Media365
Deep in the Arizona desert near Oracle sits one of the strangest experiments ever built, a sealed, 3.14-acre glass and steel enclosure called Biosphere 2. Constructed between 1987 and 1991 at a cost of roughly $150 million, funded largely by Texas billionaire Edward Bass, it was designed to test whether humans could live inside a fully self-sustaining artificial ecosystem, complete with its own ocean, rainforest, desert, savannah and farmland. On 26 September 1991, four men and four women, known as biospherians, sealed themselves inside for what was meant to be a two-year mission. What followed became one of the most closely watched, and ultimately most controversial, science experiments of the decade.
Why Biosphere 2 was built in the first place
The facility took its name from the idea that Earth itself is Biosphere 1, and this artificial version was meant to model it in miniature. According to reporting on the University of Arizona’s official description of the project, the enclosure was built as a materially closed ecological system, meaning almost nothing was meant to pass between the inside and outside world except sunlight. Beyond its scientific goals, the project was also framed as an early test of the kind of closed life support systems that future space colonies on the moon or Mars might eventually require, with the crew farming their own food and recycling their own water, waste and air.
How oxygen quietly began disappearing from the sealed atmosphere
Roughly three months into the mission, the crew began noticing a troubling pattern in their regular air quality tests: oxygen levels that had started at a normal 20.9 per cent were steadily falling. According to the peer-reviewed paper titled Oxygen loss in Biosphere 2, published in Eos, Transactions American Geophysical Union, by researcher Jeffrey Severinghaus and colleagues including Wallace Broecker of Columbia University’s Lamont Doherty Earth Observatory, oxygen levels dropped from 21 per cent to just 14 per cent over the first sixteen months of closure, a decline serious enough to begin causing health problems for the crew inside.
The chemistry that explained the missing oxygen
The mystery deepened because the drop in oxygen did not match up with a corresponding rise in carbon dioxide, the gas normally released whenever oxygen is consumed through respiration. Severinghaus and his co-authors identified the cause: microbes in the facility’s unusually rich soil were consuming oxygen at a much faster rate than expected, but crucially, the carbon dioxide they produced was reacting with calcium hydroxide in the building’s exposed concrete to form calcium carbonate, a process called carbonation. This meant carbon was effectively being locked away into the walls themselves, masking the true scale of the oxygen loss and making the problem far harder to diagnose in real time.
A crew breathing air equivalent to a Himalayan peak
By January 1993, conditions inside had become severe enough that outside intervention was required. Mark Nelson, one of the original eight crew members, later wrote for the University of Arizona Press that the crew had been living with oxygen levels around 14 per cent, equivalent to being at roughly 15,000 feet of elevation, despite the facility itself sitting at only around 3,900 feet in southern Arizona. Nelson described the moment Mission Control finally pumped supplemental oxygen into the sealed structure, raising internal levels to 26 per cent, higher than Earth’s normal atmosphere, and recalled feeling decades younger within minutes, along with hearing the sound of crew members running for the first time in many months.
Why the intervention mattered for the project’s scientific standing
The need for outside oxygen fundamentally changed how the mission was judged. As a demonstration of a truly self-sustaining closed habitat, the first mission did not succeed on its own terms, since a system that requires external resupply to keep its occupants safe cannot be considered fully closed. At the same time, the episode produced a genuinely valuable scientific finding: the interaction between rich soil, microbial respiration and curing concrete had not been anticipated at the design stage, and diagnosing it added real, lasting insight into how closed ecological systems actually behave under sustained real-world conditions.
Personal conflict overshadowed the underlying science
Beyond the oxygen crisis, tension within the crew itself became a major part of the story. According to NPR’s coverage of the project, disagreements among the eight biospherians over how the project should be managed became international headline material, often overshadowing the underlying science entirely. By the time the crew emerged in September 1993 after completing their full two-year mission, the original group had split into two factions that reportedly were no longer on speaking terms with each other, following stress from cramped shared quarters, food shortages caused by lower-than-expected crop yields, and disputes over the project’s direction.
What happened to Biosphere 2 after the first mission ended
A second, shorter mission began in 1994 but lasted only about six months before being cut short. Ownership of the facility subsequently changed hands several times, and Columbia University took over managing its research programme for a period in the mid to late 1990s before eventually stepping back. In 2007, the facility was donated to the University of Arizona, which continues to operate it today as an active Earth system science research centre, now employing dozens of staff, including researchers studying rainforest dynamics, ocean acidification and climate change using the same sealed biomes originally built for the Biospheres.
A troubled experiment that still shapes science today
Despite its rocky beginnings and lingering reputation as a 1990s punchline, complete with tell-all books and a Pauly Shore film loosely inspired by the project, Biosphere 2’s core scientific legacy has held up in ways its early critics may not have anticipated. The facility demonstrated that life support in a closed system cannot be neatly separated into biological and mechanical components, since the ecosystem, the machinery meant to support it, and even the building materials themselves all interact in ways that are difficult to predict in advance. More than three decades later, that hard-won lesson continues to inform research into life support systems for future long-duration space missions, even as the story of the crew’s two years sealed inside a shrinking supply of breathable air remains the detail most people still remember first.
