Case Study 21-2: Planet Labs and Amazon Deforestation — The Planet That Photographs Itself

Case Study 21-2: Planet Labs and Amazon Deforestation — The Planet That Photographs Itself

Background

The Amazon basin contains approximately 10% of all species on Earth. It is the largest tropical rainforest in the world, spanning nine countries, with Brazil containing approximately 60% of its area. The forest functions as a massive carbon sink — absorbing an estimated 2 billion tons of carbon dioxide annually — and as a precipitation engine for agriculture across South America. Its destruction is one of the central drivers of global climate change.

For decades, deforestation in the Amazon was documented largely through ground-level reporting by environmental advocates and journalists — a process that was slow, dangerous, and geographically limited. Satellite imagery changed this. Beginning with Landsat in the 1970s, satellite data provided systematic coverage of forest loss. But Landsat's 16-day revisit rate and limited resolution meant that deforestation events could go undetected for weeks, enabling illegal clearing operations to be completed before monitoring systems registered them.

Planet Labs, founded in 2010 by former NASA scientists, represented a fundamental shift in this equation.

Planet Labs and the Daily Planet

Planet's core innovation was not resolution — its Dove satellites capture imagery at approximately 3 meters per pixel, far coarser than Maxar's high-resolution commercial satellites. It was revisit rate. By launching a constellation of more than 200 small satellites in coordinated orbital planes, Planet achieved something no satellite operator had before: the ability to photograph every point on Earth's landmass every single day.

For deforestation monitoring, daily revisit is transformative. Illegal land clearing in the Amazon typically proceeds in a recognizable pattern: a narrow road cut into the forest, then progressive clearing outward from the road, then burning of cleared vegetation to prepare land for cattle ranching or agriculture. At 3-meter resolution, each stage of this process is clearly visible. At daily revisit, each stage can be detected within 24 hours of occurrence.

Global Forest Watch and Near-Real-Time Alerts

The World Resources Institute's Global Forest Watch (GFW) platform combines Planet data with Landsat imagery and algorithmic analysis to produce deforestation alerts. The GLAD (Global Land Analysis and Discovery) system, developed at the University of Maryland, analyzes Planet and Landsat imagery to detect forest loss and generate alerts — typically within one to two weeks of the clearing event, reduced from months or years under pre-satellite monitoring regimes.

These alerts are publicly accessible through the Global Forest Watch web platform. Government agencies, NGOs, journalists, and local indigenous communities can set up automatic notifications when deforestation is detected in a specific area. An indigenous community whose territory is being cleared illegally can receive an alert on a smartphone within days of the first clearing activity.

The Brazilian Context: Politics and Satellite Accountability

Brazil's National Institute for Space Research (INPE) operates its own deforestation monitoring system, PRODES, which uses Landsat imagery to produce annual deforestation estimates. For decades, PRODES data was the authoritative record of Amazon deforestation rates — and it showed significant deforestation throughout the 1990s and 2000s, with a notable decline following the implementation of stronger enforcement policies after 2004.

The arrival of the Bolsonaro government in 2019 brought a radical shift in Brazilian forest policy. Enforcement of illegal deforestation laws was dramatically reduced, and forest loss accelerated sharply. Deforestation in the Brazilian Amazon increased by 85% in 2019 compared to 2018. When INPE's own data documented this increase, the agency's director, Ricardo Galvão, was fired — a striking instance of a government attempting to suppress its own monitoring system's findings.

The attempt failed, in part because Planet Labs' satellite data provided an independent, internationally accessible record that could not be deleted or suppressed. Environmental organizations, journalists, and foreign governments all had access to the same imagery. The Brazilian government's political decision to reduce enforcement could be documented and published globally, regardless of what happened to INPE's institutional independence.

The Limits of Satellite Accountability

But the story does not end with clean accountability. Despite overwhelming satellite documentation of accelerating deforestation under the Bolsonaro government, the deforestation continued. International pressure — including threats of trade sanctions from European nations — had limited effect. The Brazilian government was not deterred by the existence of satellite documentation.

This illustrates a limit of surveillance accountability: documentation is necessary but not sufficient. Satellite imagery can prove that deforestation is occurring. It cannot make governments enforce their own laws. The accountability chain has multiple links — detection, documentation, reporting, political will, enforcement — and satellite imagery can only address the first two.

Discussion Point: The chapter argues that visibility does not guarantee accountability. The Amazon case demonstrates this: illegal deforestation is highly visible from space, comprehensively documented, and widely reported — and it continues. What does this tell us about the relationship between surveillance and power? Who must be watching for surveillance to produce accountability?

Indigenous Surveillance: Community-Level Monitoring

One of the most significant developments in Amazon deforestation monitoring is not institutional but grassroots. Indigenous communities — who have the strongest stake in forest protection, having lived in and managed these landscapes for centuries — are increasingly using satellite tools themselves.

Digital Democracy is one NGO that has worked with indigenous communities in the Amazon and elsewhere to integrate satellite imagery into community monitoring. The organization helped the Waorani people of Ecuador use satellite imagery alongside GPS tracking to document the boundaries of their territory and the movements of oil company vehicles near their land.

RAISG (Amazon Network of Georeferenced Socio-Environmental Information) combines satellite monitoring with on-the-ground data from indigenous and local community monitors to produce comprehensive maps of pressures on Amazonian indigenous territories.

This community-level appropriation of surveillance technology represents the synoptic gaze — many watching the few — rather than the panoptic gaze. It inverts the typical power relationship, putting satellite-derived information in the hands of the communities who are most directly affected by the activity being monitored.

The effectiveness of this inversion is real but limited. Indigenous communities with satellite tools can document illegal intrusions. They cannot, on their own, enforce laws or remove illegal actors from their territories. The surveillance provides evidence; the power to act on that evidence remains with state institutions.

Technical and Ethical Dimensions

The Resolution Floor and Its Implications

Planet Labs' 3-meter resolution is sufficient for deforestation monitoring but not for individual identification. A person walking in a cleared area is invisible at 3 meters per pixel. This resolution floor has important privacy implications: the surveillance system can track landscape change without tracking individuals. This makes it qualitatively different from higher-resolution commercial imagery that begins to approach individual detection.

As resolution improves — and commercial satellites are trending toward sub-meter resolution — this distinction will erode. The same infrastructure that documents deforestation at 3 meters per pixel will, at 0.25 meters per pixel, also document individual human presence in those landscapes. The transition from environmental surveillance to human surveillance is not a categorical shift but a resolution threshold.

Commercial Incentives and Mission Alignment

Planet Labs presents itself as a company with an environmental mission — "to image all of Earth's landmass every day, and make global change visible, accessible and actionable." This mission framing implies that the primary value of Planet's data is environmental monitoring and accountability.

But Planet is a commercial company. Its largest revenue comes not from environmental NGOs but from government contracts — including U.S. government intelligence agency contracts that are not publicly disclosed in detail. The same daily global imagery that documents deforestation is also available for the intelligence applications described in the chapter's main text.

This is not a critique unique to Planet — it applies to any dual-use technology company. But it matters for evaluating Planet's "mission" framing: the environmental applications are genuine and valuable, and they are not the full picture of how the imagery is used.

Discussion Questions

Brazil's government could not suppress its own satellite data because international commercial operators provided an independent record. Does this suggest that the privatization of satellite imagery has been, on balance, positive for environmental accountability? What are the strongest arguments for and against this position?
Indigenous communities using satellite tools to monitor their own territories represent a form of "surveillance of the powerful by the relatively powerless." What are the structural barriers these communities face in translating satellite evidence into effective protection? What would need to change institutionally for satellite evidence to produce stronger protection?
The chapter distinguishes between Planet Labs' 3-meter resolution (sufficient for landscape monitoring, insufficient for individual identification) and higher-resolution commercial imagery (approaching individual detection). At what resolution threshold should additional privacy regulations attach? Who should make that determination?
Planet Labs holds imagery of the entire Earth's landmass going back more than a decade. This archive has enormous value for tracking historical change. It also represents a permanent record of human activity on every part of Earth's surface. Who should own this archive? Who should have access to it? What governance mechanisms would you propose?
Deforestation continued in the Brazilian Amazon even when satellite documentation was comprehensive, internationally accessible, and widely reported. What does this tell us about the theory of accountability through transparency? Are there circumstances in which you believe satellite surveillance would be effective in changing the behavior of a state actor? What factors determine effectiveness?

This case study connects to Chapter 22 (environmental monitoring), Chapter 33 (activist uses of surveillance technologies), and Chapter 38 (AI-assisted analysis of satellite imagery).