Turning a New Leaf: When Tree-Rings Meet Water Policy

Forest fires, insect invasions, devastating droughts—the tales trees tell are fascinating. Their cores read like nature’s diary, as every year’s growth becomes a record of the changing world around them.

In the Southwest, the tree-ring tales of Colorado River drought are especially popular. The 1,450 mile-long river provides water for nearly 30 million people, as it stretches across seven states and part of Mexico. Yet, the fate of this vital river is uncertain, as some climate models project severe water decline in the coming decades.

Douglas fir cross section from the Zuni Mountains in New Mexico.
Credit: Henri D. Grissino-Mayer, Department of Geography, University of Tennessee

While tree rings can’t predict the river’s future, they do provide a window into its past, allowing researchers to determine whether the river’s rising and falling is in-line with its normal activity.

Tree rings collectively hold centuries of Colorado River history. When moisture-sensitive trees (usually found on steep slopes in thin, dried-out soil) get plenty of water, they pack on wide growth rings for the year. When there’s a lack of water, the trees grow very little, producing narrow rings—or in some cases, none at all.

Tree rings allow researchers to discover what the river’s streamflow was like in the years before stream gage measurements existed. Along the Colorado, stream gages—instruments used to measure water levels—have only been in place for about 100 years. Researchers choose the gage record they want to extend back, and then pair it with tree-ring samples to determine which trees are best at estimating the river’s water level, or streamflow.

Reconstructing the streamflow allows researchers to see just how common, and extreme, southwestern drought can really be.

That information is vital to water managers, tasked with developing plans to keep water flowing to the masses, even if it’s trickling in. Yet, until the ongoing basin-wide drought, many managers did not see the need to incorporate tree-ring data into their planning. Most accepted the 100-year gage record as long enough to demonstrate the Colorado River’s moodiness.

But minds began to change in 2002—the drought’s third year—when parts of the river hit water levels lower than any in the century-long gage record.

Tree-ring research began getting lots of attention then, said David Meko of The University of Arizona’s Laboratory of Tree-Ring Research (LTTR). “Severe drought brings action and interest,” he said.

It also brings phone calls, said Connie Woodhouse, a LTTR researcher and associate professor of geography and regional development.

Both Meko and Woodhouse, along with other dendrochronologists, worked on the two most current reconstructions of the Colorado River streamflow. Woodhouse was the lead author for the 2006 reconstruction, which traced the flow back to 1490. Meko took the lead for the 2007 reconstruction, which dated to 762 and nearly doubled the previous tree-ring record.

Going back that far revealed events never before seen in previous reconstructions. One drought in the mid-1100s stretched over multiple decades, reaching or surpassing the gage-record average only nine of 57 years. Two other prolonged droughts popped up in the 800s, with one hitting below-average flows 21 out of 24 years.

In the long life of the Colorado, it seems “2002-like years” aren’t anything special.

Spreading the Wood Word

Tree-ring research gives water managers a way to look back in time, and base their planning on a streamflow record that spans thousands of years instead of on an instrumental record that covers only a century.

But, convincing water managers they need to include the tree-ring data in their planning is an ongoing process.

“Many water managers have been operating their systems using the 1950s drought as the worst-case-scenario. The recent drought was worst in some areas, calling into question the assumption that the 20th century gage record contained the full range of variability in drought extremes,” said Woodhouse.

In May 2005, a conference held in Tucson, Ariz., launched a series of technical workshops hosted by Woodhouse and Jeff Lukas, of the University of Colorado’s INSTAAR Dendrochronology Lab, along with colleagues from the UA and the National Oceanic and Atmospheric Administration.

Since spring of 2006, Woodhouse and Lukas have hosted nine workshops across five states. More than 100 participants from a dozen water agencies have attended to learn how streamflow reconstructions are created and how they can be applied to planning and management.

The goal is to let people see how it’s done and why it works, Lukas said.

“We want to dispel the ‘voodooness’ ” some people associate with tree rings, he said.

So far, agencies in Colorado, Arizona, and California have employed tree-ring records to do things like broaden the scope of their planning models, or determine how often simultaneous droughts occur in different, necessary water sources.

Getting others to follow suit may be easier now that the U.S. Bureau of Reclamation offices in the Colorado River Basin have also embraced tree rings. Reclamation, the region’s largest water supplier, is a large federal agency that oversees water management in 17 western states, including the operation of the Colorado River reservoirs.

“People watch what they do,” Lukas said. “I hope it has ripple effects.”

After the current drought caused the reservoirs – Lakes Mead and Powell – to drop from nearly full to just over half full, the Bureau of Reclamation, at the direction of the Secretary of the Interior, began a process in 2005 to develop guidelines to better manage and operate those key Colorado River reservoirs.

The guidelines would also provide water users and managers—particularly those in Arizona, California, and Nevada—with more certainty as to the amount of future annual water deliveries, particularly under drought and low reservoir conditions.

Because the reservoirs have almost always been operated at normal or surplus levels, such guidelines had never been developed.

Tree-ring research played an important role in the development of the guidelines, said Bob Walsh, a Bureau of Reclamation representative.

The potential impact of the river’s natural variability on water levels has to be considered, Walsh said. “We take climate change seriously; it’s an issue of concern,” he added.

The inclusion of tree-ring research in the bureau’s guidelines is “hugely significant,” Woodhouse said. “In federal agencies, there can be a fair amount of resistance to change, so I especially admire the Bureau of Reclamation for considering the use of tree-ring data,” she said.

The guidelines, called the “Final Environmental Impact Statement for Colorado River Operations under Low Reservoir Conditions,” were completed and approved in December 2007. They went into effect in January 2008 and are designed to extend through 2026.

Branching Out: Tree Rings Join Climate Models

As for the future of water management, it appears that the tree-ring record won’t be enough on its own. Some agencies are looking to account for future global warming, while still using tree-ring data to consider the river’s natural variability.

Denver Water and the Bureau of Reclamation have both commissioned studies to bring the two aspects together, with the goal of achieving a more accurate view of what the future holds for the Colorado River.

Climate change projections have convinced some hydrologists and climate modelers that tree-ring data isn’t applicable to future streamflow at all, Woodhouse said.

While that belief is correct to some extent, the river’s natural variability will not just stop because of human-induced climate change, she added.

“Rather, in addition to the natural variability, we will have the additional regional impacts of global climate change from human activities,” Woodhouse said.

Conceptually, combining the future climate projections with the tree-ring record makes sense, said Lukas who, along with Woodhouse, developed the tree-ring data used in the Denver Water study.

But merging the Global Climate Models (GCM) with the tree-ring-based streamflow record is still an elusive task.

“What I’m not entirely sold on is how to technically combine the two types of data and then characterize it for non-technical audiences,” Lukas said.

Still, he said, “it’s where we have to be thinking.”