compared with that in IAQMS-street v1.0. We further used IAQMS-street v2.0 to quantify the contribution of local on-road traffic emissions to the O3 and NOx emissions and analyze the effect of traffic regulation policies in Beijing. Results showed that heavy-duty trucks are the major source of on-road traffic emissions of NOx. The relative contributions of local traffic emissions to NO2, NO, and O3 concentrations were 53.41 %, 57.45 %, and 8.49 %, respectively. We found that traffic regulation policies in Beijing largely decreased the concentrations of NOx and hydrocarbons (HC); however, the O3 concentration near the road increased due to the decrease consumption of O3 by NO. To decrease the O3 concentration in urban areas, controlling the local emissions of HC and NOx from other sources requires consideration.

Air quality index prediction using multivariate deep neural networks: A case study of a proposed state capital in India

Introduction: Air pollution is a major environmental challenge worldwide and predicting air quality is key to regulating air pollution. The extent of air pollution is quantified by the Air Quality Index (AQI). Air quality forecasting has become an important area of research. Deep Neural Networks (DNN) are useful in predicting the AQI instead of traditional methods which involve numerous computations. The aim of this research paper is to investigate the use of the deep neural networks as a framework for predicting the air quality index based on time series data of pollutants.
 Materials and methods: To resolve this problem, the study proposes a DNN to develop the best model for predicting the AQI. Long Short-Term Memory (LSTM) and Bi-directional LSTM have been introduced in the study to understand and predict the relationship between the pollutants affecting the AQI. The model’s performance is evaluated using the metrics, Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Mean Square Error (MSE), Root Mean Square Error (RMSE) and coefficient of determination (R2). To conduct the study, real-time hourly data for the period November 2017 to January 2020 from an air quality monitoring station was considered for the proposed capital region of the state of Andhra Pradesh in India.
 Results: The multivariate modeling considers seven pollutants as independent variables and AQI as the target variable. After experimenting and training the algorithm on the dataset, Bi-directional LSTM was shown to have the lowest MAE and RMSE values and the highest R2, indicating that it has the highest accuracy in AQI prediction.
 Conclusion: The development of a capital city involves massive construction activity resulting in air pollution. The results are helpful to the authorities to monitor the quality of air of develop air quality management programs thus avoiding the impact of air pollution on health.

High-resolution WRF forecasts in the SmartAQ system: Evaluation of the meteorological forcing used for PMCAMx predictions in an urban area

High-resolution modeling approaches are becoming a necessary step for enhancing regional air quality monitoring and forecasting. This study first evaluates the meteorological driver of the WRF- PMCAMx 1 × 1 km2 SmartAQ air quality prediction system and then assesses the impact of meteorological errors on the PM2.5 predictions. The comprehensive multi-variable verification on several timescales appraises the suitability of such a high-resolution modeling system in predicting the surface weather parameters in a major coastal urban area while gaining valuable insights into the model performance. The highest agreement between the observed and modelled values has been found for the actinometric and the thermodynamic variables, with the exception of the soil moisture which was systematically overestimated. WRF largely captures the variability of the observed values for the dynamic variables, showing better skill in the warm season, with a systematic overestimation though. The overall ability in predicting the rain events, important for the wet deposition of atmospheric pollutants is approximately 50%. Wind speed forecast errors contribute to the PM2.5 prediction errors, regardless of season and day/night hours. The impact of the wind errors is larger at almost all stations in the cold season when local emissions are elevated.

Air quality prediction based on Long Short-Term Memory Model with advanced feature selection and hyperparameter optimization

Good air quality is one of the prerequisites for stable urban economic growth and sustainable development. Air quality is influenced by a range of environmental elements. In this study, seven common air pollutants and six kinds of meteorological data in a major city in China are studied. In this urban setting, the air quality index will be estimated based on a Long Short-term Memory (LSTM)model. To improve prediction accuracy, the Random Forest (RF) method is adopted to choose important features and pass them to the LSTM model as input, an improved sparrow search algorithm (ISSA) is used to optimize the hyperparameters of the LSTM model. According to the experimental findings, the RF-ISSA-LSTM model demonstrates superior accuracy compared to both the basic LSTM model and the ISSA-LSTM fusion model.

Interpretation of Probabilistic Surface Ozone Forecasts: A Case Study for Philadelphia

Abstract The use of probabilistic forecasting has been growing in a variety of disciplines because of its potential to emphasize the degree of uncertainty inherent in a prediction. Interpretation of probabilistic forecasts, however, is oftentimes difficult, deterring users who may benefit from such forecasts. To encourage broader use of probabilistic forecasts in the field of air quality, a process for interpreting forecasts from a statistical probabilistic air quality surface ozone model [the Regression in Self Organizing Map (REGiS)] is demonstrated. Four procedures to convert probabilistic to deterministic forecasts are explored for the Philadelphia, Pennsylvania, metropolitan area. These procedures calibrate the predicted probability of daily maximum 8-h-average ozone exceeding a standard value by 1) estimating climatological relative frequency, 2) establishing a probability of an exceedance threshold as 50%, 3) maximizing the threat score, and 4) determining the unit bias ratio. REGiS is trained using 2000–11 ozone-season (1 May–30 September) data, calibrated using 2012–14 data, and evaluated using 2015–18 data. Assessment of the calibration data with the Pierce skill score suggests an exceedance threshold based on climatological relative frequency for the conversion from probabilistic to deterministic forecasts. Calibrated REGiS generally compares well to predictions from the U.S. national air quality model and operational “expert” forecasts over the evaluation period. For other probabilistic models and situations, different procedures of converting probabilistic to deterministic forecasts may be more beneficial. The methods presented in this paper represent an approach for operational air quality forecasters seeking to use probabilistic model output to support forecasts designed to protect public health. Significance Statement Because probabilistic forecasting is becoming more prevalent in the field of air quality, the purpose of this article is to draw attention to the importance of defining a framework to accurately interpret these forecasts. This work shows that 1) probabilistic forecasts are potentially more useful to forecasters when converted into deterministic forecasts and 2) that some conversion methods are more skillful than others. It is recommended that, if it begins to produce probabilistic air quality products, the National Weather Service should implement some of the strategies presented herein to help with the interpretation of such forecasts.

Air Quality Index Prediction in Six Major Chinese Urban Agglomerations: A Comparative Study of Single Machine Learning Model, Ensemble Model, and Hybrid Model

Air pollution is a hotspot of wide concern in Chinese cities. With the worsening of air pollution, urban agglomerations face an increasingly complex environment for air quality monitoring, hindering sustainable and high-quality development in China. More effective methods for predicting air quality are urgently needed. In this study, we employed seven single models and ensemble learning algorithms and constructed a hybrid learning algorithm, the LSTM-SVR model, totaling eight machine learning algorithms, to predict the Air Quality Index in six major urban agglomerations in China. We comprehensively compared the predictive performance of the eight algorithmic models in different urban agglomerations. The results reveal that, in areas with higher levels of air pollution, the situation for model prediction is more complicated, leading to a decline in predictive accuracy. The constructed hybrid model LSTM-SVR demonstrated the best predictive performance, followed by the ensemble model RF, both of which effectively enhanced the predictive accuracy in heavily polluted areas. Overall, the predictive performance of the hybrid and ensemble models is superior to that of the single-model prediction methods. This study provides AI technological support for air quality prediction in various regions and offers a more comprehensive discussion of the performance differences between different types of algorithms, contributing to the practical application of air pollution control.

Sensitivity of Burned Area and Fire Radiative Power Predictions to Containment Efforts, Fuel Density, and Fuel Moisture Using WRF‐Fire

AbstractPredicting the evolution of burned area, smoke emissions, and energy release from wildfires is crucial to air quality forecasting and emergency response planning yet has long posed a significant scientific challenge. Here we compare predictions of burned area and fire radiative power from the coupled weather/fire‐spread model WRF‐Fire (Weather and Research Forecasting Tool with fire code), against simpler methods typically used in air quality forecasts. We choose the 2019 Williams Flats Fire as our test case due to a wealth of observations and ignite the fire on different days and under different configurations. Using a novel re‐gridding scheme, we compare WRF‐Fire's heat output to geostationary satellite data at 1‐hr temporal resolution. We also evaluate WRF‐Fire's time‐resolved burned area against high‐resolution imaging from the National Infrared Operations aircraft data. Results indicate that for this study, accounting for containment efforts in WRF‐Fire simulations makes the biggest difference in achieving accurate results for daily burned area predictions. When incorporating novel containment line inputs, fuel density increases, and fuel moisture observations into the model, the error in average daily burned area is 30% lower than persistence forecasting over a 5‐day forecast. Prescribed diurnal cycles and those resolved by WRF‐Fire simulations show a phase offset of at least an hour ahead of observations, likely indicating the need for dynamic fuel moisture schemes. This work shows that with proper configuration and input data, coupled weather/fire‐spread modeling has the potential to improve smoke emission forecasts.

Application of Machine learning techniques in environmental governance: A review

In recent years, based on the continuous improvement of computer hardware performance, machine learning (ML) technology has made rapid development. Due to the strong ability of ML methods to find complex functions between associated data and the low cost of human and material resources, they are increasingly favored by environmental governance practitioners. Current research has confirmed that the application of ML technology to the field of environmental governance is of great significance in overcoming the difficulties of analysis and practice in traditional environmental work, and can greatly promote the long-term development of such directions as air quality prediction, solid waste classification, pollutant distribution mapping, and intelligent management of water environment. Therefore, by organizing the representative literature in the past five years, this paper describes the current status of the application of ML technology in four different elements of environmental governance, namely, atmosphere, water, soil, and solid waste, and aims to summarize the possible directions of the future development of ML technology and the new development brought to the field of environmental governance by exploring the basic concepts of ML, the advantages and disadvantages of the model and the constraints and other issues.

Optimizing Long Short-Term Memory Network for Air Pollution Prediction Using a Novel Binary Chimp Optimization Algorithm

Elevated levels of fine particulate matter (PM2.5) in the atmosphere present substantial risks to human health and welfare. The accurate assessment of PM2.5 concentrations plays a pivotal role in facilitating prompt responses by pertinent regulatory bodies to mitigate air pollution. Additionally, it furnishes indispensable information for epidemiological studies concentrating on PM2.5 exposure. In recent years, predictive models based on deep learning (DL) have offered promise in improving the accuracy and efficiency of air quality forecasts when compared to other approaches. Long short-term memory (LSTM) networks have proven to be effective in time series forecasting tasks, including air pollution prediction. However, optimizing LSTM models for enhanced accuracy and efficiency remains an ongoing research area. In this paper, we propose a novel approach that integrates the novel binary chimp optimization algorithm (BChOA) with LSTM networks to optimize air pollution prediction models. The proposed BChOA, inspired by the social behavior of chimpanzees, provides a powerful optimization technique to fine-tune the LSTM architecture and optimize its parameters. The evaluation of the results is performed using cross-validation methods such as the coefficient of determination (R2), accuracy, the root mean square error (RMSE), and receiver operating characteristic (ROC) curve. Additionally, the performance of the BChOA-LSTM model is compared against eight DL architectures. Experimental evaluations using real-world air pollution data demonstrate the superior performance of the proposed BChOA-based LSTM model compared to traditional LSTM models and other optimization algorithms. The BChOA-LSTM model achieved the highest accuracy of 96.41% on the validation datasets, making it the most successful approach. The results show that the BChOA-LSTM architecture performs better than the other architectures in terms of the R2 convergence curve, RMSE, and accuracy.