Air Dew Point Research Articles

Comparative analysis of artificial intelligence models for real-time and future forecasting of environmental conditions: A wood-frame historic building case study

Precise environmental monitoring within historic buildings is crucial for ensuring optimal conditions that guarantee the proper preservation of these structures, thereby maintaining their integrity and cultural legacy. Moreover, outdoor environmental conditions can potentially impact the indoor environment, especially in historic structures with deficient insulation materials and damaged envelopes. As climate change points to more recurrent and severe extreme climatic events in the coming years, a current and future comprehensive evaluation of indoor microclimate in historic structures is essential. This study investigated the utilization of machine learning and deep learning algorithms for real-time and future forecasting of the indoor microclimate, including air temperature, relative humidity, and dew point, in a historic building located in San Antonio, Texas, USA. In situ monitored data from April 2022 to January 2023 were used to train different predictive algorithms. The results indicated that Multi-Layer Perceptrons and Support Vector Machine models yielded the most accurate values in terms of real-time forecasting of the indoor microclimate, while Extreme Gradient Boosting model excelled in convergence time. Additionally, Long Short-Term Memory models were the most accurate in predicting indoor microclimate using future weather data for the current century, specifically for 2050 and 2080. The methodology developed in this study can be applied to different construction types and locations globally. As it enables the prediction of environmental conditions crucial for historic preservation, the results have the potential to assist experts in making informed decisions about conserving historic structures, both now and in the future.

The Impact of Air Pollution Information on Individuals' Exercise Behavior: Empirical Study Using Wearable and Mobile Devices Data.

Physical exercise and exposure to air pollution have counteracting effects on individuals' health outcomes. Knowledge on individuals' real-time exercise behavior response to different pollution information sources remains inadequate. This study aims to examine the extent to which individuals avoid polluted air during exercise activities in response to different air pollution information sources. We used data on individuals' exercise behaviors captured by wearable and mobile devices in 83 Chinese cities over a 2-year time span. In our data set, 35.99% (5896/16,379) of individuals were female and 64% (10,483/16,379) were male, and their ages predominantly ranged from 18 to 50 years. We further augmented the exercise behavior data with air pollution information that included city-hourly level measures of the Air Quality Index and particulate matter 2.5 concentration (in µg/m3), and weather data that include city-hourly level measures of air temperature (ºC), dew point (ºC), wind speed (m/s), and wind direction (degrees). We used a linear panel fixed effect model to estimate individuals' exercise-aversion behaviors (ie, running exercise distance at individual-hour, city-hour, or city-day levels) and conducted robustness checks using the endogenous treatment effect model and regression discontinuity method. We examined if alternative air pollution information sources could moderate (ie, substitute or complement) the role of mainstream air pollution indicators. Our results show that individuals exhibit a reduction of running exercise behaviors by about 0.50 km (or 7.5%; P<.001) during instances of moderate to severe air pollution, and there is no evidence of reduced distances in instances of light air pollution. Furthermore, individuals' exercise-aversion behaviors in response to mainstream air pollution information are heightened by different alternative information sources, such as social connections and social media user-generated content about air pollution. Our results highlight the complementary role of different alternative information sources of air pollution in inducing individuals' aversion behaviors and the importance of using different information channels to increase public awareness beyond official air pollution alerts.

Forecasting Nighttime Frost-induced Black Ice Formation on Bridges Using Atmospheric Data

Background Faced with the tragedies caused by black ice in winter, especially on bridges, it is imperative to forecast black ice for preventive maintenance and to notify drivers approaching the dangerous spots. Methods In this study, three different machine learning algorithms—Deep Neural Network (DNN), Random Forest (RF), and Support Vector Machine (SVM)—were employed to predict nighttime black ice induced by frost on three bridges in Korea. Input data consisted of atmospheric data (air temperature, relative humidity, dew point, and differences between air temperature and relative humidity over two consecutive days) provided by the weather agency. Results To assess the employed models, reference data were generated based on the physical principle that ice forms when the pavement temperature is lower than the dew point temperature and negative. The pavement temperature was obtained using an infrared surface temperature sensor mounted on a maintenance patrol vehicle. Consequently, DNN and RF showed higher performance with an accuracy of 95%, followed by SVM with an accuracy of 92.5%. Conclusion Due to the use of easily obtainable atmospheric data, the findings of this study can be practically applied to preventive maintenance and driver information, thereby enhancing traffic safety.

High potency of application on an open direct-contact thermal storage using humid air

To reuse low-temperature wasted heat as a thermal resource for high temperature, a direct-contact adsorption thermal storage was focused using humid air and zeolite 13X particles as the working fluid and adsorbent, respectively. Only a few previous studies have chosen the working fluid in gaseous form because it is unavailable for latent heat in generating heat sources. Recovering waste heat in humid air to generate hotter steam is unique and becomes an originality of our present work. The time required to regenerate zeolite particles and the maximum temperature of the generated steam were investigated assuming a warm-up device for a vehicle. The time required to regenerate zeolite was investigated by changing the dew point, temperature, and superficial velocity of the inlet humid air. It was mainly affected by the temperature of the inlet air. The absorbent was regenerated within 30 min when the humid air preheated to 200 °C was supplied. On the other hand, the maximum steam temperature was investigated by changing the superficial velocity and temperature of saturated inlet humid air. As one of the significant and novel finding in this work, the steam of >200 °C was obtained as a high-temperature heat source even with saturated humid air unavailable latent heat. Moreover, as theoretical knowledge, it was revealed that the maximum temperature of the heat source can be estimated by the relationship between the heat balance on the packed bed and adsorption equilibrium.

A universal triangle method for evapotranspiration estimation with MODIS products and routine meteorological observations: Algorithm development and global validation

The surface temperature (Ts) and vegetation index (VI) triangle method is one of the most famous and widely applied methods for regional evapotranspiration (ET) estimation. However, no robust validation has been performed yet to evaluate its accuracy on a global scale, making its accuracy unclear when compared with other global ET models. The main reason behind it is that most traditional triangle models depend on the Ts-VI feature space within a certain spatial domain to parameterize their boundary conditions. The domain dependency makes it impossible to derive universal boundaries for continental or global application. Under this background, this study presented a universal triangle method for ET estimation with Moderate Resolution Imaging Spectroradiometer (MODIS) products and routine meteorological observations. The solution for boundary conditions was performed pixel by pixel without the consideration of spatial domain based configuration. Specifically, the dry boundary was solved theoretically based on the surface energy balance equation, and the wet boundary was substituted by wet-bulb temperature estimated from dew point temperature and air temperature. The validation against ground observations collected from 33 flux towers worldwide shows that this universal triangle method achieved evaporative fraction (EF) estimation with the correlation coefficient of 0.65, the mean absolute error of 0.17, the root-mean-square-error of 0.20, and the bias of 0.13. These four metrics of daily ET estimation were 0.83, 0.71 mm/d, 0.90 mm/d, and 0.41 mm/d, respectively. This analysis is the first comprehensive validation of the triangle method on a global scale. The comparison with early work suggests that our universal triangle method on a global scale has reached a comparable accuracy to traditional triangle models at regional scale. However, the universal triangle method has the capacity to monitor ET continuously pixel by pixel without domain dependency. This is just what most traditional triangle methods do not possess.

Environmental conditions in equine indoor arenas: A descriptive study

Indoor arenas do not always include mechanical ventilation or stirring fans and occupancy by horses and humans can be sporadic and inconsistent, which creates a challenging space for understanding and predicting variations in temperature, moisture, and airflow. To understand the interior environment within indoor arenas, monitoring was conducted at 15 facilities within 200 kilometres of Lexington, KY. Environmental monitoring of dry bulb temperature, relative humidity, dew point temperature, air speeds, and solar radiation took place over 7 days in the winter and summer to examine temporal variability. Environmental data was collected every 5 minutes using the HOBO RX3000 Remote Monitoring Station with the HOBOnet Temp/RH Sensor, HOBOnet Solar Radiation (Silicon Pyranometer) Sensor, and HOBOnet Ultrasonic Wind Speed and Direction Sensor. Clear seasonal differences and diurnal patterns were evident in all environmental conditions, but the relative humidity. The relative humidity and dew point temperatures indicated moisture could be an issue in many of the indoor arenas. High relative humidity and excess moisture can negatively impact horse and human health as well as the lifespan of the facility. Similar results to previous spatial variability indoor arena characterizations were observed during the environmental monitoring with air speeds being below the threshold for still air in livestock facilities (0.51 m s-1). Sensor technology and implementation provides a better understanding of the interior environment and how indoor arena design can impact it.

THE ROLE OF AIR TEMPERATURE AND DEW POINT DURING FOG RECURRENCES ON THE ABSHERON PENINSULA (AZERBAIJAN)

The article focuses on the characteristics of air temperature and dew point changes observed during repeated fog events in the Absheron Peninsula in 2000...2022. For this purpose, continuous observation data of Heydar Aliyev International Airport was used. Here is an analysis of all types of fog by month of occurrence. The limits of total fog, Meteorological Optical Range ≤ 500 m. and 501...1000 m. are considered in the analysis. Repeating such criteria, attention was paid to the recorded air temperature and dew point indicators for I...III, IV...VI and X...XII months of the year. The analyzes show that the fogs observed in the peninsula are mostly recorded at a temperature of 6...8 0C. The most commonly observed dew point temperature ranges of 0,0...0,3 0C and 1,0...1,2 0C at all ranges of MOR in fog repeats. The results of the research are of particular importance for the planning of the work of all transport areas and the forecast of fogs.

Evaluation of three suction pretreatment systems for the air compressor

Air compressor, as the power source of pneumatic tools, suffers the problems of low efficiency and high failure rate in summer. In this paper, three suction air pretreatment systems are presented, which are characterized by the cold sources: a vapor compression refrigeration unit, a water chilling unit and a combination of water chilling unit with cooling tower. Models for the thermodynamic and economic performance evaluation of the three systems are established and solved. The moisture diffusion coefficient is applied to consider the influence of the moisture condensation and the Marshall and Swift coefficients is introduced to consider the capital change over time. Formula to calculate the efficiency of refrigeration compressor is adopted. Besides, the system total efficiency, percentage of power saving (PPS) and payback period (PBP) are defined as performance indexes. A comparative study with typical meteorological data of hot-dry regions and hot-humid regions is conducted to analyze the superiority of each system under the operation conditions of varied suction air temperature and volume flow. According to the results, optimal spray water flowrate exists for each operation condition and an air–water ratio of 0.8 is recommended. When the suction air temperature decreases, the PBP of each system increases and a sharp increase occurs when the temperature is lower than the air dew point temperature. A suction air temperature of 26 ℃ is suggested. Direct air cooling with an evaporator is preferred when the suction air volume flowrate is smaller than 60 m3/min for hot-humidity regions. The combined system is significantly superior in hot-dry regions. The PBP of the three systems is all longer than 10 months. How to guarantee the usage rate is the key for enterprises to consider the use of suction air pretreatment systems.

Application study of radiant floor compo site air-conditioning system for heating and cooling

Radiant floor air-conditioning systems are characterized by energy efficiency and comfort. However, at higher sensible heat loads and higher humidity conditions, problems such as condensation, poor comfort, slow thermal response, and difficulty in zoning control may occur. To mitigate these limitations, this study presents a composite air-conditioning system for radiant flooring that includes an air source heat pump., an underfloor coil, a fan coil and a dehumidifier for fresh outdoor air. The composite system was experimentally tested and analyzed for indoor thermal comfort under various operation modes in summer and winter in a specific thermal zone. On the basis of the experiments, the composite system comprises multiple energy supply points, and varying combinations between these functional endpoints can fulfill the comfort needs of the human body even in year-round working conditions. The radiant floor temperature varies from the dew point air temperature by roughly 7°C. Following the test day, summer radiant systems take approximately 30 minutes to reach the set indoor temperature and humidity level.. An hour after the indoor air has reached the desired temperature and humidity, the winter radiation system is activated.

Evaluating the Utility of Selected Machine Learning Models for Predicting Stormwater Levels in Small Streams

The consequences of climate change include extreme weather events, such as heavy rainfall. As a result, many places around the world are experiencing an increase in flood risk. The aim of this research was to assess the usefulness of selected machine learning models, including artificial neural networks (ANNs) and eXtreme Gradient Boosting (XGBoost) v2.0.3., for predicting peak stormwater levels in a small stream. The innovation of the research results from the combination of the specificity of small watersheds with machine learning techniques and the use of SHapley Additive exPlanations (SHAP) analysis, which enabled the identification of key factors, such as rainfall depth and meteorological data, significantly affect the accuracy of forecasts. The analysis showed the superiority of ANN models (R2 = 0.803–0.980, RMSE = 1.547–4.596) over XGBoost v2.0.3. (R2 = 0.796–0.951, RMSE = 2.304–4.872) in terms of forecasting effectiveness for the analyzed small stream. In addition, conducting the SHAP analysis allowed for the identification of the most crucial factors influencing forecast accuracy. The key parameters affecting the predictions included rainfall depth, stormwater level, and meteorological data such as air temperature and dew point temperature for the last day. Although the study focused on a specific stream, the methodology can be adapted for other watersheds. The results could significantly contribute to improving real-time flood warning systems, enabling local authorities and emergency management agencies to plan responses to flood threats more accurately and in a timelier manner. Additionally, the use of these models can help protect infrastructure such as roads and bridges by better predicting potential threats and enabling the implementation of appropriate preventive measures. Finally, these results can be used to inform local communities about flood risk and recommended precautions, thereby increasing awareness and preparedness for flash floods.

Using a Model to Compare Cooling System Design Factors For Lactating Cows

Highlights Cow mass, milk yield, solar load, air velocity, dry-bulb and dew-point temperature impact lactating cow respiration rates. Modeled respiration rates align with published studies dealing with milk yield, solar load, and evaporative cooling. The model can be used to assess cooling systems for conditions that might be difficult to capture in field measurements. Abstract. The risk of lactating cows experiencing heat stress is increasing because of increased productivity. A steady-state process-based model described by Nelson and Janni (2023) and assessed by Janni et al. (2023) was used to compare additional combinations of environmental conditions (e.g., air dry-bulb and dew-point temperatures, air velocity, solar load) and lactating cow characteristics (e.g., body mass and daily milk yield) on modeled respiration rate. Respiration rate was linked to qualitative heat stress levels, based on work by Renaudeau et al. (2012). The model demonstrates how cows with greater body mass have larger inspired volumes of respired air (i.e., liters per minute) because they have larger tidal volumes (i.e., liters per breath) while they have lower respiration rates (i.e., breaths per minute) compared to cows with smaller body masses. The modeled results support previous studies that demonstrated heat stress increases with greater daily milk yields, solar load, and air dry-bulb and dew-point temperatures. The model finds conditions where increasing air velocity at cow level can help cows reduce heat stress levels due to solar load, and high air dry-bulb and dew-point temperatures. A comparison to a field study demonstrates the need to align field data collection and model input needs to explore the actual or potential impacts of mitigation methods like shade, evaporative cooling, and air velocity (i.e., via fans). This report demonstrates how the model can be used by engineers to investigate the impact of ventilation system designs on respiration rates using current and future technologies. Keywords: Air velocity, Cow mass, Dairy, Dew-point temperature, Dry-bulb temperature, Heat stress, Milk yield, Respiration rate, Solar load.

A deep learning model for predicting risks of crop pests and diseases from sequential environmental data

Crop pests reduce productivity, so managing them through early detection and prevention is essential. Data from various modalities are being used to predict crop diseases by applying machine learning methodology. In particular, because growth environment data is relatively easy to obtain, many attempts are made to predict pests and diseases using it. In this paper, we propose a model that predicts diseases through previous growth environment information of crops, including air temperature, relative humidity, dew point, and CO2 concentration, using deep learning techniques. Using large-scale public data on crops of strawberry, pepper, grape, tomato, and paprika, we showed the model can predict the risk score of crop pests and diseases. It showed high predictive performance with an average AUROC of 0.917, and based on the predicted results, it can help prevent pests or post-processing. This environmental data-based crop disease prediction model and learning framework are expected to be universally applicable to various facilities and crops for disease/pest prevention.

Geographical quantification of the seasonality of transmission of COVID-19 in human populations as a function of the variability of temperatures

The occurrence of cases of COVID-19 suggests that it will likely become seasonally endemic in human populations. We seek to provide a quantification of the seasonality of the occurrence and severity of COVID-19 cases in human populations. Using global data, we show that the spatiotemporal distribution of COVID-19 cases is a function of distinct seasons and climates. We investigated this at the county and the country scale using a comparison of seasonal means, correlation analyses using ambient air temperatures and dew point temperatures, and multiple linear regression techniques. We found that most locations had the highest incidence of COVID-19 during winter compared to other seasons. Regions closer to the equator had a higher incidence of COVID-19 during the summer than regions further from the equator. Regions close to the equator, where mean annual temperatures have less variance compared to those further from the equator, had smaller differences between seasonal COVID-19 incidence. Correlation and regression analyses showed that ambient air and dew point temperatures were significantly associated with COVID-19 incidence. Our results suggest that temperature and the environment are influential factors to understand the transmission of COVID-19 within the human population. This research provides empirical evidence that temperature changes are a strong indicator of seasonal COVID-19 outbreaks, and as such it will aid in planning for future outbreaks and for mitigating their impacts.

Influence of geographical location and outdoor meteorological parameters on indoor humidity environment in rural residential buildings during the Plum Rains Season in the hot summer and cold winter region.

Plum Rains Season (PRS) has the typical characteristics of outdoor air temperature dramatic changes and high air humidity in the hot summer and cold winter region in China. When the outdoor temperature rises rapidly during PRS, the building envelope surface temperature is probably lower than the indoor air dew point temperature, resulting in moisture condensation. This paper evaluates the influence of geographical location and outdoor meteorological parameters on the indoor humidity environment. The effects of critical parameters such as altitude, average temperature, relative humidity, total precipitation, total precipitation days, atmospheric pressure, and wind speed on the building envelope moisture condensation in nine typical cities in the hot summer and cold winter region were simulated and analyzed. The results show that the Condensation Frequency (CFn) in the western, central, and eastern regions reached the highest in April, May, and June, respectively. Among the nine typical cities, Changsha has the highest Condensation Risk (CR). In addition, the altitude, total precipitation, and atmospheric pressure have little effect on the indoor humidity environment, and it is not directly related to CR; The average temperature and total precipitation days were not related to CR in the western and eastern regions and positively correlated with CR in the central region; The wind speed was positively correlated with CR in the western and central regions and negatively correlated in the eastern region; The relative humidity can affect the indoor humidity environment and moisture condensation on the inner surface of walls, when the relative humidity increases, the CR increases.

Single- and double-inlet PV curtain wall systems using novel heat recovery technique for PV cooling, fresh and supply air handling: Design and performance assessment

Building integrated photovoltaic (BIPV) technology has emerged as a promising solution for serving electricity and heat demands in buildings. However, PV overheating causes reduced production, increased space cooling load, and stagnation damage. To address overheating and save energy in air conditioning, this study proposed novel single- and dual-inlet ventilation PV curtain wall systems (SVPV and DVPV). In summer, the building exhaust is introduced into the channel to strengthen PV cooling, while incoming fresh air is used to preheat dew-point air. In winter, warm exhaust within the channel decreases heat loss through the facade and then raises the fresh air intake temperature. Mathematical models were developed for the systems, and a comprehensive performance evaluation criterion (PEC) was defined. Economic analysis was conducted using simple payback (SP) and net present value (NPV) models. Additionally, the impact of ambient temperature and solar radiation on system performance was investigated. The case study demonstrates that SVPV and DVPV reduced annual energy consumption by 957.23 kWh (28.07%) and 910.97 kWh (26.72%), respectively, compared to a conventional non-ventilated system. SVPV slightly outperformed DVPV, with an average PEC of 0.556 and 0.544, respectively. This study provides valuable insights for operational strategies and future large-scale promotion of BIPV applications.

INFLUENCE OF METEOROLOGICAL VARIABLES AND GEOGRAPHIC FACTORS IN THE SELECTION OF SOYBEAN LINES

This study aimed to evaluate the influence of meteorological variables and geographic factors on the selection of soybean lines concerning grain yield in Brazil and Paraguay soybean-producing regions. The study was conducted in seven different environments: Bela Vista do Norte - PY, Palotina - PR, Mangueirinha - PR, Major Vieira - SC, Três Passos - RS, Toledo – PR, and Passo Fundo - RS. The randomized block design in an incomplete factorial scheme with six soybean genotypes (G1, G2, G3, G4, G5, and G6) was used for the experiments. The harvest occurred in the first half of March, and grain yield was measured through the total harvest of the plot and expressed in kg ha-1, with grain moisture at 13%. The climatic variables used in the study were maximum air temperature (Tmax, ºC), average air temperature (Tavg, ºC), minimum air temperature (Tmin, ºC), relative air humidity (RH, %), precipitation (Prec, mm), wind speed (WS, m/s), dew point (DP, °C), incident radiation (Rad_Inc, MJ/m²), and total radiation (RAD_OL, MJ/m²); and geographic factors were altitude (ALT), longitude (LON), and latitude (LAT). The G5 genotype with a genetic value for grain yield above the general average is the most adapted to favorable environments. Altitude had the greatest influence on the biological variability of the genotypes, with a negative correlation of moderate magnitude with grain yield. Grain yield was enhanced in environments with altitudes lower than 338 m at latitudes below 24.17S.

A multi-mode packaged dedicated outdoor air system embedded with a four-coil heat pump

Dedicated outdoor air system (DOAS) plays a critical role in building ventilation. In recent years, much attention has been paid to the DOAS with heat-pump heat recovery. It is expected to have better dehumidifying performance and more operating modes to efficiently cooperate with the parallel cooling system. This work proposes a novel packaged DOAS embedded with a four-coil heat pump. Of the four coils, supply air coil and exhaust air coil are basic components of vapor compression heat pump that facilitates heat recovery from exhaust air. Return air coil is added to reduce heat loss in the mixing of return air and additional outdoor air. Supply air coil is introduced to reheat fresh air and avoid moisture condensation on air ducts. Meanwhile, refrigerant subcooling is increased in heat pump cycle for better efficiency. With simple switch the DOAS can also operate in four modes, i.e., outdoor air dehumidification, outdoor air low-load dehumidification, outdoor air heating and return air dehumidification. A functional prototype of 300 m3/h supply air is made and tested. Dehumidification capacity of 101.4 L/day and specific moisture extraction rate (SMER) of 3.22 kg/kWh are reached under the design condition of 30 °C/27 °C. Compared to the system without supply air coil, the new DOAS can see an improvement of 46.3% in SMER at the same supply air humidity ratio of 8.5 g/kga. It also keeps the supply air temperature above indoor air dew point. At last, the system's multi-mode operation for varying outdoor conditions is discussed.

Application of vine copulas to estimate dew point temperature

In this study, the accuracy of the copula-based model in the simulation of the dew point temperature in various climates of Iran was investigated, using simulations based on vine copulas such as C-, D-, and R-vine copulas. By examining the various vine copulas and their tree sequences, the best copula and best tree sequence based on AIC, BIC, and log-likelihood were selected. The results show that based on the complete similarity in our case between C-, D- and R-vine copulas, the selected best C-vine copulas fit well the dependence between the minimum and maximum air temperatures and dew point temperature. The simulation results were analyzed using root mean square error (RMSE), Nash-Sutcliffe efficiency (NSE) coefficient, and violin plots. The results show that the copula-based model has high accuracy at all stations. The min (max) RMSE is related to Kerman (Ahvaz) station with RMSE = 0.396 oC (0.617 oC). Also, the min (max) NSE is related to Ahvaz (Urmia) station with NSE = 0.925 (0.955). Also, according to the violin plot, it is possible to appreciate the acceptable certainty of the copula-based model. Due to the diversity of the tree sequences of vine copulas and the use of the rotated states of the internal vine copulas, as well as the possibility of interfering with the effective parameters in high dimensions, the simulation results are reliable and have no restrictions. This model can be used as the best model to estimate dew point temperature due to the full coverage of the range of changes in data.

PM2.5 Concentration Prediction in Six Major Chinese Urban Agglomerations: A Comparative Study of Various Machine Learning Methods Based on Meteorological Data

The escalating issue of air pollution in China’s rapidly developing urban areas has prompted increased attention to the role of meteorological conditions in PM2.5 pollution. This study examines the spatiotemporal distribution of PM2.5 concentrations and their relationship with meteorological factors in six major Chinese urban agglomerations from 2017 to 2020, using daily average data. Statistical and spatial analysis techniques are employed, alongside the construction of eight machine learning models for prediction purposes. The study also compares the feature importance of various meteorological factors impacting PM2.5 concentrations. Results reveal significant regional differences in both average PM2.5 levels and meteorological influences. The Multilayer Perceptron (MLP) model demonstrates the highest prediction accuracy for PM2.5 concentrations. According to the MLP model’s feature importance identification, temperature is the most significant factor affecting PM2.5 concentrations across all urban agglomerations, while wind speed and precipitation have the least impact. Contributions from air pressure and dew point temperature, however, vary among different urban agglomerations. This research considers the impact of urban agglomerations and meteorological conditions on PM2.5 and also offers valuable artificial intelligence-based insights into the key meteorological factors influencing PM2.5 concentrations in diverse regions, thereby informing the development of effective air pollution control policies.

Recurrent neural network modeling of multivariate time series and its application in temperature forecasting.

Temperature forecasting plays an important role in human production and operational activities. Traditional temperature forecasting mainly relies on numerical forecasting models to operate, which takes a long time and has higher requirements for the computing power and storage capacity of computers. In order to reduce computation time and improve forecast accuracy, deep learning-based temperature forecasting has received more and more attention. Based on the atmospheric temperature, dew point temperature, relative humidity, air pressure, and cumulative wind speed data of five cities in China from 2010 to 2015 in the UCI database, multivariate time series atmospheric temperature forecast models based on recurrent neural networks (RNN) are established. Firstly, the temperature forecast modeling of five cities in China is established by RNN for five different model configurations; secondly, the neural network training process is controlled by using the Ridge Regularizer (L2) to avoid overfitting and underfitting; and finally, the Bayesian optimization method is used to adjust the hyper-parameters such as network nodes, regularization parameters, and batch size to obtain better model performance. The experimental results show that the atmospheric temperature prediction error based on LSTM RNN obtained a minimum error compared to using the base models, and these five models obtained are the best models for atmospheric temperature prediction in the corresponding cities. In addition, the feature selection method is applied to the established models, resulting in simplified models with higher prediction accuracy.