In 1991, eight people entered Biosphere 2, a vast glass enclosure in the Arizona desert designed to test whether humans could live inside a materially closed ecosystem. Conceived by Space Biospheres Ventures as both an experiment in ecological engineering and a possible prototype for long-duration space habitation, the complex brought rainforest, savanna, desert, ocean, agriculture, laboratories, and living quarters under one envelope. What happened inside over the following two years was an unusually revealing account of how ecological, architectural, technological, and human systems become entangled when there is nowhere else to go.
exterior view of the University of Arizona’s 40-acre Biosphere 2 campus in Oracle, Arizona | all images courtesy of University of Arizona
Biosphere 2 used ‘lungs’ to regulate its enclosed atmosphere
Biosphere 2 emerged from an optimistic period in which ecological thinking, counterculture, systems theory, and space exploration increasingly overlapped. John P. Allen, working with Texas businessman and philanthropist Edward P. Bass, developed the project around the idea that a contained ecological system could be studied at a scale large enough to include people. The ambition was simultaneously terrestrial and extraterrestrial: to understand Earth’s cycles by reproducing them in miniature while exploring whether similar principles could eventually support human life on the Moon or Mars.
That proposition made the building itself part of the experiment. Constructed in Oracle, Arizona, between 1987 and 1991, Biosphere 2 was housed within a steel-and-glass envelope that looked like a greenhouse filled with plants, enclosing approximately 200,000 cubic meters of atmosphere, while a stainless-steel liner beneath the site was intended to prevent uncontrolled exchanges with the surrounding soil and water. Extensive piping, pumps, heat exchangers, air systems, and water-recycling infrastructure formed a hidden technosphere beneath the visible landscape.
Because the sealed structure heated and cooled under the Arizona sun, its internal atmosphere expanded and contracted. Engineers therefore installed two enormous variable-volume chambers known as the ‘lungs.’ Flexible diaphragms moved in response to changes in air volume, helping regulate pressure without simply venting the atmosphere to the outside. The system is an important reminder that Biosphere 2 was as much a machine for managing atmospheric physics as it was a miniature world.
The project attempted to compress ecological diversity into a relatively small footprint. A tropical rainforest occupied a high glass volume, while savanna, mangrove, ocean and reef, and fog desert environments formed a sequence of contrasting habitats. An agricultural zone was intended to provide much of the food for the crew, while residential spaces and laboratories were integrated into the same closed environment. The underlying strategy was to create enough biological complexity that the ecosystem could begin to organize itself. But the experiment quickly demonstrated that biodiversity cannot just be assembled from a list of species and expected to reproduce the relationships of a mature ecosystem.
When the atmosphere stopped behaving
After the first crew entered the enclosure on September 26th, 1991, oxygen concentrations began falling. By January 1993, atmospheric oxygen had dropped from approximately 20.9 percent to 14.5 percent. The conditions placed considerable physiological stress on the crew and eventually required oxygen to be introduced into the supposedly closed system. The problem was not simply that the plants were producing too little oxygen but that the artificial ecosystem contained a vast reservoir of organic matter in its soils. Microorganisms consumed this carbon through respiration, using oxygen and producing carbon dioxide. Photosynthesis was expected to compensate for that exchange, but the balance was disrupted by reduced light levels inside the structure and unusually cloudy conditions during the first winter.
The use of concrete for the building was another variable that had not been adequately accounted for. Carbon dioxide produced by microbial respiration reacted with calcium compounds in the structure’s concrete, effectively removing part of the gas from the atmosphere and resulting in a peculiar imbalance in which oxygen continued to disappear while much of the corresponding carbon dioxide failed to remain in the air.
Biosphere 2’s rainforest biome recreates a dense tropical environment beneath the glass enclosure
The ecosystem was not a miniature Earth
Other failures were biological. Pollinating insects and several introduced vertebrate species disappeared, while opportunistic species such as ants and cockroaches flourished. Aggressive vines spread through the rainforest and crop production became difficult as pollination, pests, disease, weather, and available light interacted in ways that were difficult to predict from individual ecological components.
The agricultural system nevertheless supplied a substantial proportion of the crew’s food. But the shortfall contributed to the human body adapting to scarcity, another experiment unfolding alongside the ecological one.
The crew lived under conditions of confinement, restricted food availability, altered atmospheric composition, intensive work, and continuous observation. Roy Walford, the mission physician, also used the enclosure as an opportunity to study caloric restriction, placing the crew on a nutritionally dense but low-calorie diet. Weight loss, fatigue, hunger, and physiological adaptation became part of the experiment’s lived reality.
This is where Biosphere 2 becomes more interesting than a simple story of an engineering project that did not work.
The experiment was trying to reproduce an ecosystem, but it also had to reproduce the conditions under which humans could function inside one. That meant that governance, psychology, nutrition, maintenance, labor, and interpersonal relationships became environmental variables.
Biosphere 2 is set against the Santa Catalina Mountains in Oracle, Arizona
The human system proved impossible to seal
As the mission continued, the eight biospherians became divided over how strictly the rules of the experiment should be maintained. One position prioritized material closure and experimental purity; another favored interventions when human health or operational continuity was threatened. The distinction was philosophical as much as practical. If oxygen had to be added, food imported, or mechanical systems used to compensate for ecological shortcomings, was the enclosure still a closed system?
A supposedly autonomous world required constant human care, while the humans inside it were themselves increasingly dependent on technological intervention. The social consequences became inseparable from the ecological ones. Confinement, hunger, physical exhaustion, atmospheric stress, competing interpretations of the mission, and intense external scrutiny contributed to fractures among the crew. The experiment revealed that a closed habitat is a problem of collective governance.
The second mission, begun in 1994, ended amid management disputes and a breach of the enclosure. By then, the original proposition of a fully closed human world had become increasingly difficult to sustain.
the main glass structure encloses Biosphere 2’s rainforest biome
From failed utopia to scientific instrument
The later history of Biosphere 2 may be the most useful part of the story. Under Columbia University and, subsequently, the University of Arizona, the facility was transformed from a prototype for closed human habitation into an open, highly controlled laboratory for Earth-system science. Researchers used its biomes to investigate questions around elevated carbon dioxide, ocean acidification, plant respiration, drought, and ecosystem response. Later, the Landscape Evolution Observatory introduced another form of experimental landscape through three artificial hillslopes designed to observe hydrological, geological, and biological processes as they developed over time.
Biosphere 2 began by attempting to construct a self-sufficient world and ended by becoming a place for studying the systems that make Earth’s world difficult to reproduce. The original experiment showed how many variables are hidden inside apparently stable environments and how quickly those variables become visible when planetary buffers are removed.
Biosphere 2 belongs to a lineage of utopian projects that treat buildings as instruments for producing alternative ways of living. Its glass envelope, artificial landscapes, pressure-regulating lungs, buried infrastructure, agricultural systems, and controlled atmosphere turned architecture into an operational ecosystem. But the experiment’s most enduring insight may be more modest.
A future habitat on Mars, the Moon, or elsewhere cannot be designed as a perfectly autonomous machine that simply keeps its occupants alive. It has to accommodate maintenance, failure, adaptation, unequal dependencies, ecological unpredictability, and the people operating it. The more completely a habitat is sealed from its surroundings, the more consequential every small imbalance becomes. Biosphere 2 therefore remains compelling not because it achieved its utopian promise, but because it made the promise testable.
Its glass walls separated eight people from the Arizona desert, but they could not separate biology from architecture, technology from ecology, or environmental stress from human behavior.
a network of enclosed environments
triangular glass panels cover the structure
the project reads as an artificial landscape
the glass structures form a sequence of distinct volumes across the site
the scale of the experimental complex becomes apparent against the surrounding desert terrain
dense vegetation climbs around the path inside the rainforest biome
a landscape that feels closer to a tropical forest than the Arizona desert outside
This article is part of designboom’s MODES OF HABITATION chapter, exploring how the spaces we inhabit can reshape the ways we live together. Discover more stories rethinking domesticity, shared living, and the many forms a home can take here.
The post biosphere 2: what happens when humans try to recreate earth? appeared first on designboom | architecture & design magazine.