Around the world, samples of air are regularly collected in flasks and shipped from such far-flung locations as Mongolia and Hawaii to the National Oceanic and Atmospheric Administration in Boulder, where they're analyzed for the presence and concentration of dozens of gases.

The record that's been created over the decades from the air samples -- along with measurements taken at a handful of observatories, including ones in Alaska and at the South Pole, and from a "tall tower" network, which includes a site in Erie -- have been instrumental in helping scientists understand how the atmosphere is changing.

The baseline data, which is publicly available, has allowed researchers to show that carbon dioxide concentrations are increasing and, in the 1980s, that a hole had formed in the ozone layer, among other discoveries.

Now, that atmospheric record is in danger -- according to more than 50 researchers who signed a letter published last week in the journal Science -- due to cuts to the Boulder-based Global Monitoring Division of NOAA's Earth System Research Laboratory.

"Despite the growing importance of greenhouse gas observations to humanity, substantial budget cuts at NOAA have resulted in curtailment of our ability to observe and understand changes to the global carbon cycle," the letter reads. "... As scientists, we believe that preserving the continuity of these vital time series must remain a priority for U.S. carbon cycle research."

According to the letter, a dozen flask-sampling sites already have been removed, some new tall towers that were planned to be built will be shelved, and some existing tall towers might be shut down. In an analysis released earlier this year, the American Association for the Advancement of Science said that the budget for NOAA's Oceanic and Atmospheric Research programs has been cut almost 15% between fiscal year 2010 and fiscal year 2012.

Scott Lehman, a professor at the University of Colorado's Institute for Arctic and Alpine Research, relies on data from the air sampling program in his own research, and he's concerned about the future of the record. He also points out that the cost of taking the measurements is relatively inexpensive, especially compared to their benefit.

"I've been a climate scientist for 30 years and provided expert testimony on climate change to the U.S. Senate and Congress in which NOAA measurements featured prominently," he said. 

"Regrettably, many of us have grown accustomed to policy inaction on climate change, but the targeted cuts to NOAA's research arm and their monitoring capabilities in the last few years is something new and unprecedented."

Disrupting the air sampling program erodes the primary source used for a wide variety of research, including the calibration and testing of climate models, said Pieter Tans, who leads the Carbon Cycle Greenhouse Gases Group in NOAA's Global Monitoring Division.

Climate models often differ in what they predict, and one of the most basic ways to test whether a model is useful is whether it can actually recreate conditions we see now.

"We want to know, 'How do they predict the present and the recent past?'" Tans said. "If they don't do that -- forget it. They have no credibility."

The models now also are predicting conditions in future decades. The data collected by NOAA going forward will allow researchers to see how well the models did making those predictions and tweak them as necessary.

"We're working for the future, really, to create base measurements that are going to be used to see how well we understand things," Tans said.