As an important research method, meta-analysis refers to the systematic and comprehensive statistical analysis of the results of multiple experiments with the same research purpose and independent of each other (Brander et al., 2012; Chaikumbung et al., 2016). It was originally proposed as a research synthesis method by Glass (1976). Existing meta-analyses mainly use the induction, measurements, and conclusions of a research field, cases (Chen et al., 2019), and indicators to identify scientific consensus. Meta-analysis has been used in various fields, such as medicine, pedagogy, and psychology, and has gradually been applied to geography (van Zanten et al., 2014; Eötvös et al., 2018). Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) is an evidence-based minimum set of items for reporting in systematic reviews and meta-analyses (Moher et al., 2009). As a basic data source for China’s brownfield spatial distributions, this study applies the PRISMA approach to collect brownfield case information from published literature and industrial heritage directories.

The literature review procedures are outlined in Figure 1. A literature search was performed following guidelines from PRISMA. Publications in both Chinese and English were collected by searching two major scientific citation indexing services, the China National Knowledge Infrastructure (CNKI) and Web of Science (WOS), respectively. Chinese brownfield-related literature is mainly published on CNKI. WOS is the most authoritative scientific and technological literature retrieval platform. The earliest Chinese literature on the theme of brownfields appeared in 2001 (Niu, 2001). Therefore, publication dates were limited to “2001.1.1–2019.12.31.” We used the topics “brownfield” AND “China” with WOS. Since China has no official definition of brownfields, the following search terms were use in the CNKI: “brownfield,” “industrial heritage AND transformation,” “industrial heritage AND renovate,” “industrial heritage AND regeneration,” “abandoned land AND transformation,” “abandoned land AND renovate,” “abandoned land AND regeneration” in Chinese. We searched publications ranging from journal articles to conference proceedings or degree theses. We also searched the official list of China’s industrial heritage sites and obtained 307 records.

A total of 1924 abstracts were retrieved for screening, 1,638 of which were from CNKI (in Chinese) and 286 from WOS (in English). 1) Abstracts that were not relevant to the brownfield transformation were excluded. This led to 1,217 Chinese and 32 English papers being selected for full-text review and further geo-information extraction. 2) Of the available full texts, publications that did not refer to China’s brownfield cases, lacked brownfield data, or duplicate brownfield sites were deleted, leaving 455 Chinese and 16 English publications eligible for extraction. 3) A total of 196 industrial heritage records were obtained by deleting records that did not meet brownfield requirements from 307 industrial heritage lists.

Location information was extracted from the records of the relevant primary papers. Following Zhang et al. (2019), the longitude and latitude of a location were determined using a combination of geospatial tools, including the Baidu Map, and Baidu Coordinate Picking System. We updated the place names to match the historical administrative names. We further classified all locations into four different levels according to their geographic scales and administrative levels (i.e., provincial, prefectural, county, and township and finer level). This helps potential users of this dataset extract appropriate sections for use. The locations of brownfields were then visualized using geographic information systems (GIS) software. In total, 608 Chinese, 18 English, and 196 industrial heritage records were identified.

After the data were entered, a second person thoroughly checked the dataset for validity, to avoid errors and duplication. A total of 816 records were identified, 605 of which were from CNKI, 17 from WOS, and 194 from China’s industrial heritage list. Brownfield records include 23 provinces, 255 prefecture-level counties/cities, three autonomous regions, four municipalities directly under the central government, and two special administrative regions. The Xizang Autonomous Region and Ningxia Hui Autonomous Region did not have any brownfield sites.

In the dataset of distribution of brownfields in China, each row represents a single record. The columns in the dataset are as follows:

1) Brownfield location: the geographic scale of location (provincial, prefectural, county, township, or finer level)

Based on the results, brownfields are mainly concentrated in economically developed areas and near central urban areas. As shown in Figure 2A, the overall brownfield spatial distribution in China was basically consistent with the “Heihe-Tengchong Line” as the boundary, exhibiting a spatial pattern of “East-dense-West-sparse.” There were 791 brownfield sites in the Southeast region of the “Heihe-Tengchong Line,” accounting for 96.94% of the total, and 25 in the northwest region, accounting for 3.06% of the total. The “Heihe-Tengchong Line” is an important geographical boundary that reflects China’s urbanization and economic development levels (Qi et al., 2015; Li et al., 2017). As shown in Figure 2B, the economically developed provinces of Jiangsu, Shandong, Guangdong, and Zhejiang have the largest number of brownfields, followed by resource-dependent regions such as Liaoning, Henan, Hebei, Hubei, and Hunan. The southern provinces have more brownfields than the North, followed by the northeastern provinces, with the lowest number in the Midwest. Brownfields have emerged as a result of the ongoing interaction between industrialization and urbanization, and their distribution is closely related to China’s urbanization and economic development. We used the NNI module in ArcGIS to analyze the average nearest neighbor distance of each brownfield site. According to Table 1, the NNI for the overall spatial distribution of the 816 brownfield sites identified in the study was 0.4, and the Z score was −32.90, which passed the 1% significance level test, indicating that the spatial agglomeration characteristics of brownfields in China are highly significant. As shown in Figure 2C, the KDE method was used to analyze the spatial distribution. According to the natural break method, the kernel density values were divided into seven levels, from high to low. Regions with the highest kernel density values “7.27–18.35” were determined to be high-density areas, values from “1.58–7.27” represented medium-density areas, and those from “0–1.58” represented low-density areas. China’s brownfields were characterized by multi-center agglomerations, and high-density areas were found in the Yangtze River Delta, Beijing–Tianjin–Hebei, Pearl River Delta, and Central Plains urban agglomerations, as well as the urban agglomeration in the middle reaches of the Changjiang River with a strong industrial foundation and developed economy. Shanghai, Tianjin, Beijing, Guangzhou, Nanjing, and Wuhan had brownfield density clusters.

According to China’s Code for Classification of Urban Land Use and Planning Standards of Development Land GB50137-2011, brownfield land use functions are classified as industrial, mining, military, commercial activities, storage, public facilities, and transportation facilities land. The brownfield information dataset divided brownfields into six types: industrial, mining, transportation facilities, public facilities, military, and logistics and warehouse (Alker et al., 2000; Franz et al., 2006). Industrial brownfields accounted for the highest number, followed by mining brownfields. Transportation facilities, public facilities, military, and logistics and warehouse brownfields were less common.

NNI was used to test the agglomeration characteristics of the six types of brownfield spatial distribution (Table 1). The NNI of the six types of brownfields was less than 1 in all cases, which passed the 1% significance test and formed a part of a significant agglomeration model. The degree of agglomeration is in the following order: industrial brownfields > transportation facilities brownfields > logistics and warehouse brownfields > public facilities brownfields > military brownfields > mining brownfields.

Although the NNI reflects the overall distribution of brownfields in China, it cannot intuitively reflect the degree of agglomeration. Considering this, KDE was used to further measure six types of brownfield spatial agglomeration characteristics. As shown in Figure 3A, industrial brownfields were mainly concentrated in city clusters with strong industrial foundations, such as the Yangtze River Delta, Pearl River Delta, and Beijing-Tianjin-Hebei urban agglomerations, as well as other industrially developed provinces like Taiwan, Jiangsu, Guangdong, and Zhejiang. Industrial brownfields are widely distributed, regionally concentrated, and are characterized by a large abandoned area. Because deindustrialization and urbanization were higher than in other regions, they were found to be concentrated in the Eastern region. As shown in Figure 3B, high-density concentration areas of mining brownfields appeared in the Central Plains urban agglomerations, the Shandong Peninsula’s urban agglomerations, the urban agglomeration in the middle reaches of the Changjiang River, and the Southern industrial zone. Cities with a large number of mining brownfields included Datong, Jiaozuo, Huangshi, Tongchuan, Xuzhou, Fushun, and other resource-based cities. As shown in Figure 3C, transportation facilities brownfields were concentrated in coastal cities along the Yellow River, Yangtze River, Bohai Bay, and transportation hub cities, such as Zhengzhou, Shanghai, Wuhan, and Tianjin. They mainly include the Shanghai Railway Station, Shanghai-Nanjing Railway, Shanghai Whampoa Dock, Lanzhou-Lianyungang Railway extension, and the Beijing-Guangzhou Railway. As shown in Figure 3D and Figure 3E, public facilities and logistics and warehouse brownfields belong to basic service land with a low degree of abandonment, so the quantity is relatively small. Jiangsu and Shanghai have more public facilities brownfields, and Taiwan and Shanghai have more logistics and warehouse brownfields. As shown in Figure 3F, military brownfields accounted for a small proportion and had a scattered distribution in Sichuan, Chongqing, Shanxi, Gansu, and other remote mountainous areas that are far from cities.

Industrial brownfields are more concentrated in industrially developed cities, mining brownfields are concentrated in areas with rich mineral resources, transportation facilities brownfields are concentrated in cities along the river and coastal and transportation hubs, public facilities brownfields and logistics and warehouse brownfields have no obvious agglomeration rule, and military brownfields are concentrated far from central cities.

Among the 471 basic documents searched, 252 records from 219 articles summarized the formation factors of brownfields. The frequency counting method was used to count the formation factors mentioned in the articles’ records. The factors can be classified into four categories: industrial structure adjustment, resource depletion, accelerated urbanization, and national policy orientation. Among the four types, industrial structure adjustment exhibits the highest frequency, followed by resource depletion, accelerated urbanization, and national policy orientation (see Table 2). Therefore, these four factors collaborate to produce brownfields, and their effects can be summarized as follows: industrial structure adjustment > resource depletion > urbanization process accelerated > national policy guidance.

To further demonstrate the correlation between brownfield spatial distributions and the factors that influence their formation, the distribution results were compared with the proportion of tertiary industries in China’s provinces (Long et al., 2012), 262 resource-dependent cities (GOSCPRC. General Office of the State Council of the People’s Republic of China, 2013; Feng et al., 2019), the level of urbanization in each province (Song et al., 2019), and the “Third-Front Movement” regions (Liu et al., 2014; Li H. C et al., 2019). Figure 4A compares brownfield locations with the proportion of tertiary industries in each province in 2018. Jiangsu, Guangdong, Zhejiang, Shandong, and Taiwan with a high proportion of tertiary industry also have a large number of brownfields. Figure 4B compares the brownfield locations of 262 resource-dependent cities. Mining brownfields were mostly located in resource-dependent cities, such as Huangshi, Tangshan, Jiaozuo, Fushun, Xuzhou, and Datong. Figure 4C compares those with provincial urbanization levels in 2018. Provinces with higher urbanization levels also had more brownfields. Figure 4D compares brownfield locations with the “Third-Front Movement” regions. Most military brownfields were concentrated in provinces with the “Third-Front Movement.”

The production of brownfields in China is the result of a combination of many factors. A further comparison of formation factors and types revealed that the adjustment of the industrial structure and the acceleration of urbanization are common factors, namely, those of industrial and mining, public facilities, transportation facilities, and logistics and storage land. The depletion of resources has tended to aggravate the formation of mining brownfields, while the “Third-Front Movement” policy has mainly promoted the formation of military brownfields.

The adjustment of China’s industrial structure has constituted the dominant force in its economic growth over the past 40 years of reform and opening-up. With the acceleration of industrialization and the promotion of the policy of suppressing secondary industry and developing the tertiary sector, the latter occupies a dominant position in the industrial structure. The original urban land structure and industrial patterns are becoming increasingly less adaptable to the needs of China’s economic development (Lai et al., 2020). To adapt to this change, enterprises have closed down or relocated, and leftover industrial and mining industries, logistics and storage industries, and transportation facilities have been abandoned, forming brownfields.

With the acceleration of urbanization and the expansion of urban areas, industrial, mining, public facilities, transportation facilities, and logistics and warehouse enterprises that originally belonged to the suburbs have become part of the city, forcing original industrial enterprises to relocate. This led to the formation of brownfields in China.

Resource depletion factors are key drivers behind the formation of mining brownfields. Due to the exhaustion of resources, the single industrial structure, and serious environmental pollution, the resource industry has shrunk, and industrial efficiency has declined. Resource-dependent industries face transformations and closures, resulting in more mining brownfields.

The “Third-Front Movement” was a major strategic decision made by President Mao Zedong in the mid-1960s. It was intended to strengthen wartime preparations under an increasingly tense international situation as a strategic shift from East to West to gradually change the distribution of China’s productive forces. The emphasis was on construction in the Southwest and Northwest of the country (Naughton, 1988). Because most of the “Third-Front Movement” projects were “backed by the mountain, scattered, hidden,” with the economic recession, they had no advantageous location and difficulty in terms of access led to bankruptcy or relocation, forming military brownfields.

The results of this study revealed that there were more brownfields in the South than North, Northeast, and Midwest. Possible reasons for this are as follows. First, the highest percentage of brownfields in southern provinces was related to the high level of local economic development. There was a greater degree of openness and rapid economic development in the South, coupled with high urbanization, dense population, urban land expansion, and high demand for land, which caused more attention to be paid to abandoned brownfield sites. Second, despite the gap in economic development between the South and North, the transformation of industrial structure and resource-dependent cities also produced a large number of brownfields, especially old industrial and resource-dependent cities in the North (Xie et al., 2022). Third, as the earliest developed old industrial base in the country, the Northeast should have more brownfields, but there were only 83 records in this region. This may be because the regional economy’s relatively slow growth, transformation of its industrial structure, and large stock land resulted in a low demand for abandoned brownfield redevelopment. In addition, local scholars have done relatively little research on brownfields and the limited published documents may have led to a slight discrepancy between the recorded results and the actual situation in the Northeast. Fourth, there were fewer brownfield records in the Midwest provinces, mainly due to the small local population, relatively weak industrial foundation, and insufficient attention to brownfield research.

Compared to previous studies, brownfields in China are mainly concentrated in economically developed areas, near central urban areas, and along transportation routes such as railway hubs. This result is consistent with those of Lange and McNeil (2004), Longo and Campbell (2017), and Frantál et al. (2015), who analyzed the spatial distribution pattern of brownfields in the United States, the United Kingdom, and the Czech Republic, respectively. The brownfield spatial distributions of China were consistent with the results of the Survey Bulletin on Soil Pollution in China released by the State Council in 2014. The report highlighted that soil pollution is more serious in the South than in the North, and soil pollution problems in some regions, such as the Yangtze River Delta and the Pearl River Delta, are more prominent (MEP and MLR, 2014). The spatial distribution of soil pollution is consistent with the results of the current study. Aoki et al. (2018) established a database of known Chinese industrial heritage, and found that its distribution decreased from the Southeast coast to the West, mainly concentrated in the Yangtze River Delta Urban Agglomeration, Beijing-Tianjin-Hebei Urban Agglomeration, and Pearl River Delta Urban Agglomeration. In China, most industrial brownfields are rated industrial heritage due to their own value. The agglomeration results of industrial heritage were consistent with the concentrated industrial brownfield distribution pattern in this study. This study is not only consistent with the spatial distribution of soil pollution and industrial heritage; it is the first to describe the spatial distributions of brownfields in China. The methods and results lend credibility and practical significance to data and spatial distributions of brownfields in China and offer an effective empirical meta-analysis for brownfield studies.

China’s urban development is gradually entering the adjustment and transformation period from the stage of rapid expansion, and traditional urban development based on incremental planning is gradually shifting toward inventory planning. Urban land optimization and improvement are important issues. This study proposes the following policy recommendations:

First, a database for China’s brownfields should be established to enhance the openness and transparency of brownfield information. The availability of data is a major issue in China. It is necessary to construct a comprehensive database that includes the size, complexity, original use, ownership relations, and planned/allowed use of a site. To assist various levels of government, developers, research scholars, the public, and other groups of people in obtaining brownfield information, the government should establish a dedicated website for open brownfield databases.

Second, at the macro-level, it is necessary to formulate a targeted regional brownfield development direction strategy based on regional characteristics and brownfield types. The Eastern coastal areas, especially the Yangtze River Delta, Pearl River Delta, and Beijing-Tianjin-Hebei urban agglomerations, have more industrial brownfields, and there have been more successful transformation cases. In the future, the government should increase policies and preferential measures for brownfield redevelopment in these areas, and more attention should be paid to brownfield transformation models and later project supervision. Northern areas, especially Liaoning, Shanxi, Shandong, Henan, and other resource-dependent provinces, have more mining brownfields and are relatively far from urban areas. Brownfield reconstruction projects should be based on ecological restoration and prioritize the development of an ecological park model. A considerable amount of brownfield land remains in transition as an old industrial base in Northeast China. In the revitalization of the Northeast, emphasis should be placed on identifying and investigating existing industrial wastelands, encouraging investment in brownfield redevelopment, and academic research. There were relatively few brownfields in the Midwest. Attention should therefore be paid to their identification and restoration in this area. The Midwest is both densely and sparsely populated, and there is less demand for brownfield redevelopment. Restored brownfields can be used as reserve lands and will play a role in the new pattern of western development in the future. As different types of brownfields vary in terms of their location, area, nature, pollution degree, heritage value, and redevelopment mode, the government should formulate guidelines for the redevelopment of different types of brownfields to guide practice.

Third, local governments should coordinate and prioritize brownfield redevelopment. The Chinese government controls the primary market for land and should be the first promoter of brownfield reuse. Brownfield locations, pollution degree, and stock land should be combined to evaluate the reuse potential and availability of brownfields, redevelopment priorities, and transformation directions, and identify brownfield resources to return them to the market.