Can complex networks describe the urban and rural tropospheric O3 dynamics?

Research on regional ozone prevention and control strategies in eastern China based on pollutant transport network and FNR

O3 pollution in China has worsened sharply in recent years, and O3 formation sensitivity (OFS) in many regions have gradually changed, with eastern China as the most typical region. This study constructed the transport networks of O3 and NO2 in different seasons from 2017 to 2020. The transport trends and the clustering formation patterns were summarized by analyzing the topological characteristics of the transport networks, and the patterns of OFS changes were diagnosed by analyzing the satellite remote sensing data. Based on that, the main clusters that each province or city belongs to in different pollutant transport networks were summarized and proposals for the inter-regional joint prevention and control were put forward. As the results showed, O3 transport activity was most active in spring and summer and least active in winter, while NO2 transport activity was most active in autumn and winter and least active in summer. OFS in summer mainly consisted of transitional regimes and NOx-limited regimes, while that in other seasons was mainly VOC-limited regimes. Notably, there was a significant upward trend in the proportion of transitional regimes and NOx-limited regimes in spring, autumn, and winter. For regions showing NOx-limited regime, areas with higher out-weighted degrees in the NO2 transport network should focus on controlling local NOx emissions, such as central regions in summer. For regions showing VOC-limited regime, areas with higher out-weighted degrees in the O3 transport network should focus on controlling local VOCs emissions, such as central and south-central regions in summer. For regions that belong to the same cluster and present the same OFS in each specific season, regional cooperative emission reduction strategies should be established to block important transmission paths and weaken regional pollution consistency.

Ground air pollutants explanation based on multiple visibility graph of complex network by temporal community division

In air pollution studies, the correlation analysis of environmental variables has usually been challenged by parametric diversity. Such variable variations are not only from the extrinsic meteorological conditions and industrial activities but also from the interactive influences between the multiple parameters. A promising solution has been motivated by the recent development of visibility graph (VG) on multi-variable data analysis, especially for the characterization of pollutants' correlation in the temporal domain, the multiple visibility graph (MVG) for nonlinear multivariate time series analysis has been verified effectively in different realistic scenarios. To comprehensively study the correlation between pollutant data and season, in this work, we propose a multi-layer complex network with a community division strategy based on the joint analysis of the atmospheric pollutants. Compared to the single-layer-based complex networks, our proposed method can integrate multiple different atmospheric pollutants for analysis, and combine them with multivariate time series data to obtain higher temporary community division for ground air pollutants interpretation. Substantial experiments have shown that this method effectively utilizes air pollution data from multiple representative indicators. By mining community information in the data, it successfully achieves reasonable and strong interpretive analysis of air pollution data.

Spatial clustering and spillover pathways analysis of O3, NO2, and CO in eastern China during 2017-2021

The eastern China presented the most serious O3 pollution and increasingly prominent regional characteristics. To understand the transport characteristics of O3 and its precursors and identify their potential relationships are of great guiding significance for interregional joint prevention and control. In this study, the annual and seasonal transport networks of O3 and its precursors (NO2 and CO) during 2017-2021 were constructed by applying the complex network method to air quality observations. And the key spatial clusters, the spillover paths and the potential links among pollutants were comprehensively analyzed based on the topological characteristic analysis of the established air pollutant transport networks. As the results showed, O3 pollution in the eastern China was affected by active regional transports of O3 and its precursors. Regional transports of O3, NO2, and CO were more prominent in autumn and showed high synchronization. The regional transport of precursors, especially NOx, was an important cause of regional O3 pollution. Air pollutant transport characteristics varied with seasons and regions, which demonstrating the importance of regulating seasonal and regional differentiated joint prevention and control strategies, especially for NOx. The results of this study can provide science-based guidance for the regional cooperative control of O3 pollution in the eastern China, and the application of complex networks can also provide a new methodological perspective for the study of air pollution transmission.

Graph Machine Learning for Improved Imputation of Missing Tropospheric Ozone Data

Gaps in the measurement series of atmospheric pollutants can impede the reliable assessment of their impacts and trends. We propose a new method for missing data imputation of the air pollutant tropospheric ozone by using the graph machine learning algorithm "correct and smooth". This algorithm uses auxiliary data that characterize the measurement location and, in addition, ozone observations at neighboring sites to improve the imputations of simple statistical and machine learning models. We apply our method to data from 278 stations of the year 2011 of the German Environment Agency (Umweltbundesamt - UBA) monitoring network. The preliminary version of these data exhibits three gap patterns: shorter gaps in the range of hours, longer gaps of up to several months in length, and gaps occurring at multiple stations at once. For short gaps of up to 5 h, linear interpolation is most accurate. Longer gaps at single stations are most effectively imputed by a random forest in connection with the correct and smooth. For longer gaps at multiple stations, the correct and smooth algorithm improved the random forest despite a lack of data in the neighborhood of the missing values. We therefore suggest a hybrid of linear interpolation and graph machine learning for the imputation of tropospheric ozone time series.

O3 transport characteristics in eastern China in 2017 and 2021 based on complex networks and WRF-CMAQ-ISAM

Increasingly prominent pollution levels and strong regional characteristics of O₃, especially in economically developed eastern China, called for a regional cooperation strategy based on transport quantification. This study adopted the complex networks to construct the O₃ Transport Network (OTN) to explore characteristics in eastern China in the summer of 2017 and 2021, whose results were afterward verified with spatial source apportionment results simulated with WRF-CMAQ-ISAM. As OTN suggested, O₃ transport showed stronger and faster characteristics in eastern China in 2021 than in 2017, judging from changes in the network density, number of connections, transport ranges, and transport paths. Among all cluster communities, inland Shandong was the most important O₃ transport hub, the Central Community was the largest community, and the Southern Community showed the closest inter-city transport relationships. In- and out-weighted degrees in OTN showed relatively superior consistency with the transport matrix obtained with WRF-CMAQ-ISAM, and can be explained by wind fields. Generally, O₃ pollution in the whole eastern China showed more frequent intra-regional transport and more strengthened inter-city correlations in 2021 than in 2017, meanwhile, northerly and southerly cities exhibited strengthening and weakening trends in O₃ transport, respectively. Despite the completely different principles of complex networks and air quality models, their results were mutually verifiable. This study presented a comprehensive understanding of O₃ transport in eastern China for further formulation of regional collaborative strategies and provided the methodological verification for applying complex networks in the atmospheric environment field.

Doing more with less: How to design a good subgroup governance model for the air pollution transport network in "2+26" cities of China?

Air pollution shares the attributes of significant spatial spillover effects and environmental public goods, leading to the territorial governance model that easily falls into a state of failure. Despite a growing number of studies on the local spatial spillover effect of air pollution, scant evidence currently exists on its global spatial association effect and a good subgroup governance model. Based on a panel data set of the daily prefecture-level city data on air quality measured by the air quality index (AQI) in "2 + 26" cities of China in 2015 and 2018, this study first builds an air pollution transport network (APTN), i.e., the cities as the nodes and the association relationships between the nodes as the edges. Furthermore, this paper reveals the spatial association effect and the temporal lagged attribute of the APTN using the Social network analysis (SNA) and the Generalized impulse response function (GIRF). The results are summarized as follows. (1) Every city has significant spatial association effects of air pollution with at least another city in the APTN, and northern APTN affects most to the air pollution of other cities, while southern APTN is obviously always affected by air pollution in other cities. (2) Transport strength peaks on the second day of an air pollution transport process, and the transport process lasts for 7-12 days. (3) The APTN is divided into four subgroups: Sycophants, Primary, Bidirectional, and Brokers, with Baoding, Zhengzhou, Heze, and Hengshui as the central cities of each group, respectively. Overall, our study provides a networked, modular, and early-warning governance model for policymakers.

Spatiotemporal heterogeneity and influencing factors on urbanization and eco-environment coupling mechanism in China

High-quality urbanization is the core for realizing human well-beings, for which reason investigating how the relationship evolves between urbanization and eco-environment is of crucial importance. Differing from the rationale of revealing spatial spillover effects using traditional tests, we consider spatial network characteristics to enrich the notion of local coupling and telecoupling from a relational perspective. First, we adopt coupling coordination degree model (CCDM) and decoupling model (DM) to calculate the urbanization and eco-environment coupling coordination degree (UECCD) and the decoupling index (DI) in 30 provinces and municipalities of China from 2008 to 2017. Second, we use gravity model to construct urbanization and eco-environment coupling coordination network (UECCN), in which provinces are nodes and spatial connection relationships of UECCD are edges between nodes. Third, we introduce social network analysis (SNA) to reveal spatial network characteristics of UECCN without using local spatiotemporal heterogeneity. Finally, we employ spatial econometric model to reveal factors that influence urbanization and eco-environment coupling effect. The major findings and conclusions of this study are summarized as follows. (1) The main subclasses of UECCD and DI are basically uncoordinated patterns with eco-environment lagging and weak decoupling, respectively. (2) Only two spatial agglomeration types of UECCD exist, the high-high (H-H) clustering in Shanghai and the low-low (L-L) clustering in western China, whereas no significant spatial agglomeration effect is observed among most provinces. (3) The distribution characteristics of UECCN are sparse in western China and dense in eastern China, and the spatial correlation strength of UECCN improves. (4) Technological innovation plays a critical role in promoting UECCD, while the total population, per capita disposable income, coupling network structure, and environmental regulations exert significant impact on UECCD. Collectively, we propose to prioritize governance provinces with low UECCD in western China as well as adequately utilize the positive externalities of key node provinces in eastern China. Equally importantly, we suggest that it is also critical to fully exert a driving force of technological innovation on improving the UECCD by promoting renewable energy utilization.

Influential Nodes Identification in the Air Pollution Spatial Correlation Weighted Networks and Collaborative Governance: Taking China's Three Urban Agglomerations as Examples

Nowadays, driven by green and low-carbon development, accelerating the innovation of joint prevention and control system of air pollution and collaborating to reduce greenhouse gases has become the focus of China’s air pollution prevention and control during the “Fourteenth Five-Year Plan” period (2021–2025). In this paper, the air quality index (AQI) data of 48 cities in three major urban agglomerations of Beijing-Tianjin-Hebei, Pearl River Delta and Yangtze River Delta, were selected as samples. Firstly, the air pollution spatial correlation weighted networks of three urban agglomerations are constructed and the overall characteristics of the networks are analyzed. Secondly, an influential nodes identification method, local-and-global-influence for weighted network (W_LGI), is proposed to identify the influential cities in relatively central positions in the networks. Then, the study area is further focused to include influential cities. This paper builds the air pollution spatial correlation weighted network within an influential city to excavate influential nodes in the city network. It is found that these influential nodes are most closely associated with the other nodes in terms of spatial pollution, and have a certain ability to transmit pollutants to the surrounding nodes. Finally, this paper puts forward policy suggestions for the prevention and control of air pollution from the perspective of the spatial linkage of air pollution. These will improve the efficiency and effectiveness of air pollution prevention and control, jointly achieve green development and help achieve the “carbon peak and carbon neutrality” goals.

Spatiotemporal evolution of NO2 diffusion in Beijing in response to COVID-19 lockdown using complex network

The COVID-19 pandemic and the corresponding lockdown measures have been confirmed to reduce the air pollution in major megacities worldwide. Especially at some monitoring hotspots, NO₂ has been verified to show a significant decrease. However, the diffusion pattern of these hotspots in responding to COVID-19 is not clearly understood at present stage. Hence, we selected Beijing, a typical megacity with the strictest lockdown measures during COVID-19 period, as the studied city and attempted to discover the NO₂ diffusion process through complex network method. The improved metrics derived from the topological structure of the network were adopted to describe the performance of diffusion. Primarily, we found evidences that COVID-19 had significant effects on the spatial diffusion distribution due to combined effect of changed human activities and meteorological conditions. Besides, to further quantify the impacts of disturbance caused by different lockdown measures, we discussed the evolutionary diffusion patterns from lockdown period to recovery period. The results displayed that the difference between normal operation and pandemic operation firstly increased at the cutoff of lockdown measures but then declined after the implement of recovery measures. The source areas had greater vulnerability and lower resilience than receptors areas. Furthermore, based on the conclusion that the diffusion pattern changed during different periods, we explored the key stations on the path of diffusion process to further gain information. These findings could provide references for comprehending spatiotemporal pattern on city scale, which might be help for high-resolution air pollution mapping and prediction.

A complex network approach for the model of vehicle emission propagation and intelligently mine the interaction rules

With the increasing number of motor vehicles, exhaust emission has become a major source of urban pollution. Most studies are limited to the prediction of pollutant concentration, which cannot clearly indicate the change of pollution emissions and regional relationship. In this paper, we propose an emission propagation model of vehicle source pollution based on complex network in order to intelligently mine the interaction and propagation rules hidden behind dynamic spatiotemporal data. First, aiming at the problems of low resolution and insufficient data volume of vehicle emission data, a high-resolution pollution emission data is generated based on the COPERT (Computer Program to Calculate Emissions from Road Transport). For study the influence of causality between regions, a propagation model is designed based on the convergent cross mapping method to transform the emission time series into a complex network. In addition, we propose a novel key node mining algorithm using hybrid local and global information to identify areas of heavy pollution. Experimental results on real datasets demonstrate that the spread of pollution follows certain rules and is also affected by regional influences. Moreover, the proposed algorithm is superior to the state-of-the-art methods.

Analysis of Air Mean Temperature Anomalies by Using Horizontal Visibility Graphs

The last decades have been successively warmer at the Earth’s surface. An increasing interest in climate variability is appearing, and many research works have investigated the main effects on different climate variables. Some of them apply complex networks approaches to explore the spatial relation between distinct grid points or stations. In this work, the authors investigate whether topological properties change over several years. To this aim, we explore the application of the horizontal visibility graph (HVG) approach which maps a time series into a complex network. Data used in this study include a 60-year period of daily mean temperature anomalies in several stations over the Iberian Peninsula (Spain). Average degree, degree distribution exponent, and global clustering coefficient were analyzed. Interestingly, results show that they agree on a lack of significant trends, unlike annual mean values of anomalies, which present a characteristic upward trend. The main conclusions obtained are that complex networks structures and nonlinear features, such as weak correlations, appear not to be affected by rising temperatures derived from global climate conditions. Furthermore, different locations present a similar behavior and the intrinsic nature of these signals seems to be well described by network parameters.

Improving graph-based detection of singular events for photochemical smog agents

Recently, a set of graph-based tools have been introduced for the identification of singular events of O₃, NO₂ and temperature time series, as well as description of their dynamics. These are based on the use of the Visibility Graphs (VG). In this work, an improvement of the original approach is proposed, being called Upside-Down Visibility Graph (UDVG). It adds the possibility of investigating the singular lowest episodes, instead of the highest. Results confirm the applicability of the new method for describing the multifractal nature of the underlying O₃, NO₂, and temperature. Asymmetries in the NO₂ degree distribution are observed, possibly due to the interaction with different chemicals. Furthermore, a comparison of VG and UDVG has been performed and the outcomes show that they describe opposite subsets of the time series (low and high values) as expected. The combination of the results from the two networks is proposed and evaluated, with the aim of obtaining all the information at once. It turns out to be a more complete tool for singularity detection in photochemical time series, which could be a valuable asset for future research.

Checking complex networks indicators in search of singular episodes of the photochemical smog

A set of indicators derived from the analysis of complex networks have been introduced to identify singularities on a time series. To that end, the Visibility Graphs (VG) from three different signals related to photochemical smog (O₃, NO₂ concentration and temperature) have been computed. From the resulting complex network, the centrality parameters have been obtained and compared among them. Besides, they have been contrasted to two others that arise from a multifractal point of view, that have been widely used for singularity detection in many fields: the Hölder and singularity exponents (specially the first one of them). The outcomes show that the complex network indicators give equivalent results to those already tested, even exhibiting some advantages such as the unambiguity and the more selective results. This suggest a favorable position as supplementary sources of information when detecting singularities in several environmental variables, such as pollutant concentration or temperature.

A sliding window-based algorithm for faster transformation of time series into complex networks

A new alternative method to approximate the Visibility Graph (VG) of a time series has been introduced here. It exploits the fact that most of the nodes in the resulting network are not connected to those that are far away from them. This means that the adjacency matrix is almost empty, and its nonzero values are close to the main diagonal. This new method is called Sliding Visibility Graph (SVG). Numerical tests have been performed for several time series, showing a time efficiency that scales linearly with the size of the series [O(N)], in contrast to the original VG that does so quadratically [O(N²)]. This fact is noticeably convenient when dealing with very large time series. The results obtained from the SVG of the studied time series have been compared to the exact values of the original VG. As expected, the SVG outcomes converge very rapidly to the desired ones, especially for random and stochastic series. Also, this method can be extended to the analysis of time series that evolve in real time, since it does not require the entire dataset to perform the analysis but a shorter segment of it. The length segment can remain constant, making possible a simple analysis as the series evolves in time.