"We may not talk to trees," begins a plaque at an entrance to the University of Arizona Laboratory for Tree-Ring Research on the west end of Arizona Stadium, "but they talk to us."
Slabs of wood can be found wherever the eye wanders; the aroma is reminiscent of a carpentry shop.
"The basic idea," explained Thomas Swetnam, the director of the lab, "is that if it's wet, rainy, the trees will grow fast and you get a big ring, a larger annual growth ring. If it's dry and lack of rainfall or lack of snowfall in the winter, the trees are stressed and they grow very little, and so the tree rings are very narrow." Trees growing in dry landscapes have ring widths that closely correspond to the rainfall.
The lab's founder, Andrew Douglass, used the principle to date the ancient ruins of the Anasazi and cliff dwellings at Mesa Verde in Colorado and Chaco Canyon in New Mexico. Douglass' successor, Edmund Schulman, took to the headwaters of the Colorado River to develop a water flow chronology.
"The basic idea there," said Swetnam, "is you collect samples from living trees and dead trees in the upper watershed of the Colorado River – in the mountains of Colorado, of Western Colorado over into Utah. You collect many sites around the headwaters of the Colorado River, and you measure the tree rings from lots of places and combine those measurements into one long time series. What you find is that the relationships between the ring widths and those sites are correlated with the gauged records of water flow.
"You calibrate that instrumental record with the ring-width record for the period that they have in common," Swetnam said, then you can use the ring widths to go back as far as you have the tree-ring chronology.
In 1922, officials of seven Western states including Arizona and California signed the Colorado River Compact, which apportioned the water from the river among the seven states.
All seemed well until Chuck Stockton, a professor at the Tree-Ring Lab, published a paper in the 1970s that showed the agreement based the allocation from the river on a period in the early 1900s that was determined to be "the wettest period in something like 500 years," Swetnam said. Stockton was able to chart water flow in the Colorado River back to the 1520s.
Enter David Meko, an associate research professor at the lab, who had been a student of Stockton. The California Department of Water Resources, concerned about a drought in 2004, wanted Meko to extend the chronology even earlier.
"They asked me what we could to improve that information, and the one thing that I thought was time extension," Meko said, meaning to go earlier than 1500. "We know from other tree-ring work in the West, from spotty locations, that the really old tree-ring records show big climate fluctuations in the medieval period. So I said we could search for wood. The only way we could get back before the medieval period with drought-sensitive trees that are good for stream flow reconstruction is to find remnant wood" – logs, stumps or dead wood lying on the ground. They funded a year-and-a-half study to take a trip to "some of the classic tree-ring sites."
They headed in three-man crews to the headwaters of the upper Colorado River basin – six or seven spots in the high Rockies of Colorado, and in Utah. "We're using the trees as if they're surrogate rain gauges up in the mountains," Meko said.
Connie Woodhouse, an associate professor of geography, happened to be doing a study for the U.S. Geological Survey in Colorado to develop tree-ring chronologies relating to lakes, so they were able to divide the sites.
Meko had hoped to get past 1300 AD, but couldn't promise that he would. "But as it turned out," he said, "we found a good set of sites with wood that went back to 1200, and the oldest ones going way back before that."
What they documented, once the tree-ring samples were brought back to the lab and evaluated, was sobering – a drought of about six decades in the 12th century, "almost the whole lifetime of a person without having a wet year," he said.
"This was just a really unusual period in the mid-1100s, mainly for not having wet years. This gap of not having individual wet years – for around 60 years there was no real wet year – that type of scenario really has the water managers upset. We really depend on those single really wet years to fill up reservoirs."
The storage capacity of the reservoirs on the Colorado is about four years, Meko said. "We can't have just normal conditions to refill the reservoirs," he said. The worry is that with climate change there won't be the wet year that recharges the reservoirs.
Sharon Megdal, director of the UA's Water Resources Research Center, said water managers and analysts are working to deal with the implications of the tree-ring discoveries.
"Water managers have known for some time that the Colorado River allocations were based on a particularly wet period of time," Megdal said. "For some time, the Arizona Water Banking Authority, a state agency established in the mid-1990s, has been storing water for times of drought on the Colorado River. What people are working on is improving the modeling of the Colorado River to better take into account climate variability and climate change. All of this will necessarily require analysis of alternative scenarios."