The idea that the Earth can be “searched” like a database has circulated for several years in academic and technical circles. Earth Index, developed by the nonprofit Earth Genome, brings that idea into practical use. Earth Index allows users to scan satellite imagery by visual similarity. A user can highlight an example—a patch of deforestation, a mining site, a trawler, or an airstrip—and instruct the system to find comparable patterns elsewhere. The underlying approach relies on “foundation models” trained on vast archives of Earth observation data, enabling the system to recognize features across geography and time. Until recently, such analysis required specialized teams, bespoke models, and significant computing resources. Even well-funded investigations could take months to develop. Tools like Earth Index reduce that burden. In one Mongabay-specific case, our journalists used it to identify previously unreported narcotrafficking airstrips in the Peruvian Amazon, combining automated detection with on-the-ground reporting (Spanish) Earth Index is now available without a waitlist, through an “Open” tier that provides global access and core features to any user. More advanced capabilities—such as higher usage limits, API access, and a more computationally intensive “Deep Search”—sit behind a separate tier, though the developers say they intend to keep access free for high-impact users. This expansion reflects a shift in how geospatial AI is being deployed. Satellite data has been publicly available for decades, yet much of it remains underused because of its complexity. Foundation models alter that dynamic by allowing users to interact with imagery in more intuitive ways, translating raw pixels into searchable patterns. Earth Index’s design emphasizes accessibility. Non-technical users can upload their own reference data, generate training labels, and share results publicly. The system supports iterative workflows in which human judgment refines machine output, enabling datasets that once took months to assemble to be built in days. [caption id="attachment_318313" align="alignnone" width="1200"] Earth Genome worked with Amazon Conservation and Pulitzer Center to build Amazon Mining Watch.[/caption] Early applications suggest a wide range of uses. Investigations have mapped illegal gold mining in the Amazon, quarry operations in the Balkans, watersheds in California’s farming heartland, and methane emissions from cattle. Others have used the tool to catalogue industrial livestock facilities or track changes in land use across regions. Much of the underlying imagery comes from public satellites with moderate resolution and update frequency, making the system less suited to real-time monitoring. Interpretation still depends on user judgment, and false positives remain a risk. The availability of such tools also raises questions about oversight and misuse, even if current data constraints reduce the likelihood of surveillance at a granular level. Its main impact, for now, is that access extends beyond specialists. By lowering the technical threshold, it extends capabilities once confined to governments and specialists to a wider set of actors: journalists, researchers, advocacy groups, and local communities. Whether it improves environmental outcomes will depend on who uses it, and to what ends. Disclosure: Mongabay has a partnership with Earth Genome. Earth Genome did not have editorial influence over this piece.

How do you count and keep track of bats? Advances in videography, including thermal cameras, have made it easier to spot bats. But these animals move fast, travel in big groups and are often found in the dark — making analysis of the data a tough task. Scientists have developed a new software called Thrutracker Analytics that uses traditional computer vision and artificial intelligence to count bats. Watch the latest episode of Then vs Now to learn

Every year, monarch butterflies make their iconic migration across North America. The journey spans thousands of miles and three countries. However, very little is known about this migration, resulting in the lack of concrete data about a very important life stage of these butterflies. Scientists are now using lightweight radio tags to get insights into the mysterious migration of monarch butterflies. Using the technology, they have been able to understand how and where the butterflies move, filling crucial gaps in the data. Watch the latest episode of Then vs Now to learn more.

Biologging trackers have long been used to track and monitor marine animals like whales, sharks and dolphins. But it has been a challenge to use them on stingrays because of their smooth skin and the lack of a prominent fin. Scientists have now developed a multisensor tag which can be securely attached to sting rays. The tags gather a wide array of data including how the animals move, how they interact with other species and move through their habitats. Watch the latest episode of Then vs Now to learn more.

Perched high above the forest floor, the tropical canopy is a reservoir of biodiversity that has long resisted scrutiny. Its inaccessibility has left many of its inhabitants — orchids, epiphytes, ants, monkeys, frogs — poorly studied and poorly protected. But a new study offers a workaround: let the rain do the climbing. Scientists led by Lucie Zinger at the France-based Center for Biodiversity and Environmental Research (CBRE) have shown that water dripping from the canopy carries traces of DNA, or environmental DNA (eDNA), from the organisms above. By capturing and analyzing this “rainwash” in low-tech collectors, they identified hundreds of species across plants, insects, birds, amphibians, and mammals. The plants and animals detected in rainwash of an old-growth forest compared to a plantation plot. The study, conducted in French Guiana, compared samples from a mature Amazonian forest and a nearby tree plantation. The results were striking. Diversity was markedly higher in the undisturbed forest, where passive collectors accumulated eDNA over a 10-day period. Crucially, the rainwash signal was both local and persistent. Even after heavy rain, biodiversity signatures remained spatially distinct at the scale of tens of meters and stable for up to 40 days. Unlike airborne or stream-based eDNA, which can drift and muddle geographic origin, rainwash captures a sharp snapshot of the immediate canopy. It can also be deployed at scale with minimal cost. The researchers propose that this method could become a cornerstone of biodiversity monitoring in tropical forests — habitats that are increasingly threatened and chronically undersurveyed. That’s not to say it’s a panacea. Detection remains limited to species that shed detectable DNA, and to wet seasons when rainfall is sufficient. But in a field stymied by logistical and financial constraints, the ability to “listen” to the canopy through its own runoff is a conceptual advance. In the future, the hum of the rainforest may be traced not through what can be seen and heard, but through what the rain leaves behind. Banner image of a strawberry poison dart frog, by Rhett A. Butler/Mongabay.  

Fifteen teams have advanced to the semifinal round of XPRIZE’s $5 million Autonomous Wildfire Response Track, moving one step closer to proving that autonomous systems can detect and extinguish wildfires within 10 minutes across 1,000 square kilometers of challenging terrain. Part of a four-year, $11 million global competition launched in 2023, the initiative seeks to spur breakthroughs in rapid-response firefighting technology as climate-driven wildfires grow more frequent and destructive. Selected from a global pool, these teams represent a range of institutions, from defense contractors to university research labs and even a high school in California. Each presents a unique solution that blends robotics, artificial intelligence, and wildfire science. “The convergence of exponential technologies such as AI, robotics, drones, and sensors offers us the opportunity to detect wildfires at inception, and put them out in minutes before they spread—that’s the mission of this XPRIZE,” said Peter H. Diamandis, Executive Chairman of XPRIZE, at the time of the competition’s launch in 2023. Many of the semifinalists take a “system-of-systems” approach. AeroWatch, a Spain-based consortium, is integrating components from over a dozen partners to create a unified interface for fire managers. Crossfire, based at the University of Maryland, deploys scout UAVs for surveillance and “Firejumper” drones for suppression. Its system was validated in a live-fire demonstration earlier this year. Others focus on scale and speed. Canada’s FireSwarm Solutions is developing heavy-lift drone swarms capable of operating at night. Germany’s Dryad combines solar-powered sensors with reconnaissance and suppression UAVs to detect fires at the smoldering stage. Meanwhile, Anduril, a U.S. defense tech firm founded by Oculus VR’s Palmer Luckey, is fielding its AI-enabled Lattice OS platform paired with advanced sensor towers and aerial reconnaissance. Student-led teams are also in the mix. Wildfire Quest, from Valley Christian High School in San Jose, has built a solar-powered, self-replenishing drone system with an expandable suppression network. RAINDROPS, a collaboration between Norwegian and American universities, is piloting a low-cost system-of-systems architecture designed for global scalability. This summer, the XPRIZE Wildfire team will visit each semifinalist’s test site to evaluate the systems in live field conditions. These trials will assess technical readiness, safety, connectivity, and resilience in real-world settings, including high heat, wind, and complex terrain. The Alaska Center for Unmanned Aircraft Systems Integration will oversee data capture for judging. The stakes are high. According to XPRIZE, Extreme Wildfire Events now account for over 80% of fire-related damages worldwide. “With over 30 years of experience in fire management, I’ve seen firsthand how devastating wildfires can be,” said Shawna Legarza, former director of fire and aviation at the USDA Forest Service, in the 2023 launch statement. “To better protect our land and ourselves, we need to change the way we detect and manage wildfires now.”