Attributing the Causes of Climate Change

Blake Thomas looked across the table from under his dark, tangled hair and crossed his arms. He was trying to explain how, when faced with a mountain of data that appeared to establish a link between changes in western river flows and human-induced climate change, he could still be uncertain of the connection.

“It is obvious that humans have caused changes in air temperature,” he said. “But stream flow is complex.” Thomas, a hydrologist with the United States Geological Survey in Tucson, Arizona, shares a sense of caution that is common among those studying climate.

The San Pedro River one and a half miles north of the Charleston Bridge (looking north).
Credit: Sandy Anderson, Gray Hawk Nature Center

The link between human activity and the observed global rise in temperatures is widely accepted for several reasons:

• Scientists have identified an increase in global carbon dioxide levels that coincides with the beginning of the industrial revolution.
• The ability that these chemicals have for trapping heat in the atmosphere has been reproduced in laboratories.
• Ice core samples taken from Antarctica allow scientists to view the atmosphere’s content back as far as 890,000 years ago.
• The record has allowed scientists to recognize that the change in carbon dioxide levels since the 18th Century is unprecedented over the last million years.

So, what makes linking changes in stream flow—or any other natural occurrence—to climate change so much more difficult?

In his research on trends of stream flow in the Southwest, Thomas has been constantly reminded of two things: there are numerous possible means to any single end, and stream flow data only dates back about 100 years.

To confidently determine if a change is due to the natural order of things—climate variability—or is caused by humans—anthropogenic climate change—scientists may need more than 10 decades of data, Thomas said.

“It’s nice to have 100 year records, but sometimes you don’t,” said Dan Cayan, a researcher with the Climate Research Division of Scripps Institute of Oceanography.

Cayan was part of a group headed by Tim Barnett, also of the Scripps Institute of Oceanography, that published a study of human-induced changes in the climate and waterways of the West in February 2008. He has few hang-ups about linking his observations to climate change.

Barnett and the team compared the climate trends they observed over a 50-year period with downscaled climate models that predicted the precipitation and temperature levels of the West with and without the effects of greenhouse gases. They found that the actual data looked eerily similar to models that included the effects of greenhouse gases, and thus were able to conclude that up to 60 percent of the climate-related trends they observed were human-induced.

In the scientific world, this is called a detection and attribution study—detection meaning that the changes they observed were different from what they would normally expect in nature, and attribution meaning that those changes were consistent with what they would expect of human-induced trends.

Call off the search, the numbers are in. Right?

Figure 1. Regional streamflow trends in Arizona and New Mexico, 1950 - 2002.
| Enlarge This Figure |
Credit: Adapted from 2006 USGS report by Joe Abraham, CLIMAS, The University of Arizona.

“There is still a lot to look at,” said Julio Betancourt, a researcher with the U.S. Geological Survey. The Barnett study is compelling, he said, but it does not silence all doubts.

Downscaling models, the process by which scientists focus a model on a smaller, more specific geographic or chronologic space, is not failsafe, Betancourt said. There is more than one method of downscaling, and the results don’t always click.

Thomas pointed out that, although the group observed clear trends in river flow in the West and Pacific Northwest, these trends did not hold up in Arizona and New Mexico.

He pushed a bound copy of his study on stream flow trends in southeastern Arizona and southwestern New Mexico across the table.

“If you look at regional streams in Arizona, there is no consistent pattern,” he said. “If it’s human-caused, you would expect regional trends.”

Digging deeper into the patterns of stream flow unearths even more variables.

Changes in vegetation, human regulation, and even fires, which burn vegetation and are capable of changing the chemical properties of the soil, can all affect the way rain and snow flow into waterways, Thomas said.

The scientific community will continue to study links between environmental deviations and human-induced climate change. As Thomas and others have pointed out, however, there are still questions to be resolved.

“In the long run, nature’s going to supply the answers,” Cayan said.

Settling back in his chair, Thomas was in agreement with his colleague.