Distinct urban-rural gradients of air NO2 and SO2 concentrations in response to emission reductions during 2015-2022 in Beijing, China

Does urban-rural integration contribute to environmental health? Exploring the interplay between urban-rural integration and air quality dynamics in Yangtze River middle reaches city cluster

Background

Exploring the coordinated relationship between urban-rural integration and air quality has significant implications for promoting urban-rural development, preventing air pollution and ensuring residents' health. This study takes Yangtze River middle reaches city cluster as a case study, calculates the levels of urban-rural integration and air quality development, analyzes their coupled coordination relationship and driving factors, and explores the path of coordinated development.

Methods

This study constructs a coupling coordination degree model to analyze the relationship between the urban-rural integration development level and air quality development level. We use the trend surface method to analyze the spatial divergence characteristics of the coordination degree between urban-rural integration and air quality. In addition, we used a geographic detector to analyze the factors affecting the coordination degree.

Results

(1) The overall level of urban-rural integration development showed an upward trend. High-value regions were concentrated in the Wuhan, Chang-Zhu-Tan, and Nanchang metropolitan areas. (2) The Air Quality Index showed an overall decline, with the most significant improvements observed in Wuhan, Changsha, and Jiujiang. (3) The coupling degree increased from 0.570 in 2013 to 0.794 in 2021, and the coordination degree increased from 0.337 in 2013 to 0.591 in 2021. The link between urban-rural integration and air quality has deepened over time, and the two promote each other, making city cluster develop towards environmental friendliness. The spatial distribution of coordination degree shows a "high in the west and low in the east, high in the north and low in the south" trend. (4) Per capita GDP, non-agricultural employment ratio, urban-rural spatial circulation media, population urbanization level, and fixed asset investment were identified as the core factors driving the coordination degree between urban-rural integration and air quality.

Conclusion

This study confirms that the urban-rural integration and air quality of Yangtze River middle reaches city cluster are gradually changing in the direction of high-quality coordination. However, there are great differences among cities, regional imbalance is prominent, and coordination degree is driven by multidimensional factors.

Improved PM2.5 prediction with spatio-temporal feature extraction and chemical components: The RCG-attention model

Deep learning models are widely used for PM2.5 prediction. However, neglecting temporal and spatial characteristics leads to low prediction accuracy. In this work, a new deep learning model (RCG - Attention model) was developed, which combines the residual neural network (ResNet) and the convolution gated recurrent network (ConvGRU) and is applied to extract the spatio - temporal features for predicting PM2.5 concentration over the subsequent 24 h. The ResNet extracts the spatial distribution features of pollutants, and the ConvGRU extracts temporal features. The spatial and temporal features are fused by the multi - head attention mechanism to obtain multi - dimensional features. These features are finally fed into a series of fully connected layers to predict the future results. Incorporating these chemical components enhances the scientific validity of the dataset and strengthens the inherent logical connections among variables. The Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), and R - squared (R2) results indicate that the prediction performance of the RCG - Attention model surpasses that of other baseline models. The model demonstrates superior prediction performance across multiple monitoring stations, suggesting robust generalization capabilities and adaptability for various regions in one city. The SHAP results show that PM10, NO2, RH, NO3-, OC and NH4+ are significant influencing features. The RCG - Attention model provides a comprehensive solution for PM2.5 concentration prediction by integrating spatial and temporal feature extraction with chemical components.

When and why PM2.5 is high in Seoul, South Korea: Interpreting long-term (2015-2021) ground observations using machine learning and a chemical transport model

Seoul has high PM2.5 concentrations and has not attained the national annual average standard so far. To understand the reasons, we analyzed long-term (2015-2021) hourly observations of aerosols (PM2.5, NO3-, NH4+, SO42-, OC, and EC) and gases (CO, NO2, and SO2) from Seoul and Baekryeong Island, a background site in the upwind region of Seoul. We applied the weather normalization method for meteorological conditions and a 3-dimensional chemical transport model, GEOS-Chem, to identify the effect of policy implementation and aerosol formation mechanisms. The monthly mean PM2.5 ranges between about 20 μg m-3 (warm season) and about 40 μg m-3 (cold season) at both sites, but the annual decreasing rates were larger at Seoul than at Baengnyeong (-0.7 μg m-3 a-1 vs. -1.8 μg m-3 a-1) demonstrating the effectiveness of the local air quality policies including the Special Act on Air Quality in the Seoul Metropolitan Area (SAAQ-SMA) and the seasonal control measures. The weather-normalized monthly mean data shows the highest PM2.5 concentration in March and the lowest concentration in August throughout the 7 years with NO3- accounting for about 40 % of the difference between the two months at both sites. Taking together with the GEOS-Chem model results, which reproduced the elevated NO3- in March, we concluded the elevated atmospheric oxidant level increases in HNO3 (which is not available from the observation) and the still low temperatures in March promote rapid production of NO3-. We used Ox (≡ O3 + NO2) from the observation and OH from the GEOS-Chem as a proxy for the atmospheric oxidant level which can be a source of uncertainty. Thus, direct observations of OH and HNO3 are needed to provide convincing evidence. This study shows that reducing HNO3 levels through atmospheric oxidant level control in the cold season can be effective in PM2.5 mitigation in Seoul.