Adaption to Tianjin, China, Based on a Retrospective Pattern Study on the Petrochemical Industry Development and the Correlated Process of SO₂ Abatement in Yokkaichi, Japan

Abstract

Yokkaichi is one of the four major Japanese cities facing air pollution after World War II, owing to modern urban industrialization in the 20th century. Tianjin City, in China, also showed similar industrial patterns in the petrochemical industry. For decades, the petrochemical industry development has been deteriorating the environment with its by-product, sulfur dioxide (SO₂). In this paper, we summarized the characteristics of air pollution in Yokkaichi through a retrospective approach by comparing common features of Yokkaichi and Tianjin. We believe that Yokkaichi is at Stage 4, after the pollution stage, whereas Tianjin is currently in Stage 3. We believe that the efficacy of regional environmental policies in Yokkaichi related to SO₂ pollution can help predict the pollution pattern in Tianjin. We used an extended stochastic regression on a population, affluence, and technology model as a reference to demonstrate the feasibility of Yokkaichi’s pattern and the comparison between Yokkaichi and Tianjin. Fossil fuels, especially crude oil, may continuously be exploited as the main energy source in the next few decades. Thus, experiences of SO₂ air pollution in Yokkaichi and Tianjin’s could be of universal value. As it has been 50 years since the final judgment of the Yokkaichi Asthma and Yokkaichi Air pollution joint lawsuit, we attempted to reflect on Yokkaichi’s history to strengthen efforts to achieve future sustainable development goals.

1. Introduction

1.1. Retrospective of Air Pollutions

Among all kinds of pollution, air pollution is one of the most complicated and difficult to prevent and control. Air pollution refers to the presence of contaminants or pollutants in the air/atmosphere that interferes with human health or welfare or produces other harmful environmental effects [1]. Modern urbanization and industrialization have created enormous requirements for energy and fuel. Since the Industrial Revolution, the utilization of fossil fuels has been widely moved from mega-industrial cities to rural areas. Globally, large-scale air pollution incidents occur in industrial development, commercial activities, and even daily life. Until the 19th century, the most prominent air pollutants were smoke and ash from the burning of coal or oil in the boiler furnaces in power plants, locomotives, marine vessels, and home heating fireplaces and furnaces [2]. However, with time, the composition of air pollutants got more complex. Air pollution originating from industrialization first began in Great Britain. However, from the invention of the pumping engine and reciprocating engine to the British Parliament to start addressing the issues of pollution nationally, over 30 years have passed. However, smoke and ash abatement in Great Britain lasted until the early 20th century. Issues related to simple sources of pollutants, e.g., smoke and ash, needed quite a century to be resolved. In the first half of the 20th century, oil started replacing coal and became the pillar fuel of industrialization and urbanization, and complex air pollution incidents began to appear in, e.g., Meuse Valley of Belgium in 1930 and Los Angeles in the 1940s. Cases of interaction between the former and isolated air pollution showed a continually steady inclination. After World War II, to catch up with the pace of developed countries, condensed development strategies were accepted by Latin American and Asian countries, which caused air pollution in these countries much harder to address. Their work elaborates on general and specific approaches adopted to control emissions-small and large-scale, mobile and stationary, combustion and non-combustion. Other than the origins and natural factors of air pollution, its negative effects on human health were also discussed. A review study by B. Brunekreef and S. Holgate stated that people exposed to airborne particulate matter and ozone (O₃) tend to have a higher risk of mortality and get respiratory and cardiovascular diseases [3]. Moreover, the economic growth of a country is associated with air pollution as a by-product. In the study of G. Grossman and A. Krueger, three air pollutants across the urban areas located in 42 countries were used to determine the relationship between air quality and economic growth, and multiple aspects were analyzed to study possible changes and pressures that are given to the environment with the aim of encouraging more active trade and investment. The concentration of sulfur dioxide (SO₂) was found to be increasing with per capita gross domestic product (GDP, Table S1) at a low level of national income but decreasing with GDP growth at higher levels of income [4]. In comparing the air pollution incidents before and after World War II, air pollution in Asian counties is more intricate. In Japan, four major epidemics: Minamata, Miigata Minamata, and Itaiitai disease, and Yokkaichi Asthma, occurred within 15 years after World War II due to water, soil, and air pollution. Pollutants do not only endanger human health directly but also indirectly through agriculture and fisheries. In China, public health has been constantly threatened by all kinds of pollution since 1990. Thus, the correlations between urbanization and pollution in China have been rigorously studied by scholars and researchers. Either in long- or short-term exposures, the associations of daily SO₂ concentrations with overall health and cardio-respiratory diseases are apparent. Therefore, urgent action must be planned to resolve pollution issues in China, even though SO₂ pollution was not specifically targeted [5,6]. Daily emergency hospital visits were marked correlated with a particulate matter with a diameter of less than 2.5 and 10 μm (PM_2.5 and PM₁₀, respectively), nitrogen dioxide (NO₂), and SO₂ at 3.34%, 3.96%, 5.90%, and 5.38%, respectively [7].

Many studies have discussed the environmental impacts of human economic activities. Among them, the analysis of the impacts of the North American Free Trade Agreement was an attempt [4]. Through different types of models, such as the Environmental Kuznets Curve (EKC) [8,9] and models developed from IPAT (Environmental Impact = Population ×Affluence × Technology), different types of environmental impacts, which are related to air pollution, were analyzed, such as carbon dioxide (CO₂) emission, energy use, SO₂, dust, NO₂, PM_2.5, and exhaust gases. EKC was demonstrated by researchers who were mainly dedicated to describing the pressure given to our natural environment by anthropogenic CO₂ emissions worldwide [10]; however, it does not represent a lack of attempts to evaluate sulfur emissions. Most of the sulfur EKC studies focused on the period during 1970–1990, and the turning point considered usually appears after income per capita surpassed USD 6000 [11,12,13,14,15]; therefore, the EKC model for sulfur is restrained here.

For all kinds of analyses, one of the popular and successful models is stochastic regression on population, affluence, and technology (STIRPAT). It was derived from a fundamental theory of IPAT given by P. Ehrlich in the 1970s, reworked, and rechecked by T. Dietz and E. Rosa to discuss the effects of population on the affluence of CO₂ emissions [16]. Then, it passed the I = PBAT discussion, in which a special driver behavior was also included [17]. A variant of the IPAT model was ImPACT, in which T is for consumption per unit of GDP (C) and impact per unit of consumption (T), so I = PACT, which discussed its developing process from the original IPAT-ImPACT-STIRPAT and compared it to explore the different kinds of driving forces for environmental impacts [18]. The defects in the application of the STIRPAT model have also been reconsidered before. It was believed that the STIRPAT model’s inconclusive results and potential knowledge gap geographically imbalanced its scope, monotonously focusing on carbon emissions, lacking a common agreement on data selection, additional explanatory variables, and regression models [10]. Still, we reckon the STIRPAT model as a useful tool for analyzing the correlation between development and environmental pollution.

1.2. SO₂ Pollution in Japan and China

SO₂-related air pollution frequently happened during 1960–1970 in multiple cities all around Japan, including Yokkaichi. The former study on Yokkaichi Air Pollution demonstrated thorough research and helped the establishment of the Yokkaichi Air Pollution Study and Yokkaichi Study in Japan [19,20]. The decade-long study elaborated on the medical relationship between the health damage of residents and SO₂ [21]. A comprehensive demonstration of the history of Yokkaichi Asthma and the related countermeasures was given by Yukimasa et al. [22]. The past dataset of Yokkaichi was examined many times to prove that SO₂ exposure increases the risk of all-cause and cause-specific mortality [23]. Air pollution incidents in multiple Chinese cities were also mentioned in the study of K. Yoshida as a comparison with Yokkaichi. The frequently occurring air pollution incidents in China since the 1990s have also drawn the attention of Japanese researchers and scholars. However, most of the studies tend to focus on isolated research problems rather than the air pollution problems of Chinese cities. The comparisons at the country level can commonly be seen in Japan–China comparative environmental studies.

In a review study about Chinese air pollution control policies, it is commonly accepted that those policies implemented before 2010 were mostly ineffective; however, the national SO₂ emission reduction goal was accomplished during 2006–2012 [24]. At the same time, the impact of environmental regulations and national plans was also frequently discussed in different modalities. In the Beijing–Tianjin–Hebei region (henceforth: BTH region, Table S1), where Tianjin belongs, the interior correlations of the environmental regulations were negated, and both direct and indirect spatial effects were verified, and the long-term promotion effect was expected [25]. In northern China, drivers of ambient air quality in Beijing were analyzed, and the cause of pollution in Beijing changed from being predominantly related to coal burning to mixed traffic exhaust and coal burning [26]. As an industrial city that is adjacent to Beijing, Tianjin has a heavier pollution burden. Its crude oil-related industry is caused by the fact that it once belonged to the Two Control Zones (henceforth: TCZ, Table S1) of China. In China, the TCZ prefectures had a higher share of industrial activities than the non-TCZ prefectures [27]. With desulfurization technologies getting off the ground in China and the implementation of more strict regulations, TCZ cities reached a better environment with less SO₂, and the concept of TCZ was abandoned after the 2000s. The SO₂ concentration in the BTH region did not follow simple linear trends but instead reflected a repercussion of environmental measures and political and economic activities. In recent decades, decreasing trends have been observed due to government efforts to restrain emissions from power and industrial sectors. National environmental pollution control measures were believed to be evidently effective [28].

The SO₂ pollution in Sichuan and Chongqing was studied. However, the substantial terrain situation in Sichuan and Chongqing restrained air circulation. Even though we tried to make a comparison between air pollution in Yokkaichi and SO₂ pollution in the southwestern China region, the factor of terrain cannot be ignored; therefore, Tianjin was selected as another subject of this study. We understand that the difficulty of comparing two cities’ processes of industrialization and pollution abatement would be harder than we imagine. However, as a typical petrochemical industrial city, Yokkaichi could give more valuable experiences to other cities having petrochemical industries. Therefore, we cannot help but wonder if there is a pattern existing in the Yokkaichi Air Pollution abatement process. Can the patterns from Yokkaichi also be applied to another city, such as Tianjin? How do we verify the feasibility of the city when applying the pattern to Tianjin? Therefore, we decided to make the following hypothesis and consider it in this research.

Hypothesis 1. 

Using SO₂ emission and industrial shipment value as the main index, there is a pattern that can be summarized to divide Yokkaichi’s pollution abating process, and the pattern can be applied in Tianjin.

This research consists of two parts. In the first part, a hypothesis is verified. In the second part, as discussed, the extended STRIPAT model is utilized to characterize the possible SO₂-industry-related environmental impacts that the objects are sustaining. First, an inductive demonstration of argumentation is adopted to summarize Yokkaichi’s process into a more intuitionistic modality.

2. Consideration for Hypothesis

2.1. Elemental Information about Yokkaichi Air Pollution

Yokkaichi (136°24′–136°40′ E, 34°54′–35°04′ N) is under the governing of Mie Prefecture, Japan (Figure 1). It is located on the east coast of Japan. On its east side is the famous Ise Bay. In Figure 1, on the west side of Yokkaichi, the Suzuka Mountains stand. Several rivers run from the mountains and meander through Yokkaichi to Ise Bay (Table S2).

During World War II, Yokkaichi city was selected for stationing No.2 Japanese Navy Fueling Depot. The navy factory laid the foundation of Yokkaichi as an industrial city. In 1955, the first Japanese large-scale petrochemical complex was established at Yokkaichi by the Japanese cabinet to operate. From 1959 to 1963, the No.1 and 2 Petrochemical Complexes officially started their operations [21]. With the help of the industrial complex, Yokkaichi finished its energy structure transformation faster than the whole country. Crude oil became the dominating energy source in 1960, whereas coal only took up approximately 5% [29,30]. To meet the needs of the rapid economic resurgence and domestic energy consumption growth of Japan, Yokkaichi has become one of the largest petrochemical industrial cities in Japan. The petrochemical complexes were initially welcomed by residents and local government because of the lucrative profit they brought into personal income and local finance. Possible environmental issues were not under consideration till then. The complex mostly imported crude oil from the Persian Gulf in the Middle East, where sulfur content was distinctly higher than in other production areas. In the meantime, efficient desulfurization measures were not taken. For a long time, heavy dust, stinky odor, and other problems that are caused by air pollutants with sulfide have been noticed [30]. Even the fishes nearby were polluted by the wastewater from the complexes. TO deal with the pollution, “The Smoke and Soot Regulation Law” was implemented by the local government. The death rate due to respiratory diseases related to increased in Yokkaichi’s polluted area since 1966, and a considerable difference in death rates was noticed between the polluted and non-polluted areas [21]. Petites and complaints by residents frequently happened during 1960–1970. Patients of Yokkaichi Asthma filed a joint lawsuit against the petrochemical industrial corporations and the local government. Efforts were made by the local government, the country, and individuals to reduce pollution [31]. On the national and municipal levels, one of the representative measures was the concept of “total emission control”. During late 1980–1990, Japan’s modern air pollution control laws, policies, and regulations were implemented. Petrochemical companies were forced to transform their production mode and reduce the emission of pollutants, yet the SO₂ pollution kept affecting Yokkaichi despite all the measures taken. Petitions and complaints from residents also reached a plateau with a high frequency. Aside from all kinds of law amendments, policies, and regulations, in 1990, the establishment of organizations such as the International Center for Environmental Technology Transfer Center (henceforth: ICETT) was encouraged by Yokkaichi, Mie Prefecture, and the Japanese government for rooting in the historical experience of Yokkaichi Air Pollution and creating an environmentally friendly future [29].

2.2. Yokkaichi Abatement of SO₂

According to Tobler’s First Law, all attribute values on a geographic surface are related to each other, but closer values are more strongly related than distant values [32]. After combining the air pollution characteristics of both cities, SO₂ was found to be used as the main index to measure the reductions in the discharged pollutants. The industrial shipment value (ISV, Table S1) data of Yokkaichi were acquired directly from the Yokkaichi government’s Information Communication Technology Strategy Section [33]. The changes in the Yokkaichi population were cross-checked with the national census of Japan. Further, data on the annual average concentration (AAC, Table S1) of SO₂ in Yokkaichi and Isozu Region (IR, Table S1) were acquired from the Mie Prefecture Environment White Book, which was published annually during 1967–2019 [34]. The GDP data of Mie Prefecture were acquired from the official website of the Mie Prefecture government, and the energy consumption data of Mie Prefecture were acquired from the official website of the Agency for Natural Resources and Energy, Ministry of Economy, Trade, and Industry of Japan.

According to the SO₂ AAC of Yokkaichi since 1965 (Figure 2), during the 1960s and early 1970s, when industrial development was a major pillar supporting high economic growth in Japan, the SO₂ discharge from petrochemical wastes also reached its peak. Before 1962, SO₂ pollutant monitoring facilities did not exist in Yokkaichi. The SO₂ AAC in IR from one of the monitoring facilities, which was located at the center of the Yokkaichi industrial complex territory. Owing to the expanding Japanese economy and high-speed developmental policies, an energy shortage occurred during the 1950s–1960s. During 1960–1965, the proportion of crude oil in response to Japan’s energy consumption increased from 37.7% to 58.4% and reached 75% in 1975. Further, with governmental support, the first, second, and third Yokkaichi industrial complexes were rapidly developed during 1959–1972. As shown in Figure 3, the ISV of the petrochemical industry began accounting for more than half of the city’s ISV shortly after 1960.

The hourly SO₂ concentration in IR in 1961 was measured and was found to reach a maximum level of 1.64 ppm, which was 10 times more than the permissible limits. Until 1964, the morbidity rate among nearly 3000 residents in IR was approximately 2.3%, which was particularly higher among the elderly and children. During almost the same period, the medical causation between factory-generated SO₂ emission and increasing asthma morbidity rate was affirmed. In recent decades, Yokkaichi’s dominant industry has changed from the petrochemical industries to the high-technology electronic component industry. After the 1980s, the industrial structure was adjusted. Particularly, the mechanical industry started developing, while the proportion of the petrochemical industry started decreasing continuously. Until the 2010s, the electronic component-manufacturing industry ranked first among all of Yokkaichi’s industries. According to the changes in Yokkaichi’s annual ISV data (Figure 4), compared with the past Yokkaichi air pollution period, the annual ISV of the industry was present at least three times more. Further, according to the changes in Yokkaichi’s GDP structure during 2006–2016 (Figure 4), secondary industries were the main industries until 2015. With the change of dominant industry, complaint cases from Yokkaichi residents also descended accordingly (Figure 5).Currently, tertiary industries are dominating Yokkaichi’s GDP. Therefore, Yokkaichi completed its industrial structure from the high pollution possibility type to the low pollution possibility type.

Previous empirical studies on Yokkaichi air pollution have focused on causation analysis between asthma and SO₂ pollution, mortality and SO₂ pollution, epidemiology with pollution [19,37,38,39], and the chronological study on Yokkaichi air pollution usually used Yokkaichi Air Pollution joint lawsuit to divide the entire process into four periods, and this division method considered both citizens’ awareness and governmental measures [21,29,30,31,37]. Moreover, this method has been proven to be rational and useful for decades. However, our pattern summarized from the Yokkaichi Air Pollution process divided into four new stages as follows:

Stage 1, the Early Stage, occurred after World War II until 1967. In this stage, the demand for economic development was high. Therefore, both local government and citizens encouraged the introduction of the industrial complex in Yokkaichi in exchange for economic growth [21,30,31]. Further, the first and second Yokkaichi industrial complexes were established and started operation in this stage. Subsequently, many petrochemical factories were assembled because of the establishment of the first and second complexes. Although the data before 1965 are lacking, it is still reasonable to speculate that SO₂ emissions increased rapidly in this stage. Acid rain was observed in Yokkaichi, and the pH value of the rain decreased from above 6.0 to nearly 4.0 during 1961–1967 in the earlier empirical research on Yokkaichi asthma [21]. This proved that during Stage 1, the SO₂ pollution continued to deteriorate. Since industrial development was not as rapid as assumed, local citizens started observing the environmental changes and started questioning the impact of the industrial complexes. This fact was confirmed by the number of petitions against the factories in Shiohama. However, the impact of pollution was relatively small at this stage.

Stage 2 occurred during 1968–1972. Although this stage comprised a relatively shorter period than the other stages, the third Yokkaichi industrial complex started its operation, showed higher industrial growth, and exhibited a strong fluctuation in the SO₂ AAC. Notably, after Stage 1, citizens’ awareness regarding air pollution also improved. The Yokkaichi Air Pollution joint lawsuit and its final judgment verified the improvement in public awareness [30]. Moreover, the impact of pollution was larger in Stage 2 than in Stage 1.

Stage 3 occurred during 1972–1980. Industries developed rapidly in this stage. Additionally, owing to the efforts undertaken to abate air pollution and the public’s increased awareness after the Yokkaichi Air Pollution final judgment, SO₂ concentrations decreased slowly yet stably in this stage. Moreover, the impact of pollution was continually increasing from Stage 1 to Stage 3, with the maximum impact observed in Stage 3 and photochemical smoke frequently appearing in Yokkaichi.

After 1980, Yokkaichi finally entered Stage 4, the after-pollution stage. In this stage, the simulation that the Yokkaichi industrial complex brought to Yokkaichi’s industry started fading. Economic growth, which relied on the complexes, slowed down. Evident changes in Yokkaichi’s industrial structure were not observed in this stage, and the industrial development was stable. Similar trends were observed for the SO₂ AAC. After entering the last stage, air pollution in Yokkaichi was under control. Moreover, the impact of pollution started systematically summarizing and learning from previous pollution experiences. The gradually increasing old industrial structure motivated the optimization of the industrial structure. Accordingly, Yokkaichi’s industrial structure was transformed in this stage. Additionally, the petrochemical industry was no longer the dominant industry, and Yokkaichi’s industrial structure reached a new, stable level. Therefore, a pattern can be summarized based on the industrial development and SO₂ concentration of Yokkaichi, as shown in Figure 6.

2.3. SO₂ Pollution in Tianjin

The air pollution issues of Tianjin usually were discussed corresponding to Beijing and Hebei Province as part of the BTH Regional Integration Strategy. Compared with the studies of related policies in Beijing, benefits in Tianjin were neglected. Moreover, the studies on PM_2.5 almost predominate air pollution studies in China. Most data on Tianjin were collected from the Yearbook of Tianjin by the Statistics Bureau of Tianjin. Tianjin Statistical Yearbooks were published by the China Statistics Press, and the data were compiled by the Tianjin Municipal Bureau of Statistics and the Survey Office of the National Bureau of Statistics in Tianjin [40]. In both the 2019 and 2021 Tianjin Statistical Yearbooks, although the industry proportion data for 2000–2018 were recorded, contrasting findings were observed in both years. Accordingly, we adopted the data before 2018 from the 2019 Yearbook and the 2019 data from the 2021 Yearbook. Further, the monthly SO₂ concentrations in Tianjin were acquired from the National Key Cities’ Air Quality Monthly Report by the Ministry of Ecology and Environment of China [41], and Tianjin’s annual SO₂ concentration data by districts were acquired from Tianjin Yearly Environment Statistics Report [42].

Tianjin (116 °43′–118°4′ E, 38°34′–40°15′ N) is located near Bohai Bay and has a terrain similar to that of Yokkaichi, as shown in Figure 2 and Figure 7. Geographically, Tianjin’s terrain hinders easy atmospheric flow in winter. Therefore, the diffusion of air pollutants in Tianjin evidently differs between summer and winter. After the formal establishment of the People’s Republic of China (“China”), Tianjin was designated as one of the few municipalities and heavy industry bases of China. Since the 1970s, to achieve rapid economic growth and sustain the national “Reform and Opening-up,” industrial companies, including petrochemical companies, were widely welcomed by the Chinese government. Similar to other Chinese cities, Tianjin’s desire to achieve rapid economic growth resulted in severe environmental problems. Based on the local crude oil production and Tianjin’s location as a harbor city, petrochemical industries developed rapidly in this region. Moreover, before the onset of PM_2.5 pollution, from the 1970s to 2000s, part of Tianjin belonged to the TCZ because of its large-scale SO₂ emissions and acid rain problem. Even presently, the petrochemical industry is one of Tianjin’s industrial pillars. Apart from the petrochemical industry, Tianjin, which belongs to the BTH region, also owns a large proportion of steel, cement, and glass industries. Moreover, a large amount of Tianjin’s annual SO₂ emissions originates from coal consumption for household heating in winter and from daily production in local factories (non-petrochemical industries). Although PM2.5 gained public attention in China, SO₂ continued to affect the environment and human health. In the recent decade, a new “Law of China on the Prevention and Control of Atmospheric Pollution” and joint prevention and control activities were implemented by the Chinese government to reduce the emission of air pollutants. Moreover, as an important city of the BTH region, Tianjin’s air pollution issue was related to the joint control policies and reforms that have been implemented on a large-scale in the household energy structure (Table S3).

SO₂ emissions were monitored from the 1980s. Presently, there are 27 air pollutant monitoring stations in Tianjin. Before 2016, the annual total emission of SO₂ in Tianjin ranged from 150,000 tons to 300,000 tons per year (Figure 8). Since 2004, during which the AAC was published for the first time, SO₂ emissions have shown a decreasing trend in both the total values and concentration. Moreover, among the three SO₂ emission sources in Tianjin, the annual emissions from industries were dominant. Urban residential sources mainly included emissions arising from energy consumption related to household heating. Tianjin is also witnessing industrial structure adjustment. With measures such as tax benefits, the local government stimulates the development of high-technology companies. Furthermore, only in the second half of 2017, approximately 21,000 high-pollution small and medium-sized enterprises in Tianjin were ordered to shut down or clean up. These measures helped the high-technology manufacturing industry in achieving 4.4% growth in 2018, in addition to the continuous decrease in pollution. However, Tianjin still has traditional heavy industry, and the path to industrial reform has not been stable (Figure 9 and Figure 10). The structure of total profits (pre-tax) of main economic indicators of competitive industries above the designated size of Tianjin has not changed greatly during 2012–2019, with the petrochemical industry still dominating almost half of the industrial profits. Based on the SO₂ AAC of Tianjin, the total SO₂ emission changes evidently declined in the recent 20 years. In fact, the administrative authority did not issue a reasonable penalty. However, civil law actions appeared more frequently by year (Figure 11). Although the instability in the administrative punishment implied that the enforcement of laws and regulations related to air pollution was not systematized, compared with the previous “all-blank” situation, the increase in the number of punishments could be observed based on the increased environmental awareness among citizens of Tianjin.

The key features of the pattern in the Yokkaichi Air Pollution process are industrial development speed, change in the SO₂ values, and citizens’ awareness (Tables S4–S7). Before adopting this method in Tianjin, Tianjin’s industrial development and SO₂ AAC should be demonstrated again. According to Tianjin’s annual emission statistics (Figure 8) for 1978–2015, the industrial development process in Tianjin was almost consistent with the key features of Yokkaichi in Stages 1 and 2. Stage 1, the early stage in Tianjin, could be from 1978 to 2008. In three decades, Tianjin’s industrial growth maintained a slow speed. By 2008, its total industry value was only three times more than that in 1978. Although data on the changes in SO₂ in this stage were lacking, a peak SO₂ value for Tianjin observed before 2004 could be inferred. Further, although the pollution condition was serious, residents were not sufficiently aware of air pollution. Therefore, the impact of pollution in Tianjin at this stage was also relatively small. Subsequently, during 2008–2015, Tianjin was in Stage 2, during which it experienced relatively fast industrial development. After the 2008 Beijing Olympics, awareness about air quality increased among the citizens. This was confirmed by the proportion of civil actions. In this stage, SO₂ concentration fluctuated, but the total concentration continued to decrease. Since 2016, the SO₂ AAC has started decreasing gradually. The total industry value data from 2016 to 2018 showed no signs of any sharp increase; additionally, citizens’ awareness about air pollution was higher in this stage than in Stage 2. According to the number of cases related to air pollution during 2016–2019, the number of civil actions was 20 times more in 2019 than in 2015. Currently, more measures have been undertaken to adjust the city’s industrial structure, which witnessed an evident increase in the high-technology and low-pollution industries. According to Tianjin’s Yearbook of 2018, high-technology manufacturing industries accounted for 13.3%, and new industries related to fields, such as information technology, biology, and high-technology equipment, accounted for approximately 21.8% of the entire city’s industry value. However, these phenomena cannot be used to verify the entry of Tianjin to Stage 4. We believe that the key factor to enter Stage 4 would be a drastic change in the industrial structure wherein the petrochemical industry would no longer be the pillar for this region’s development. Presently, traditional industries, such as oil, petrochemical, and metallurgical industries, still dominate Tianjin’s industrial structure (

Table 1. Energy consumption by 3 industries and living in Tianjin, 2015–2019 (Unit: SCE).

Year	Final Consumption	Primary	Secondary	Tertiary	Living
2015	8137.29	105.43	5713.81	1304.36	1013.69
2016	7875.03	110.18	5368.22	1336.63	1060
2017	7687.75	116.68	5101.45	1366.86	1102.76
2018	7917.81	107.81	5263.17	1355.01	1191.83
2019	8261.29	107	5538.91	1413.37	1202

and Figure 10). Particularly, the profit from oil and natural gas industries accounts for over 40% of the total industrial profits of Tianjin, implying that Tianjin’s industrial structure has not yet been completely adjusted, and thus, the conventional structure will continue affecting its economy. Based on the above findings, Tianjin can be considered in Stage 3 (Figure 12).

In conclusion, according to our four-stage division of Yokkaichi’s process, a specific pattern can be observed in Yokkaichi’s air pollution abatement process. Moreover, Yokkaichi’s pattern can be applied to Tianjin, as Tianjin is currently consistent with Stage 3 of Yokkaichi’s pattern.

3. Discussion

3.1. Data and Model

To obtain a better understanding of the four-stage pattern, it is rational to cross-check the data of both cities’ SO₂ concentration with industry value. SO₂ concentration stands for the regional SO₂ pollution, and we deem it appropriate that SO₂ pollution is considered an environmental impact here for coherence. To study the environmental impacts, we adopt the original IPAT equation, I = PAT, as the basic model. The concept of the IPAT model is Pollution = (Population) ×(Production/Capita) × (Pollution/Production), i.e., Pollution = Pollution [45]. It has been considered useful as a starting point for testing different drivers of environmental impacts. When discussing anthropogenic environmental change, three elements, population (P), affluence (A), and technology (T), are considered determinants [46], with the other drivers that can be included in the variation of the original equation. Moreover, the reformed IPAT equation—is demonstrated as follows:

$$I=a{P}^b{A}^c{T}^{d}e$$

(1)

In the application of the original IPAT equation, the scale of population, production per capita, and pollution per production are more flexible. To observe and understand the environmental impacts, in many cases, appeared as CO₂ emission, P can be population or the density of the population. A can be the GDP, GDP per capita, T as energy efficiency. Sometimes T can also be considered a residual term [46]. Parameters a, b, c, d, and e are, as a matter of fact, equal to 1 as a constant, which brings no extra influence on the equation. However, the reformed IPAT equation can be added to other possible variables as long as they are considered to correlate with the environmental impacts. In this paper, I is considered as an annual concentration of SO₂, P is the number of residents, and T is the efficiency of energy utilization. In the meantime, we consider that other factors could also be included in discussing the environmental impacts. Therefore, we need to introduce an extended stochastic equation from the IPAT equation, hereby, the extended STIRPAT model.

In this discussion, we used an extended STIRPAT model as a reference to check the flow extracted from the economic development with SO₂ abatement in Yokkaichi (

Table 2. Variable, symbol, and definition of Yokkaichi.

Variable	Symbol	Definition
Environmental impact	I(yc)	Annual average concentration of SO2
Population	P(yc)	Number of residents in Yokkaichi
Affluence	A(yc)	Industrial Shipment Value
Technology	T(yc)	Efficiency of energy utilization in Mie Prefecture *
Behavior	B(yc)	Case of complaints filed to municipal authorities by residents
Industrial Structure	S(yc)	Proportion of secondary industry by three main industries

* Efficiency of energy utilization stands for the GDP value produced per unit of energy consumption. Here, because the data from Yokkaichi are insufficient, we acquired the Mie Prefecture’s data instead.

). Meanwhile, we also checked Tianjin’s data for reference (

Table 3. Variable, symbol, and definition of Tianjin.

Variable	Symbol	Definition
Environmental impact	I(tj)	Annual average concentration of SO2
Population	P(tj)	Number of residents in Yokkaichi
Affluence	A(tj)	Industrial Product Value
Technology	T(tj)	Efficiency of energy utilization
Behavior	B (tj)	Case of punishment filed by the administrative authorities of Tianjin
Industrial Structure	S(tj)	Proportion of secondary industry by three main industries

). The extended STIRPAT model has provided a different perspective for the analysis of data. However, it is necessary to explain that in this model, both sides of the equation are not necessarily equivalent. Other than the P, A, T, we included the important factors that appeared above, anthropogenic behavior, as B, hereby the case of complaints filed to municipal authorities by residents in Yokkaichi or the case of punishments filed by the administrative authorities of Tianjin. Furthermore, we added S, industrial structure, which in both cities is the proportion of secondary industry by three main industries, as the adjusted and extended STIRPAT equation based on the realistic requirement of this study is as follows:

$$I=a{P}^b{A}^c{B}^d{S}^h{T}^{k}e$$

(2)

In general, population growth increases the burden on the environment; thus, Population negatively correlates with I. We decide the effect of the variable Affluence on I be negative mainly because we drop the income or GDP per capita as Affluence, but industrial product value, therefore more growth on industry causes more pressure on I. Higher efficiency in energy utilization helps reduce pollution, therefore, positively correlates with I. Moreover, Behavior, no matter government’s behavior of issuing punishment or residents’ behavior as filing complaints, even though they were caused by pollution, which is a high environmental impact, all can objectively cause more behavior on reducing pollution, hence, a negative correlation. S for industrial structure, with a higher proportion of the secondary industry, I will also be more intensive; therefore, S is positively correlated with I (

Table 4. Elaboration of positive “+”/negative “–”correlation between variables with I.

Symbol	+/−
I	N/A
P	+
A	+
T	−
B	−
S	+

). In this extended STIRPAT model, the parameters’ relations are a = b = c = d = h = k = e = 1, with e as a residual term. To make a clearer view of the data and the tendency, the logarithms of Equation (2) were considered as follows:

$$LnI=Lna+bLnP+cLnA+dLnB+kLnT+hLnS+Lne$$

(3)

The original data for Yokkaichi is shown in

Table 5. Source data for Yokkaichi on STIRPAT equation.

Year	I(yc) AAC of SO2 (ppm)	P(yc) Population (1 000)	A(yc) ISV (Million JPY)	T(yc) Technology (JPY/unit)	B(yc) Behavior (case)	S(yc) Industrial Structure *
2007	0.04	313.403	2.48368 × 108	0.04145	122	0.376
2008	0.03	314.805	2.68521 × 108	0.04396	113	0.36
2009	0.02	314.577	2.7044 × 108	0.0442	102	0.382
2010	0.02	314.393	2.23067 × 108	0.04611	72	0.41
2011	0.02	314.623	2.46814 × 108	0.04744	72	0.355
2012	0.02	313.915	2.61461 × 108	0.04613	68	0.461
2013	0.02	313.277	2.68495 × 108	0.0453	62	0.509
2014	0.02	312.857	3.08802 × 108	0.04564	60	0.521
2015	0.01	311.031	3.17992 × 108	0.04297	80	0.513
2016	0.01	310.674	3.35594 × 108	0.04122	31	0.469

* The industrial structure data of Yokkaichi is acquired from ICT Strategy Section of Yokkaichi municipal.

. After we took the logarithm of each variable, the data on Yokkaichi can be described in

Table 6. Logarithm calculation based on source data of Yokkaichi (Table S8, Figures S1–S6).

Year	Ln (I(yc))	Ln (P(yc))	Ln (A(yc))	Ln (T(yc))	Ln (B(yc))	Ln (S(yc))	Ln (ΣPATBS(yc))
2007	−5.52146	5.74749	19.40844	−3.18325	4.80402	−0.97817	3.25032
2008	−5.80914	5.75195	19.41556	−3.12459	4.72739	−1.02165	3.24838
2009	−6.21461	5.75123	19.22298	−3.11903	4.62497	−0.96233	3.23938
2010	−6.21461	5.75064	19.32414	−3.0768	4.27667	−0.8916	3.23408
2011	−6.21461	5.75138	19.38179	−3.04834	4.27667	−1.03564	3.23183
2012	−6.21461	5.74912	19.40834	−3.07635	4.21951	−0.77436	3.23971
2013	−6.21461	5.74709	19.54821	−3.09448	4.12713	−0.67531	3.24465
2014	−6.21461	5.74575	19.57754	−3.08703	4.09434	−0.65201	3.24566
2015	−6.90776	5.73989	19.63141	−3.14718	4.38203	−0.66748	3.25573
2016	−6.90776	5.73874	19.36595	−3.18885	3.43399	−0.75715	3.20245

and

Table 7. Coefficient Analysis of Ln(I) with Ln (P), Ln (A), Ln (T), Ln (B), and Ln (S) of Yokkaichi in time sequence.

Variable	Coefficient (Pearson’s r)	R2	Adj. R2
Ln (P(yc))	0.7347	0.53978	0.48225
Ln (A(yc))	−0.24813	0.06157	−0.05574
Ln (T(yc))	0.12589	0.01585	−0.10717
Ln (B(yc))	0.71771	0.51511	0.4545
Ln (S(yc))	−0.59696	0.35636	0.27591

. Since we did not expect the left side and right side of the equation to be equivalent, and the purpose for adopting the models is to describe the yearly changes between drivers and the annual concentration of SO₂, in the logarithm calculation process, we also enumerated the changes on time of each variable, as LnP(yc), LnA(yc), LnT(yc), LnB(yc), and LnS(yc). Moreover, the most important, the logarithm of “$a{P}^b{A}^c{B}^d{S}^h{T}^{k}e$” showed as Ln(Σ_PATBS(yc)).

A simple linear fitting analysis is given in Figure 13 to compare numerical Ln(I(yc)) value with calculated Ln (Σ_PATBS(yc)). And the whole span of Yokkaichi’s Ln(I(yc)) is as showed in Figure 14.

Here we apply the same method in Tianjin in adding comparison. As shown in Table 3 above, we chose the case of punishment issued by the administrative authorities of Tianjin to emphasize the regional change rather than a change on a national scale. Moreover, the data on the industrial structure are acquired from Tianjin’s Statistics Yearbook. The five-year consistent data of Tianjin was demonstrated as in

Table 8. Source data for Tianjin.

Year	I(tj) AAC of SO2 (μg/m3)	P(tj) Population (10 000)	A(tj) IPV (Million CNY)	T(tj) Technology	B(tj) Behavior (Case)	S(tj) Industrial Structure
2014	49	1516.81	727,168	0.00748	97	0.47
2015	29	1546.95	719,654	0.00748	62	0.424
2016	21	1562.12	680,513	0.06861	114	0.409
2017	16	1556.87	686,398	0.00617	178	0.409
2018	12	1559.6	696,271	0.00593	41	0.405

, the logarithm calculation of I, P, A, T, B, S and “$a{P}^b{A}^c{B}^d{S}^h{T}^{k}e$” in

Table 9. Logarithm calculation based on 5-year consistent source data of Tianjin (Table S9, Figure 7, Figure 8, Figure 9, Figure 10, Figure 11 and Figure 12).

Year	Ln (I(tj))	Ln(P(tj))	Ln(A(tj))	Ln(T(tj))	Ln(B(tj))	Ln(S(tj))	Ln (ΣPATBS(tj))
2014	3.89182	7.32436	13.49691	−4.89605	4.57471	−0.75502	2.9829
2015	3.3673	7.34404	13.48653	−4.89559	4.12713	−0.85802	2.95512
2016	3.04452	7.3538	13.4306	−2.67925	4.7362	−0.89404	3.08864
2017	2.77259	7.35043	13.43921	−5.08731	5.18178	−0.89404	2.99524
2018	2.48491	7.35218	13.45349	−5.12854	3.71357	−0.90387	2.91706

with its data description and linear fitting analysis (

Table 10. Coefficient Analysis of Ln(I) with Ln(P), Ln(A), Ln(T), Ln(B), and Ln(S) of Tianjin in time sequence.

Variable	Coefficient (Pearson’s r)	R2	Adj. R2
Ln (P(tj))	−0.88348	0.78054	0.70739
Ln (A(tj))	0.7669	0.58814	0.45085
Ln (T(tj))	0.02674	0.715266	−0.33238
Ln (B(tj))	0.16967	0.02879	−0.29495
Ln (S(tj))	0.91957	0.84561	0.79415

and Figure 15), and the linear fitting of whole span (since 2004) of Tianjin’s Ln(I(tj)) (Figure 16) are as follows:

3.2. Discussion on Model Utilization and Pattern Utilization

In the sequence of two data groups of Yokkaichi and Tianjin, based on the comprehensive data of the whole process, it was found that the stability of Yokkaichi’s process of pollution abatement was stronger than that of Tianjin. We have elaborated on how the SO₂ concentration of Yokkaichi has descended to a low and stable level in recent decades. Even any small change in its annual concentration will cause a comparatively huge fluctuation. However, if we analyzed the whole span (Yokkaichi since 1967, as shown in Figure 14, Tianjin since 2004, Figure 16), the environmental impacts of Yokkaichi were at their Stage 4 and appeared to be more cohesive with its right-side SUM value from equation (3), which also covered our hypothesis above that Yokkaichi has entered the Stage 4, after pollution stage, and whereas Tianjin is still in Stage 3. Here, we can also interpret the result from the comparison through the extended STIRPAT model, which was more applicable in Stage 3 of Tianjin than in Stage 4 of Yokkaichi. On observing the data of Yokkaichi, we considered that the deviation was caused by the long-term low-level SO₂ after entering Stage 4.

Hitherto, there were no similar patterns for a city-level comparison in the field of air pollution studies. This study tried to summarize Yokkaichi’s air pollution abatement process and to prove a realistic and viable model exists in it. There is an established connection between the SO₂ concentration of petrochemical industrial cities and industrial development, which is based on the target city’s petrochemical industry ratio, SO₂ emission and concentration, economic development, and residents’ awareness about environmental protection. We only consider this study as a start, as it is just an attempt at something new, and still crude and with many flaws. Due to the long span of time, the available data for both Yokkaichi and Tianjin was scarce. All these prerequisites restricted this four-stage division method to the Yokkaichi and Tianjin in this study’s entire time span. There are still many uncertainties when discussing the future development of Tianjin because we cannot be sure that it will proceed similarly to Yokkaichi. If we must give Tianjin some suggestions from Yokkaichi’s pattern to help speed up its transition from stage 3 to stage 4, we would propose a tighter regional SO₂ emission control regulation and a more intensive residents’ involvement in environmental issues. Further research may make it possible to find other petrochemical industrial cities around the world that were suitable for this method as well. Even if methods of low carbon emissions and low air pollution were adopted by all the members of the United Nations and all kinds of reproducible and clean energy were promoted for research worldwide, it would still take decades for clean energy to replace fossil fuels completely. As long as the petrochemical industry influence human activities and daily life, we hope the experience we draw from the air pollution abatement process in Yokkaichi can become useful in the future for cities having petrochemical industries.

Since 2020, soon after the global COVID-19 pandemic, great recessions have hit all countries, trades, and professions, including the environment. After the official end of the pandemic, will the possible redemption cause a loosening in the supervision of pollution control? Will the information on the present energy structure also be slowed down outside the developed countries? We are currently standing at the final 10 years of the United Nations’ plan for achieving sustainable development goals (SDG), and not much time is left for us to meet the targets. It also means that to fulfill the missions, every country, organization, and individual needs to make an extra effort. In this study, we focused on the data before the pandemic and assumed that the pandemic may have given some change to the pattern acquired from the past. However, under these complicated circumstances, we still want to believe that the retrospective study on air pollution in Yokkaichi would be useful for other Asian cities and able to maintain its unique and important value. After entering the 21st century, Yokkaichi has been dedicated to creating a Pollution-Environment-SDGs-Carbon-neutral society based on its experience from the past [47]. Even though there are differences that cannot be neglected between Yokkaichi and Tianjin in population, industrial scales, culture, and so on, we reckon that if the experience form Yokkaichi air pollution can be flexibly adopted, it could still be proper for Tianjin to speed up its process to enter the after-pollution stage. At the same time, by reviewing multiple pollution-reducing methods regarding air pollution in Yokkaichi and Tianjin and comparing the results of enforcing environmental policies in both countries, we hope to have a better understanding of Tianjin’s current situation of improvement and a clearer view of how and where this process may lead.