Chinese sociotechnical imaginaries of Earth observation: From sight to foresight

Abstract

Although Earth observation (EO) is considered a universal scientific technique with a hegemonic gaze, national sociotechnical imaginaries shape its practice. Historically established and organized by national governments, EO, which is commonly pursued via satellite remote sensing, depends on domestic perceptions of the technique's scope and applications. EO is also a political technology with the capacity to mediate and potentially regulate life on Earth at a range of scales. In China, the state and scientists endeavor to transform EO into a data-driven, inductive, and predictive method with a focus on social coordination, especially in cities—a key scale for scientific study and the organization and reproduction of state power. Contributing to China's reimagining of EO as a governance and governmentality tool are advances in social sensing, which leverages big data continuously generated by individuals’ devices. By analyzing big EO and social data, rather than only observe past and present changes to the Earth's surface, the Chinese state and scientists seek to predict future social events. While these changes are narrated as a popularization of remote sensing enabling more responsive, human-centric governance, they may augur the rise of antipolitical technologies that attempt to prevent dissent by determining and even guiding future behavior. As China, a satellite power, aims to export its services for observing and predicting Earth and upscale them through global governance, critique of their associated sociotechnical imaginaries is crucial. Such research reveals the heterogeneity of the satellite gaze and may identify where and how people hold power within EO systems.

Introduction: Ground control to Mao Zedong

In 1970, a radio transmitter aboard the first satellite of the People's Republic of China (PRC), Dong Fang Hong 1 (东方红一号, “The East is Red”), broadcast the de facto national anthem—a paean to the country's communist leader—down to Earth. Seventy years later, as one of the world's leading satellite powers, the Chinese state is capturing electromagnetic radiation reflected from the planet to support increasingly digitized governance. Building on the pioneering satellite's motto of “Go up, catch, hear, see” (上得去，抓得住，听得到，看得见), the state perceives space as a vantage point for not only sensing human–environmental processes, but intervening in them, too.

Many countries have an interest in Earth observation (EO)—gathering data about the planet using remote sensing instruments like satellites, aircraft, and drones—to manage the global environment. EO produces synoptic and consistent data at regular time intervals that can be input into complex models. China, however, additionally seeks to use EO data to orchestrate social dynamics, placing a strong faith in its evidentiary weight. Already in 2002, a Chinese district court used satellite imagery to convict an individual of deforestation (Chinanews.com, 2002). In contrast, internationally, while satellite imagery is increasingly introduced as evidence in human rights cases, many domestic courts continue to debate its admissibility (Purdy and Leung, 2012).

As the Chinese state expands from government to governance and governmentality, which targets social production rather than social control (Taylor, 2007), it is relying on and reimagining EO. A Chinese sociotechnical imaginary—defined as “collectively imagined forms of social life and social order reflected in the design and fulfillment of nation-specific scientific and/or technological projects” (Jasanoff and Kim, 2009: 120)—of EO can thus be identified with a distinct scope, scale, temporality, and applications. China's sociotechnical imaginary of EO is data-driven, inductive, predictive, and societally focused. EO is also perceived as key to national development. The 2022 white paper on China's space program declares, “Under the overarching goal of building a safe, healthy, beautiful and digital China, we will intensify the integration of satellite application with the development of industries and regions, and space information with new-generation information technology such as big data and Internet of Things” (State Council Information Office of the PRC, 2022).

Chinese remote sensing scientists are working to achieve national aims, turning political discourses into political technologies: tools used by the state to pursue its objectives, like managing territory, resources, and populations. The Chinese state and science are deeply entangled, particularly in the capital- and resource-intensive fields of satellite EO and big data. Researchers are combining satellite imagery collected at improving spatial, spectral, and temporal resolutions with what Chinese and foreign scientists often call “social sensing” (社会感知, shehui ganzhi) or data generated by individuals and their devices (Janowicz et al., 2020; Liu et al., 2015). In the era of the Internet of Things, these include a widening range of sensor-equipped objects from automobiles to refrigerators and, of course, mobile phones. Smartphone data is voluminous in China, which has the world's highest number of mobile devices despite urban–rural inequalities in their adoption. Given the global rise of Chinese science and the Chinese state's interest in commercializing, exporting, and standardizing EO services, Chinese EO may reshape international EO.

Critical scholars have long recognized that EO has never been merely a scientific technique. As Lövbrand et al. (2009: 10) observe, satellite imagery and Earth system models open a “new political space for government intervention.” Critiques of EO as a political technology focus on foreign and military surveillance (Nardon, 2007; Shim, 2014), interrogating how the power to see and be seen from afar affects (inter-)state behavior and relations.

As the Chinese state and science broaden the remit of EO from monitoring large-scale lands and resources to anticipating and governing individuals’ behavior, attention should be paid to EO's transformation into a domestic disciplinary technology. Satellites’ ability to see at the submeter scale empowers the surveillance state, which public concerns in many liberal societies over the data's threat to individual privacy reflect. More consequential, however, may be the accumulation of big EO data to support inductive techniques in which general inferences and theories are made from specific observations. The trend of using big data to generate rather than test hypotheses is apparent across international remote sensing. Techniques like machine learning applied to large-scale, multitemporal image sequences are pushing out cognitive approaches, which favor pre-existing knowledge and ontologies (Arvor et al., 2019; Zhu et al., 2017). As image analysis moves toward starting with pixels rather than spatial categories, it may even “revisit and consider characteristics of (human) vision as pre-requisite” (Sudmanns et al., 2020: 843). Recognizing this technical shift and transcending visual critiques of EO focused on discrete images, I consider the political risks of redeploying EO, in the age of big data, as a practice of prediction rather than observation. To quote Antoinette Rouvroy, “The hallmark of power in the algorithmic regime is not so much that it would look at us, spy on us, but rather that it literally bypasses the visual episteme” (Groos, 2023).

This paper's empirics are gathered from a review of scientific papers published by Chinese remote sensing scientists largely in English (identified on Web of Science) and occasionally in Chinese (identified on the China National Knowledge Infrastructure (CNKI) database) and from Chinese white papers and media reports. Together, this material informs a genealogy of China's sociotechnical imaginary of EO. I approach these empirics as a political geographer trained in remote sensing. My interpretation is ultimately partial and limited by my lack of Chinese language skills; select translations were carried out with the generous assistance of Mandarin-speaking colleagues. At the same time, the analysis is enhanced by my understanding of geospatial science and capacity to critique its scientific and political ramifications. I also spent three years as an academic in Hong Kong, where I was regularly exposed to EO science from Mainland China, much of which focused on smart cities.

While publications in Chinese scientific journals disseminate knowledge domestically, they are little referenced by their international counterparts (Zhou and Leydesdorff, 2006). Interrogating the practice and imagining in China of scientific fields like EO diversifies accounts of the intellectual history of Earth system science, which privilege the West (Rindzevičiūtė, 2018). Such work also contributes to critiques of China's governance of data and cyberspace (e.g. Wang, 2022), especially as the latter diverges from the West while still shaping global standards (Seaman, 2020). With China's impact on global science and governance growing, geopolitically and scientifically informed “anticipatory accounts,” following Egliston and Carter (2022), of the country's sociotechnical imaginaries are timely. Forward-looking analysis may help problematize and resist technologies narrativized as rational, scientific, and apolitical—if not necessarily in China, at least before they are exported and globalized.

The optics of observation: Chinese visions from space

Critiques of EO typically consider satellites’ monolithic, totalizing views of the Earth's full disc (Cosgrove, 2001; DeLoughrey, 2014). Recent research unpacks the idiosyncrasies of EO cultures and practices, with scholars deconstructing the politics and economics of satellite ecosystems (Alvarez Leon, 2022; Cirac-Claveras, 2021; Jirout, 2017). These studies reveal that despite satellites’ presumed “gaze from nowhere” (Haraway, 1988: 581), EO is a heterogeneous practice. A variety of norms, standards, and objectives—many of which are determined by national governments—shape EO from country to country, meaning that no two satellites see alike. Even if they had identical specifications, their data would be interpreted and applied in different ways depending on the individual analyst, national political and scientific cultures, and many other variables. These differences confirm Krige's (2008: 39) contention that “we should not shy away from admitting the complexity and diversity of the space effort, nor pretend that the view of the world from Washington is the only view worth recording.” Extending this argument, Cirac-Claveras (2021: 4) suggests that there is no “single universal space age vision.” There is also no single satellite vision. Deleuze (1992: 6) asserts, “Types of machines are easily matched with each type of society—not that machines are determining, but because they express those social forms capable of generating them and using them.” In China, satellites—once paradigmatic “vision machines”—are being designed to achieve “sightless vision,” as Virilio (1994: 59) anticipated. By promising to foretell the future, computers and algorithms may overcome the cardinal empirical sense, at least in Western cultures (Jenks, 1995): sight.

Applying the concept of sociotechnical imaginaries, which involves analysis of technoscientific and political discourses (Sadowski and Bendor, 2019), exposes the social and scientific construction of satellite EO. Due to its photorealism and detachment from its subject, the technique is less obviously political and subjective than cartography, whose inherent artistry makes it easier to grasp that “all maps lie” (Monmonier, 2018). Still, just as maps shape the world, satellite images can, too. The development of EO services, which states such as China seek to empower with decision-making abilities, may lend political agency to such techniques, producing worldmaking technologies—particularly once adopted and “locked in” by society (Foxon, 2007).

Sociopolitical imaginaries can be studied at various scales, though the national is most common. While risking methodological nationalism, nationally framed studies are appropriate for critiquing EO satellites. Until 1986, national governments (individually and through bilateral and multilateral collaborations) were the only entities operating the spaceborne instruments. Since the 1990s, commercial EO has expanded in countries like the USA, but in China, the national government still predominates (Liu et al., 2019). Moreover, state-run scientific bodies like the Chinese Academy of Sciences (CAS) call for developing science and technology to help “orchestrate the needs of both the nation and society” (Li et al., 2016: 4). As a result, studies of “Chinese” sociotechnical imaginaries are arguably more justifiable and coherent than in countries where the state has a looser grip on science.

The relevance of sociotechnical imaginaries to China studies has led to the concept's adoption across research into the country's scientific and technological projects. At a range of scales in China, researchers critique sociotechnical imaginaries of genetically modified rice (Chen, 2015), offshore wind (Korsnes, 2016), ecocivilization (Hansen and Svarverud, 2018; Huang and Westman, 2021), the circular economy (Schulz and Lora-Wainwright, 2019), the Great Yangtze River Protection Programme (Sheng et al., 2022), artificial intelligence (AI) in military policy (Bächle and Bareis, 2022), digital sovereignty (Cong and Thumfart, 2022), and terraforming (Bashford and Galka, 2022). Few studies, however, interrogate Chinese sociotechnical imaginaries of a scientific discipline. One exception is Au (2022)'s examination of precision medicine, though scientific discourse rather than research is scrutinized. By attending to the scientific rather than social construction of sociotechnical imaginaries, I depart from such studies based on narrative analysis (see also Egliston and Carter, 2022; Pickersgill, 2011).

As the Chinese state advances the use of EO domestically and abroad, consideration must be given to both the domestic origins of sociotechnical imaginaries and what happens once they are exported. Unrestricted by scale or location and seemingly objective, EO data, models, and services are easily shared and sold. Yet satellites are the products of major national technology programs and cement “particular imaginations of risk and benefit, public good, and nationhood” (Jasanoff and Kim, 2009: 122). Satellite imagery produced in any one state reflects specific and situated concerns over issues such as individual privacy versus safety (risk/benefit); clandestinity and commercial versus open-access data (private/public good), and public versus private satellite programs (nationhood and the role of the state/market). Consequently, pixelated pictures of Earth bear the invisible imprint of sociotechnical imaginaries that fuse “cultural and political-moral virtues” with “technological, judicial, and political values” (Hansen et al., 2018: 196).

While the international community heralds EO's potential to manage the global environment, Chinese officials and scientists imagine its potential to manage and produce society. Such use would render EO not only a political technology but also an antipolitical technology when seen through a liberal lens. In democratic societies, politics is fundamentally understood as the valuing and protection of public contestation, dissent, and disagreement (Barry, 2002). There, even as techniques like predictive policing take hold, doubts remain regarding their validity and efficacy (Mugari and Obioha, 2021), as does skepticism of their attempts to divine future behaviors. Antipolitics “treat[s] the social world analogously to the natural world as a set of dependent variables,” viewing political resistance as symptomatic of “ignorance” or “irrationality” (Schedler, 1997: 12). Antipolitical technologies try to determine and even influence forthcoming events, preempting the radical and spontaneous new beginnings made possible by politics (Arendt, 1963). Through predetermination, antipolitical technologies terminate humanity's “world-building capacity” (Arendt, 1963: 175) and replace it with computation.

Satellite state: Earth observation for a modern government

Conventional histories of EO are periodized according to advances in satellite technology. Applying this approach to Chinese EO, three stages are identifiable: reliance on foreign satellites (1980–2008), development of environmental EO satellites (2008–2013), and application of high-resolution (Gaofen (GF)) satellites (2013–2020) (Gao et al., 2020). This article instead brackets sociotechnical shifts in Chinese EO and situates them within broader political and economic changes.

Chinese EO developed when the Chinese government was promoting economic development and industrialization during the Reform and Opening-Up period beginning in 1978. This context shapes the national sociotechnical imaginary surrounding EO's initial conception, development, and deployment. In the USA, photographic reconnaissance satellites were first engineered during the Cold War to surveil enemy territories. The data's scientific value for monitoring processes such as forestry and urbanization was quickly recognized, giving rise to scientific EO in the USA. Meanwhile in 1975, China became the third country to launch recoverable high-resolution satellites with the military's Fanhui Shi Weixing (返回式卫星, “Recoverable Satellite”) program (Harvey, 2013). Since these only managed to orbit briefly, Chinese satellites could not provide the large-scale, repeat imagery needed for scientists to fulfill the tasks of modern government like managing lands and resources. As the USA began making its EO satellite imagery available to the global scientific community under its “open skies” policy (Draeger et al., 1997), Chinese scientists purchased hard copies and magnetic tapes of US Landsat imagery, representing their first sustained encounter with EO data (Jirout, 2017).

While Chinese applications of EO data to target large-scale land and resource development resembled those of satellite powers such as the USA, Soviet Union, France, and Japan, Chinese scientists had to rely on foreign as opposed to domestic data. EO initiatives were included in the Chinese Communist Party (CCP)'s Five-Year Plans, which spell out the state's social and economic goals and technologies of national interest. China's Sixth Five-Year Plan (1976–1980)—the first following the Cultural Revolution—called for the “Teng Chong” EO pilot project, through which CAS scientists produced information about land resources, forestry, and land cover in Yunnan using print-outs of multispectral Landsat imagery. Teng Chong and the Tianjin Bohai Bay and Ertan hydropower remote sensing experiments together formed the “three campaigns” underpinning China's EO push. Other projects that followed continued to use EO data in a piecemeal fashion. The first national land survey (1984–1996), for instance, involved mixed methods including “aerial photos, Landsat images, and maps” and required 2 million people, over 1 billion yuan, and 12 years to complete (Lin and Ho, 2003: 88).

Seeking a more systematic and efficient route to land surveying and management, in the late 1980s, the China Academy of Space Technology (CAST) began designing its own multispectral EO satellites. The country launched its first meteorological satellite in 1988 as part of the Fengyun (风云, “Wind Cloud”) program (Borowitz, 2020). That same year, CAST and its Brazilian partners commenced cooperation on two satellites to map Chinese and Brazilian national resources (de Oliveira Lino et al., 2000). The first China–Brazil Earth Resources Satellite (CBERS) was launched in 1999, followed by its successor, CBERS-2, in 2003. Chinese officials suggested that the latter instrument could help fill the data gap created by the malfunctioning Landsat 7 satellite (Covault, 2003), indicating interest in disseminating EO data internationally. CBERS laid important foundations for Chinese-led EO, with applications in “agriculture, forestry, water resources, national territory resources, ecological environment and regional planning application” (Song et al., 2006: 22).

In 2001, China National Space Administration published its Tenth Five-Year Plan, which aimed to create “a varied remote sensing system that has long-term stability and an integrated space-ground application system” and launch eight new civil satellites: “one Haiyang ocean monitoring satellite, two Fengyun-3 meteorological satellites, two earth-space science satellites, and three environmental and disaster monitoring satellites” (Pollpeter, 2008: 5). Marking a rapid advance in Chinese EO's synopticism, five years later, China's second white paper on outer space sought “to form an all-weather, 24-h, multispectral, differential-resolution EO for stable operation, and achieve stereoscopy and dynamic monitoring of the land, atmosphere, and sea” (State Council Information Office of the PRC, 2006).

China has since become a global EO satellite power. As of December 31, 2022, out of the 542 known Chinese satellites currently in orbit, 296 are for EO, most of which the government operates (Union of Concerned Scientists, 2023) (Figure 1). Their capabilities range from very high-resolution (VHR) multispectral imagery to C-band synthetic aperture radar, which can pierce through clouds and darkness. China's EO program has collected 1500 petabytes of data—the largest national “stock in the world,” according to CAS scientist Li Deren (Sohu.com, 2017). Li has led efforts to transform EO from an observation tool into what he and his co-authors term an “EO brain:” an “intelligent Earth observation system which simulates the brain perception and cognition process” (Li et al., 2017: 136). Reimagined as cognition rather than observation, EO is becoming a vital predictive tool for Chinese governance and governmentality.

Figure 1.

Known Chinese satellites currently orbiting Earth, by purpose, operator, and year launched.

Governance by satellites: Crystal balls as the long arm of the law

Since Reform and Opening-Up, the Chinese state has expanded its targets from the economy to society and the environment. As the state has broadened from government to governance and governmentality (Hoffman, 2007; Jeffreys and Sigley, 2009), it has refined its political technologies. While the post-1949 Chinese state concentrated on managing lands, resources, and the economy with a heavy hand, in the 1990s, it began intervening in social and environmental issues with more sophisticated techniques (Han and Lai, 2012). Human quality began to form “an object of enquiry, intervention, and surveillance of human behavior” within Beijing's pursuit of an “exemplary society” (Bakken, 2000: 1). Ensuring the “quality” of citizens is a national objective, as exemplified by social credit systems and regulations guiding “harmonious” behavior (Wong and Dobson, 2019). To manage the nascent disciplinary society in which technology plays a key role in enforcement and individuals’ self-regulation (Huxley, 2008), the Chinese state is incorporating EO into governance.

Soon after establishing a national satellite EO system in 1999, Chinese scientists and officials expressed interest in carrying out social observations and applications from space. In wider histories of EO, VHR EO imagery is sometimes considered agentive, compelling states to monitor and track individuals (Froomkin, 1999). Yet Chinese scientists appear to have envisioned human-focused EO before they had ready access to VHR imagery. In 2002—the same year satellite imagery was used in a Chinese court to convict an individual of illegal deforestation—Li Chuanrong, Deputy Director-General of the China Remote Sensing Satellite Ground Station, noted, “If we realized that the EO is being in transition from science supported to EO business, and if we agree that EO business is principally an information business, we have to operationally provide the valuable information for the demands of economy, social strategy, politics and environment in daily bases [sic]” (Li, 2002: 1611). In 2010—three years before China would launch its first VHR satellite—one Chinese master's thesis observed that remote sensing had “played an important role in farmland protection, land monitoring, law enforcement and deterrent” (Li, 2010). Evidently, interest in using EO for regulating human behavior predated high-precision vision.

Leading calls to use EO to intervene in society, Guo Huadong, Director-General of the CAS Institute of Remote Sensing and Digital Earth, has asserted that China should “build an effective and sustainable Earth observation mission and planning system…to serve the country's significant political needs and sustainable socio-economic development in the context of globalization” (Guo, 2013: 98). Guo champions leveraging the expanding amount of EO data, referred to as “big Earth data” or “big EO data,” which he argues can support real-time decision-making (Guo, 2017; Guo et al., 2020b). His research helps realize state goals of operationalizing big data for governance, which are laid out in documents like the State Council's Action Plan on Promoting Big Development, released in 2015. “Facing the needs of economic and social development,” the white paper says, China should “develop a scientific big data application service center to support the resolution of major issues of economic and social development and national security” (State Council, 2015). This objective reflects an antipolitical view of big data, seeing its promise lying in not only providing societal solutions but also resolutions that preclude future conflict. Chinese law enforcement has increasingly adopted EO, particularly for land management. In 2016, Guo Renzhong of Shenzhen's Municipal Commission of Planning and Land Resources gave a presentation to the United Nations that explained, “In order to enforce the land use plan, it's necessary for the central government to monitor and review the overall land use conditions” (Guo, 2016). Using EO to regulate people's land use—in other words, their conduct—transforms EO into a disciplinary technology.

Chinese scientists’ and officials’ embrace of EO to guide individual and collective behavior is reflected in the establishment of the Chinese High-Resolution Earth Observation System (CHEOS): a civilian VHR satellite system with environmental and societal applications. CHEOS is unique, as similar systems in other countries are all clandestine or commercial. It comprises seven GF satellites, the latest of which, GF-7, has a 10-cm resolution. Chinese scientists explain that “…with the characteristic of high spatial, temporal, and spectral resolution,” CHEOS “is aimed at newly establishing an all-day, all-weather coverage Earth observation system for satisfying the requirements of social development” (Chen et al., 2022a: 1). Chinese scientists’ access to VHR imagery may allow them to hone its applications faster than their foreign counterparts, who generally must purchase expensive commercial imagery with grants or rely on donated imagery.

Chinese research using GF imagery exhibits a predictive bent, aspiring to turn the state's optical powers into oracular powers. One project, organized in response to a call by the Ministry of Lands and Resources issued to China's State Land Agencies, developed an early-warning and real-time monitoring platform for land supervision. In 2017, the Jinan Bureau of State Land Supervision piloted the platform, which “focus[ed] on the situation of illegal land use, permanent farmland protection and urban exploitation boundary breakthroughs in the supervision areas, and carried out real-time monitoring and early warning for all levels of land management departments to provide real-time image services” (Sun et al., 2018: 1585). By the following year, the authors noted that the platform had processed 1400 satellite images in real time and had “truly achieved ‘early warning, early detection and early deterrence’” (Sun et al., 2018: 1585). An article by two scientists from the Surveying and Mapping Institute of the Lands and Resource Department of Guangdong Province similarly used GF data for land use enforcement. In their paper, “Research on the Application of Gaofen-7 in Law Enforcement and Monitoring of Lands Resources,” the scientists state that satellite imagery is crucial to the CCP's push for strict land management and regulation. They argue:

…to carry out the monitoring and law enforcement of land resources is to give full play to the role of satellite remote sensing technology in the dynamic supervision of land resources utilization. The major decisions of strengthening the supervision of land law enforcement and standardizing the order of land management can effectively curb the rising trend of land violations, effectively realize the protection and supervision of cultivated land resources, and promote the economical and intensive use of land resources (Liu and Liu, 2020: 71).

With early warning and deterrence integral to Chinese EO, its temporality is shifting from sight to foresight. The CCP's Five-Year Plans and associated white papers reflect this change to the technology's sociotechnical imaginary. The Tenth Five-Year Plan (2001–2006), Eleventh Five-Year Plan (2006–2010), and Twelfth Five-Year Plan (2010–2015) focused on engineering satellites. While the Fourteenth Five-Year Plan (2021–2026) says little regarding EO, the Central Commission for Cybersecurity and Informatization's Fourteenth Five-Year Plan for National Informatization (2021–2026) spotlights satellite services with an eye to prediction. Though EO typically entails collecting historical observations, Chinese EO extends to future processes. In the National Informatization Plan, under the “major task” of “Building Jointly Constructed, Jointly Governed, and Jointly Shared Digital Social Governance Systems,” a subsection on “Emergency Response Management Modernization Capability Enhancement Projects” alludes to prevention rather than response:

Enhance risk monitoring and early warning capabilities. Build comprehensive monitoring and early warning systems for natural disasters and production safety risks; gather monitoring resources for natural disasters, production safety and urban and rural supervision and control, build the ‘SkyEye Net’ satellite constellation emergency response system; enhance comprehensive monitoring of multiple disaster categories and disaster chains, and early risk identification and early warning capabilities (Creemers et al., 2022: 35).

A second example relating to “AI Social Governance Experimentation Projects” proposes combining satellite EO and AI:

Launch social experiments on risk prevention. Explore and research the integration of artificial intelligence and satellite remote sensing, audiovisual supervision and control, the Internet of Things, emergency response broadcasts, etc.; identify risks early and provide precise early reports and warnings in areas such as production safety, urban safety, and natural disasters, reducing harm to the popular masses’ lives and property (Creemers et al., 2022: 34).

While Western sociotechnical imaginaries see EO as a guarantor of the international order by deterring “rogue states” from unwarranted actions, Chinese imaginaries perceive EO as a guarantor of domestic order.

China's application of EO for domestic governance is exemplified by the Ecological Conservation Redline Supervision Platform announced in April 2023 (Ministry of Ecology and Environment, 2023a). The “nationwide satellite and automated monitoring system” (Lewis, 2023: 225), which is being applied while still under development, combines observations from 30 foreign and domestic satellites, aerial remote sensing, ground observations, and video surveillance. The director of the Department of Natural and Ecological Protection of the Ministry of Ecology and Environment claimed that 30,000 patterns of human activities and “2000 clues to ecological damage” have already been identified (Ministry of Ecology and Environment, 2023b). The “sky–space–ground” monitoring network is intended to protect 30% of China's land and 150,000 km² of marine areas from illegal development—areas bounded by the Ministry of Ecology and Environment with “ecological redlines” (Bai et al., 2018).

Strategically, the platform is designed to have a public participation component. This addition allows people to provide information on suspected illegal development, complimenting top–down surveillance with bottom–up informing. Under digital authoritarianism, the popular legitimacy of EO and AI as political technologies derives not only from their capacity to maintain public order and deliver economic growth (Zeng, 2020) but from making people feel they can influence the black box. To sustain digital political technologies, algorithmic inclusion may prove more vital than algorithmic transparency. Once included, however, users can be exploited for their labor and severed from their data (Owen, 2023). Inclusion, exploitation, and alienation are exacerbated on platforms that gather data passively, which are becoming inescapable parts of public infrastructure in societies such as China (de Kloet et al., 2019).

Seeing like a state: People as sensors

Upending EO's top–down gaze, the Chinese state and science are situating its techniques within individuals through the development of social sensing. The technique involves “outsourcing sensing tasks to the general public” (Zhu et al., 2022: 1), with “each individual play[ing] the role of a sensor” (Liu et al., 2015: 512). Mobile phone location-based services, taxi data, and social media check-ins represent social sensing data (Chen et al., 2018). Since much of this data is generated without active consent, social sensing resembles a distributed panopticon analogous to satellites, which surreptitiously take pictures from above. Scientists at Peking University and the University of California, Santa Barbara describe social sensing as “a category of spatio-temporally tagged big data that provide an observatory for human behavior, as well as the methods and applications based on such big data,” adding that it should be “viewed as an analogue of remote sensing that excels at collectively sensing our socioeconomic environments” (Liu et al., 2015: 526).

Some Chinese scientists see social sensing as improving upon remote sensing due to its more precise detection of human activities. This opinion contrasts with sociotechnical imaginaries of EO in places like the USA, where commercial VHR satellite imagery providers seeking to bolster their legitimacy and trustworthiness reassure the public that they are unable to see individuals. The introduction to a special issue on “Social Sensing and GeoAI” expresses that with “geographic big data, the inversion process ‘from people to the ground’ can be realized, which to a certain extent makes up for the inadequacy of traditional remote sensing technology that focuses on sensing physical geography” (Yang et al., 2022: 327). Regardless of whether social sensing is superior to remote sensing, Liu et al. (2015: 527) believe that integrating social and remote sensing data will be “an attractive research topic.” Social sensing has indeed proven popular among Chinese scientists, reflecting both the empirical availability of geolocational data generated by the country's billion-plus mobile phone users and state desires to perfect disciplinary technologies. One exemplary study produced precise city and county population estimates by combining NASA nighttime light satellite data and European Space Agency land use/land cover data with location-based service data for 800 million users of Tencent, one of China's largest internet service providers (Xu et al., 2021). More than just counting the population, such data synthesis may support its manipulation. Xu et al. (2021: 6) conclude, “The use of these highly precise open data sources to monitor urban population mobility and to predict the behavior of various citizens at the street level will be beneficial for intelligent urban planning and community management.”

If China seeks to “order people and things” (Larner and Walters, 2004: 3) and “life itself” (Greenhalgh, 2009), it does so above all in cities. In the smart cities that China is engineering (Cowley et al. 2018), urban productivity, orderliness, and quality depend on the behavior of citizens-turned-sensors. Approximately 65% of China's population lives in cities, which constitute a key organizing unit for reproducing authoritarian state power (Cartier, 2015). Chinese research integrating remote and social sensing focuses on urban environments, mapping “vitality” block by block (Yang et al., 2021), urban land use (Huang et al., 2020), residential land suitability (Huang et al., 2019), and the social functions of urban green space (Chen et al., 2018). High-resolution satellite imagery and social sensing data have been integrated to identify “fine-grained urban villages” to inform “future urban renewal decisions” (Chen et al., 2022b: 12). Such research demonstrates the potential for inductive models to drive urbanization policies. Chen et al. (2022b: 12) propose applying their method to “slums in India and Brazil,” raising the possibility of algorithms trained on Chinese data traveling to foreign places with more uneven digital landscapes.

Exposing the enrollment of social sensing science into national surveillance, numerous studies have been published by scientists affiliated with state security organs like the Third Research Institute of the Ministry of Public Security in Shanghai. Its work informs domestic policing and intelligence and counterintelligence efforts (e.g. Tang et al., 2019). One study in Neurocomputing led by an affiliated author explores “mobile crowd sensing of human-like intelligence using social sensors": humans who form “not only data consumers, but data producer” [sic] generating real-time data on platforms like Weibo and Twitter (Xu et al., 2021: 3). Mobile crowd sensing's potential applications include “public security, smart cities, location based services, etc.” (Xu et al., 2018: 3). The former two represent domains in which the Chinese state perceives opportunities for centralizing and instrumentalizing big data (Große-Bley and Kostka, 2021). These opportunities crystallized during the pandemic, when the state normalized individualized interventions such as regulating people's movements based on their smartphone-reported health status (Liu and Bennett, 2020).

Chinese scientific research and state discourses justify social sensing by claiming that it popularizes governance. Yet its alienation of decision-making and attempted foreclosure of the future render the technology deeply antipolitical. Mao Mingrui, an executive of various smart city organizations, argues in a report published by the Beijing Municipal Bureau of Coordinated Administrative Law Enforcement for Urban Management that social sensing, when defined as citizens’ hotlines and mailboxes for reporting problems, has always informed urban governance (Mao, 2020). He criticizes this antiquated approach, however, as “passive governance.” If social sensing is properly leveraged, he contends, “Citizens are not required to take the initiative to report. Rather, before the problem occurs or there is any obvious problem, city managers learn about urban problems in advance in order to inform governance decisions” (Mao, 2020).

A study examining street vending violations echoes this belief in social sensing's ability to perfect and popularize governance. Li et al. (2022) compare “bottom–up” social sensing data (phone calls to a citizen complaint hotline) with “top–down” data collected by a “grid-based smart urban governance system” (surveillance videos and reports filed by law enforcement and citizens). The authors recognize the limits to both forms of data: the “top–down” data does not actually offer “a “God's-View observation” (p. 9), while the social sensing data exhibits perception bias. Still, they claim that once detected, perception bias can be harnessed to “nudge” and improve urban governance (p. 1). Li et al. (2022: 1) determine, “This research has contributed to promoting the people-oriented transformation of urban governance.”

While contending to put individuals at the core of governance, Chinese sociotechnical imaginaries of EO, particularly when combined with social sensing, turn the body politic into a data apolitic. The population's ongoing subjectification requires ceaseless data production and consumption from all angles. China's National Informatization Plan urges the country to “formulate grassroots smart governance standard systems [and] promote the application of technologies such as information recognition technology, high-resolution satellite sensing and imaging, three-dimensional mapping, audiovisual imagery as well as smart sensing” (Creemers et al. 2022: 50). Scientists like Gao et al. (2020: 716) heed the state's call to invert the panopticon by encouraging the “popularization and application of domestic satellite remote sensing data.” They urge:

The state should innovate the management mechanism, advocate and encourage people from all walks of life to give priority to the use of domestic satellite remote sensing data, form the active rather than passive behavior of ‘I want to use’, and improve the comprehensive efficiency of domestic satellites and the level of scientific and technological development (Gao et al., 2020: 716).

Ensuring public buy-in is important to the success of China's expanded vision of EO, which relies on feedstocks of satellite and social data. While including individuals and their information in digital decision-making processes may help grant digital authoritarianism popular legitimacy, equally critical is that the new remote sensors stay plugged in.

Globalizing EO “with Chinese flavors”

Following Western actors, who pioneered the application of spaceborne EO to development interventions in the 1980s (Paul and Mascarenhas, 1981) and the sustainable development goals (SDG) in the 2010s (Kavvada et al., 2020), the Chinese state seeks to export its Earth observation and prediction systems. Chinese scientific research and political discourses integrating remote and social sensing mirror US narratives in the 1960s, which framed satellite EO as benevolent to the global community (Black, 2019).

Yet more than exporting EO data, China seeks to export EO services. Describing the aforementioned CHEOS program, Chinese scientists Chen et al. (2022a: 2) state, “CHEOS promotes the transformation and upgrading of satellite systems from ‘focusing on research and test’…, formulates the satellite product and application system with Chinese flavors, and progressively provides global services.” Commodifying EO services “with Chinese flavors” advances a suggestion made in 2002 by Li Chuanrong, Deputy Director-General of the China Remote Sensing Satellite Ground Station. He recommended that EO technologies be used to serve “national economic construction” and exported (Li, 2002: 1610). Already, China has made headway in disseminating other satellite services like its BeiDou Navigation Satellite System. The global position, navigation, and timing (PNT) service offers an alternative to the US Global Positioning System (GPS) and the European Union's Galileo. BeiDou, whose data is freely available like GPS and Galileo, has proven popular with countries participating in China's Belt and Road Initiative and Digital Silk Road (Sewall et al., 2023), potentially paving the way for their use of Chinese EO services.

As satellite remote sensing lends itself to planetary governance (Gabrys, 2020), China may encounter success in exporting its EO services and sociotechnical imaginaries. Whereas China's “datafication” of marginalized minorities such as Uyghurs (Ma, 2022) via techniques such as face recognition and forensic genetics sparks demands from foreign scholars for journals to retract “unethical work” (Moreau, 2019), nationally scaled EO and social sensing research is less controversial. Moreover, journals such as International Journal of Digital Earth and Big Earth Data, published by Taylor & Francis on behalf of the Beijing-based yet cosmopolitanly named International Society for Digital Earth with CAS support, offer outlets for research that circulates and legitimates Chinese EO norms, data, and services. Public-facing media also promote Chinese EO services. In 2019, an English-language YouTube video was released to promote Alibaba Cloud Planet Engine, a Chinese AI-integrated satellite imagery and remote sensing platform (https://www.youtube.com/watch?v=bqUcQW7VEK4). A voiceover mimicking an American astronaut intones, “I’m not just a selfie stick for Planet Earth. I have intelligence, too.” The video's Chinese subtitles differ subtly: “Am I just a pair of eyes that sees far enough? No, I can think, too,” refocusing EO from the optical to predictive.

China's effort to internationalize its EO services is exemplified by a satellite developed by CAS and launched in 2021: the Sustainable Development Science Satellite 1 (SDGSAT-1). While under development, the instrument was called the CASEarth satellite and deemed the world's first “human trace satellite” (Guo et al., 2020a: 914). Perhaps to avoid suggestions of surveillance, three years later, Guo et al. (2023: 34) described the quietly renamed satellite as able to “precisely depict traces of anthropic activity” with its thermal infrared spectrometer, glimmer imager, and multispectral imager. Whereas the CASEarth satellite was intended to provide insights into “socioeconomic development in China at a very fine scale” (Guo et al., 2020a: 913) and “provide a new model for sustainable development decision making” (Guo et al., 2020b: 761), the rebranded SDGSAT-1 provides a “continuous stream of data on the Earth system and human-environment interactions to enable multi-scale data support to policymakers and fill in data gaps” (Guo et al., 2023: 37). The strategic repurposing of the CASEarth satellite as SDGSAT-1 suggests a state desire to legitimize Chinese EO through international organizations such as the United Nations, much as in other standard-setting bodies (Camfferman, 2020). While Western countries are rejecting the adoption of many Chinese technologies, resisting technical standards is trickier, partly because they seem non-political (Rühlig, 2023). The task at hand is to expose not only the geopoliticization of apparently apolitical practices but attempts to render them antipolitical, too.

Conclusion

Since the 1960s, the West's dominance of EO has made the satellite gaze seem natural and universal. Yet as Chinese sociotechnical imaginaries dislodge hegemonic theories, practices, and applications of EO, its politicization and subjectivity are becoming more evident—as is the fact that it was never merely a scientific technique. EO has always been a political technology, whether to police the world, surveil city streets, or simply manage farmland and forests. Only the targets of that political technology have changed, from enemy territories to unruly citizens.

China is shifting EO away from observation toward prediction. The country has advanced from using foreign satellite imagery to launching its own satellites and, now, turning the country's billion-plus smartphone users into remote sensors. Their data is feeding urban governance and law enforcement, transforming EO from an act of passive environmental monitoring into one of continuous political intervention. In post-pandemic China, where collecting individuals’ data and hypervigilant governance are normalized, multivalent EO spanning planetary orbits and individual bodies may become technologically and politically possible.

While Chinese governmentalities have been interrogated domestically (Hoffman, 2007; Jeffreys and Sigley, 2009), their global extension merits scrutiny. When any technology travels—particularly ones forged through close cooperation between the state and science—it brings with it the sociotechnical imaginaries of the source country. These are well-hidden in technologies based on inductive reasoning, which makes them seem flexible and adaptable while veiling their normative frameworks. As China exports big data-driven political technologies narrativized as scientific and rational, their uptake may spur adoptive states and publics to make point-and-click interventions in society and the environment without full appreciation of their political consequences. Wider digital divides in certain regions where China seeks to export its services, such as Africa, the Middle East, and Southeast Asia, may also result in a smaller proportion of the population being represented in social sensing than in China.

Future research should consider how Chinese scientific methods, models, practices, and norms are “going out,” helping to bridge work on Chinese governmentalities, Chinese science and technology studies (STS), and global China. Ethnographic STS could prove eye-opening, though such work may be challenging to pursue, particularly for foreigners. Advances in studying networks of science with citation data offer alternative means of analyzing and mapping Chinese scientific collaboration (e.g. van der Wouden, 2022). Attention should also be paid to the funders of Chinese EO and social sensing research given the regular involvement of state security organs. Research into Chinese EO publication and funding networks may help determine whether Chinese EO is being globalized and indicate its potential for becoming a hegemonic practice with worldmaking powers. Finally, histories of Chinese scopic regimes (Wei, 2017) and their gradual displacement by algorithmic regimes require unveiling, too.

While EO once produced knowledge through sight, EO with Chinese characteristics privileges foresight. If to see an object is to begin to destroy it (Virilio, 1989), what happens when a state can pull back the curtain on a presupposed future? As EO and social data explode and people become both remote sensors and users of remotely sensed data, identifying where and how they hold power within EO and social sensing systems will be critical. Somewhere between the pixelated interstices of submeter anarchy and planetary governmentality, the “simulacral protagonist” (Baudrillard, 1995: 24) may be found.