In an era accustomed to the overexploitation of the earth’s resources, the impacts of industry activity on ecosystems and communities that rely on them are well understood. However, the extent of these extractive activities now appears to go beyond what protectors of those resources previously thought. With advancing technologies, new research has been able to uncover illegal modes of extraction going unmonitored over land, air, and sea, creating an expansive network of ghost paths that allow resource exploitation to flourish. “Dark vessels”, “ghost roads”, and “clandestine airstrips” are recently identified avenues within this hidden transport system.
Since these modes span over vast areas of the earth, it has historically been difficult to monitor all activity taking place. Using the growing power of AI and widely available, high quality satellite imagery, researchers can now detect clues indicating ghost paths on a larger scale than ever before. Although this field of research is still in its early stages and has only highlighted the prevalence of these activities, recognizing their existence is the first step toward stopping them. Emerging data can be used to inform governments and protection organizations, providing them with a better understanding of the size of the threats they need to fight.
Dark Vessels
A map of fishing activity around the world, showing tracked and untracked vessels. Global Fishing Watch.
A study published this year in Nature by Global Fishing Watch (GFW) used geospatial artificial intelligence (GeoAI) to reveal the massive amounts of unmonitored human activity in the oceans, or “dark fleets,” which support fishing and offshore energy development. Researchers discovered that 75 percent of the world’s industrial fishing vessels aren’t publicly tracked, with nearly a third of transport and energy vessel activities also missing from public tracking systems. Without documentation of human activity on the ocean, it is challenging to control industries that have left a third of fish stocks exploited beyond biologically sustainable levels, and an estimated 30–50% of critical marine habitats lost due to human industrialization.
The primary tool for tracking vessels, the Automatic Identification System (AIS), uses radio signals but can be easily circumvented by ships wishing to remain undetected. To gain better insight, AI technology was employed to analyze 2,000 terabytes of satellite data from the European Space Agency’s Sentinel-1 constellation. This data was then combined with GPS information to identify ships that would otherwise be undetected. A machine learning program further examined this extensive data set to locate undocumented fishing vessels, a result that the director of GFW called “the beginning of a new era in ocean management and transparency.”
Ghost Roads
Mapped roads in northeastern Borneo using a leading road database, OpenStreetMap (left), and with the ghost roads identified in this study added (right). Author
Networks of illegally-built roads, known as ‘ghost roads’, are proliferating across tropical forests, as revealed by a recent study published in Nature. Usually built for the purpose of logging, poaching, or mining, these roads pose a severe threat to these hotspots of biodiversity and are often a precursor to environmental destruction. As these roads do not appear on official maps, they can escape environmental governance, driving deforestation and compromising protected areas. Forests across Indonesia, Malaysia, and Papua New Guinea are under particular threat of destruction from large-scale palm oil operations.
Researchers mapped an area in the Asia-Pacific tropical region to illustrate the abundance of these illegal roads by comparing data from leading global road datasets and roads manually mapped from satellite imagery. In an effort requiring 7,000 hours of work from trained volunteers, the team found that the true density of roads is severely underestimated by official sources. Lacking accurate information on human activities poses challenges to conservation efforts, creating blind spots in zoning and law enforcement and leaving extractive activities left unmonitored. Research has shown that road establishment typically precedes drastic spikes in deforestation.
Applying strict legal protections to the land would seem to be a clear solution, however, even protected areas are not immune to the ghost road phenomenon. The same study found that roads cause similar rates of deforestation inside protected areas as outside of them. Once road density was accounted for, protected land only slightly reduced forest loss. The findings indicate that protected areas can effectively prevent deforestation only if they prevent road construction.
Clandestine Airstrips
Flights from illegal airstrips used for mining in Yanomami territory in Brazil, found using a combination of satellite imagery, crowdsourced databases, and geospatial analysts. The New York Times.
In remote areas that can’t be reached by car or boat, miners and loggers have furthered vast networks of criminal activity by using untracked flights and illegally-built airstrips. A recent investigation featured in The New York Times, a collaboration with The Intercept Brazil and The Rainforest Investigations Network, found that over 1,200 clandestine airstrips have been constructed throughout the Brazilian Amazon to support the illegal mining industry, including 61 within the protected Yanomami Indigenous territory. Mining activity not only accelerates the destruction of the world’s largest rainforest but also endangers the lives of its long-standing inhabitants. These operations, part of a thriving illicit industry, go largely unchecked due to the neglect or ineffectiveness of regulatory agencies in Brazil which lack the resources to effectively curb illegal mining.
Investigations by the federal police and environmental agents depict a chaotic situation in the Amazon’s airspace. Planes and helicopters with revoked permits continue to fly undeterred to illegal mines, often with their transponders turned off and frequently crossing the border in and out of Venezuela. Even when the military or law enforcement shut down an illegal airstrip, mining operations inevitably resume and continue to fly under the radar. To verify the locations of these airstrips and link them to illicit mining, the investigators developed an AI algorithm that could identify features of illegal mining sites and runways. Analyzing thousands of historical satellite images, they documented indicators of nearby mining activities, such as clear-cut rainforest areas and pools used by miners to separate dirt from ore. To help them determine which airfields were unregistered, they cross-referenced their findings with OpenStreetMap, a community-built global map of roads and infrastructure.
Conclusions
While AI offers a promising route for detecting illicit transport networks due to its ability to analyze large quantities of data, there remain limitations to how well emerging technologies can address ghost networks. According to researchers who uncovered the ghost roads in Borneo, the human eye is still better at detecting roads than AI. Efforts are ongoing to develop automated road-monitoring technologies using satellite imagery, which could save the thousands of hours of labor required to confirm the roads manually. A similar challenge exists for monitoring planes flying over the vast expanse of the Amazon forest. Though the New York Times was able to estimate flight paths from observed airstrips, aircraft supporting illegal mining in the Amazon remain difficult to monitor as they can fly at low altitudes to evade radar detection. A thorough analysis of these pathways will require enhanced technologies to keep up with the unpredictable landscape of illegal activity.
Despite geotech’s current limitations, most governments alone do not have the capacity to oversee millions of acres of forest or entire oceans to monitor illegal activities. To keep up with the expansion of ghost networks, AI-based mapping systems could be vital in monitoring them on a global scale. Once these informal networks are mapped, the data can help prove that these activities are taking place, potentially prompting law enforcement or other organizations to take actions to protect resources under threat. Although it can be assumed that illegal activity is linked to increased exploitation, the direct effects of these illegal networks on ecosystems are also not fully understood. Knowing that these activities are occurring could provide avenues for further research on the extent of damage caused by ghost networks. Some of the most dangerous threats to our world’s ecosystems could be the ones we are unaware of.
Sources:
Satellite mapping reveals extensive industrial activity at sea | Nature
Ghost roads and the destruction of Asia-Pacific tropical forests | Nature
The Illegal Airstrips Bringing Toxic Mining to Brazil’s Indigenous Land | Pulitzer Center
Amazônia tem mais pistas de pouso clandestinas que legais
