In the first half of the 20th century, measurements of the greenhouse gas CO2 weren't very precise, said Steve Ryan, physical scientist for the National Oceanic and Atmospheric Administration.
Scientists had primitive measuring techniques and weren't aware of the possibility of contamination, he explained. "We breathe it out, plants take it up, combustion produces it. It's just all over the place."
Those early results ranged widely. For example, currently there is a concentration of about 385 parts per million of CO2 in the atmosphere, but early measurements varied from the high-200s to 4-500 ppm. At the time, scientists concluded it might be 100 years or more before they could tell if there was an increase in CO2.
But Charles Keeling of the Scripps institution of Oceanography noticed when taking measurements in the western United States that the CO2 concentration in the atmosphere, during the daytime, would go to a fairly uniform value from month to month, because the air is vertically mixed when the sun comes out, Ryan said.
Keeling also saw the uniform concentration at a higher altitude, away from vegetation and other carbon receptacles, at a place like an island. So in 1958, he set up a station at Mauna Loa to track CO2.
The results, known as the Mauna Loa curve, show a steady increase of CO2. The seasonal variations in CO2 levels make the upward line appear as a jagged staircase.
Over the years, more sampling sites were set up in other countries, islands and oceans on commercial ships to provide a fuller picture of the gases in the Earth's atmosphere. (Map of observation sites around the world)
Instruments called CO2 analyzers at Mauna Loa continuously suck in air and measure the carbon dioxide content 24 hours a day, seven days a week. Since it would be too expensive to install such devices throughout the global CO2 measuring network, the sites, including Mauna Loa, employ another technique called "flask sampling".
Air samples are collected in 5 liter containers once a week and sent for analysis to NOAA's central repository in Boulder, Colo.
Ryan demonstrates CO2 flask sampling:
More and more measurements are now taken at Mauna Loa, such as of other greenhouse gases, the ozone, and particulates in the atmosphere. After the ozone hole was recognized in 1984, and the Montreal Protocol banned certain chemicals that were depleting it, measurements showed a decrease in the banned chemicals and indicated that the ozone layer was starting to recover, Ryan said.
Similarly, the CO2 measurements may show if countries' efforts to reduce CO2 emissions are having a cumulative effect.
Meanwhile, scientists at Mauna Loa are scratching their heads at the behavior of another greenhouse gas: methane. After years of steadily increasing -- much like CO2, methane concentrations began leveling off by the 1990s, much to everyone's surprise, Ryan said.
"It just flattened out and they're still debating what the causes of that are."