Political Geography 78 (2020) 102127 Available online 23 December 2019 0962-6298/© 2019 Elsevier Ltd. All rights reserved. Is a pixel worth 1000 words? Critical remote sensing and China’s Belt and Road Initiative Mia M. Bennett Department of Geography and School of Modern Languages and Culture (China Studies Programme), The University of Hong Kong, Room 1023, 10th Floor, The Jockey Club Tower, Centennial Campus, Pokfulam Road, Hong Kong A R T I C L E INFO Keywords: Critical remote sensing Development Night lights Infrastructure China Belt and Road Initiative ABSTRACT As a novel means of researching China’s Belt and Road Initiative (BRI), this article advances a critical remote sensing agenda that connects the view from above provided by satellite imagery with the grounded, qualitative methodologies more typical of political geography such as ethnographic feldwork. Satellite imagery is widely used to produce empirics relating to the BRI, and the Chinese state is showing increasing interest in applying Earth observation data to governance. A more critical approach attentive to the politics of remote sensing, especially in light of China’s emergence as a space and satellite power and its embrace of big data, is needed to more precisely reveal what changing pixels represent on the ground and expose the potential issues with data captured from high above the planet. This paper offers three theoretical and methodological objectives for critical remote sensing. First, I refect on the geopolitics involved in the production and analysis of satellite imagery. Second, through analysis of night light imagery, which captures illuminated anthropogenic activities, I interrogate metanarratives of development. Third, I engage with qualitative methods by “ground-truthing” remote sensing with ethnographic observations along China’s borders. I also seek to avoid the methodological nationalism often present in remote sensing research by situating these mixed-methods case studies at scales above and below the nation-state. As one of the largest development interventions in history materializes, pursuing critical remote sensing can create opportunities for social scientists to leverage quantitative and geo- spatial methods in support of more equitable and sustainable futures. 1. Introduction: the power relations of remote sensing In 2013, Chinese President Xi Jinping launched the “One Belt, One Road” policy, now known as the Belt and Road Initiative (BRI), to improve trade and transportation networks across Eurasia and beyond (Mayer, 2018). The BRI assembles variegated processes of Chinese do- mestic and international investment under one heading with a US $4–8 trillion dollar price tag (Jochec & Kyzy, 2018). By Beijing’s count, as of July 2019, 115 countries were partnering in the BRI. But there is still no offcial map of included areas, routes, or projects (Narins & Agnew, 2019). Indeed, the initiative’s seductive power arguably fows from its “amorphous and ever-changing scope” (Oliveira, Murton, Harlan, Rippa, & Yang, 2020). While offcial cartographic representations of the BRI are scarce, the Chinese government and academic institutes are increasingly relying on satellite imagery – photographs of the Earth taken from space – to support research relating to the BRI, which may also inform future policy interventions. The state’s reticence to publish offcial maps compared to its eagerness to employ remote sensing points to a key difference between cartography and remote sensing. Whereas maps are considered malleable representations, satellite imagery is imagined as objective, neutral, and, importantly, rational – a key word in Chinese narratives of development and modernization (Anagnost, 1995; Yeh, 2009). Yet remote sensing, too, involves representation, inference, and analysis, meaning the practice is inherently political. Despite its passive appearance, observing and monitoring the Earth is never neutral (Alvarez Le� on & Gleason, 2017; Gleason & Hamdan, 2017). Satellite imagery is open to numerous “interpretations and political uses” (Parks, 2001, p. 589), and remote sensing can be used for purposes like war and surveillance (Anson & Cummings, 1991). Such technologies, however, can equally promote more socially and environmentally equitable fu- tures. If the crux of critical research hinges on contesting hegemonic representations (Nicholas, 2006), then more critical remote sensing should recognize the politics and positionalities involved in satellite imagery, expose rather than conform to teleological metanarratives of development, and engage with qualitative methods. As Plummer and E-mail address: mbennett@hku.hk. Contents lists available at ScienceDirect Political Geography journal homepage: http://www.elsevier.com/locate/polgeo https://doi.org/10.1016/j.polgeo.2019.102127 Received 11 December 2018; Received in revised form 26 November 2019; Accepted 2 December 2019