Every three months, an ocean in the Arizona desert took a trip through time. In 1996, Columbia University researchers used Biosphere 2’s marine mesocosm to recreate three sets of ocean conditions: those of a glacial ice age, the present day, and the 22nd century. According to the University of Arizona’s Biosphere 2 ecosystems page, the researchers alternated between these conditions at three-month intervals over three years. The experiment examined how changing carbon dioxide levels and seawater chemistry affected photosynthesis and coral calcification.
The tank behind the experiment is no ordinary aquarium. Originally designed to simulate a Caribbean coral reef, the enclosed ocean system has a surface area of 35 by 20 meters and holds 2.6 million liters of water. Mechanical systems reproduce or replace natural processes, allowing researchers to adjust temperature, water mixing, gas exchange, nutrient concentrations, and carbon dioxide pressure independently. Although the facility models a tropical reef, its water supply did not need to come from the Caribbean. A 1999 paper in Ecological Engineering, The Biosphere 2 coral reef biome, describes an initial mixture of roughly 10% natural seawater from the Southern California coast and 90% local well water combined with commercial sea salt.
The three experimental scenarios differed in carbon dioxide concentrations and seawater chemistry. The glacial scenario used a CO2 partial pressure of 200 microatmospheres and a pH of 8.29. Present-day conditions used 360 microatmospheres and a pH of 8.08, while the 22nd-century scenario used 900 microatmospheres and a pH of 7.89. The declining pH values indicated increasingly acidic seawater. In a 2000 study published in Global Biogeochemical Cycles, Chris Langdon and colleagues reported that their long-term measurements predicted a 40% decline in net coral reef calcification between 1880 and 2065 as seawater carbonate saturation decreased. This finding describes the projected change in net reef calcification over that period, rather than a uniform decline across all three experimental scenarios.
Reasons behind the 40% reduction
{{/usCountry}}The three experimental scenarios differed in carbon dioxide concentrations and seawater chemistry. The glacial scenario used a CO2 partial pressure of 200 microatmospheres and a pH of 8.29. Present-day conditions used 360 microatmospheres and a pH of 8.08, while the 22nd-century scenario used 900 microatmospheres and a pH of 7.89. The declining pH values indicated increasingly acidic seawater. In a 2000 study published in Global Biogeochemical Cycles, Chris Langdon and colleagues reported that their long-term measurements predicted a 40% decline in net coral reef calcification between 1880 and 2065 as seawater carbonate saturation decreased. This finding describes the projected change in net reef calcification over that period, rather than a uniform decline across all three experimental scenarios.
Reasons behind the 40% reduction
{{/usCountry}}Deep Dive
What did the 1996 Biosphere 2 coral experiment find about future CO2 levels?
Why does ocean acidification make it harder for corals to build their skeletons?
Did a real-reef experiment also find that added CO2 reduces calcification?
Calcification is the process through which corals build their hard calcium carbonate skeletons, which form the physical structure of a reef over time. When calcification slows, corals build their skeletons more slowly, potentially making it harder for reefs to maintain or increase their structure as erosion and other processes continue to remove material.
That raises a fair question: can a tank in Arizona stand in for a real reef? The 1999 Ecological Engineering paper describes a fore-reef and lagoon modeled on Caribbean fringing reefs, with 25 coral genera, although algae made up much of the community. The researchers also acknowledged important differences from natural tropical reefs. Community metabolism was about half that of a natural tropical reef flat and more closely resembled that of shallow, high-latitude reef lagoons. Despite these limitations, the authors concluded that the system functioned as a recognizable coral reef community. Their findings also showed that changes in carbonate concentration were associated with corresponding changes in reef calcification, providing a basis for studying how ocean chemistry affects coral growth.
A second life for the ocean
The ocean habitat later deteriorated after a period of neglect. The University of Arizona’s research account explains that algae and cyanobacteria overgrew the substrate, and nearly all the original animals disappeared, leaving about two dozen hardy fish. Since then, researchers have worked to rehabilitate the ecosystem. By February 2026, that effort had reached a major milestone: researchers had introduced 108 coral fragments into the Ocean Habitat for the first time in more than 30 years. The team attached the corals to specially designed structures as part of research into how they respond to higher temperatures. Updated lighting, circulation, and temperature-control systems support the work. Researchers can now use the 2.6-million-liter habitat to investigate coral growth, environmental stress, and possible approaches to reef restoration under controlled conditions.