Will We Have Enoughby Garnet Bass
It takes only a few inches of water to float a kayak. That’s one reason I like them. I can scoot back into shallow coves and slide over downed trees visible just below the surface. Yet there I sat, stuck on a mudflat in the middle of Falls Lake, the primary water source for Raleigh, NC. In 2005, for the second time in 4 years, summertime drought had dropped the lake level more than 7 feet, leading the city to impose conservation measures—and leaving me high and, if not quite dry, exceedingly muddy. Was this a portent of things to come?
Fifteen miles south of where I sat, a team with the SRS Southern Global Change Program has been studying that very question. Led by project leader Steve McNulty and research hydrologist Ge Sun, they’re creating computer models that will help local officials across the South understand the potential for water shortages in the coming decades.
“If you realize that time after time you’re going to have these severe water shortages, now is the time to start measures to try to prevent them—whether it’s developing networks to move water from other areas, building reservoir systems, or starting water conservation measures to try to reduce the water demand during those stress periods,” says McNulty. “Those are options municipalities can have open, if they’re given enough time to prepare for it. Part of what we do here is to give them that time by developing models that accurately predict what will happen in the future.”
The models they’ve created cover almost 700 watersheds and take into account the effects of climate change, population growth, land use, and vegetation patterns. The models also show the effects of ground water depletion and water use by key sectors such as agriculture and thermoelectric power plants.
By the end of the year, the team hopes to have a Web-based program up and running that will allow local planners and policymakers to run their own what if scenarios. Officials will be able to see not only the probability of drought, but the effects of potential responses; for example, whether a new reservoir would reduce water shortages from 6 years out of 10 to 2 years out of 10. The team is also working to expand the model nationwide.
“Traditionally the Southern Research Station has conducted a lot of locationbased forest hydrologic research,” says Sun. “The question is, how do we extrapolate those data and scale them up? That’s really what our program is focused on. We use modern technology like GIS (geographical information systems) and computer simulation models to scale up to larger areas and make research more relevant to policymaking.”
Working with Sun and McNulty are Jennifer Moore Myers and Erika Cohen, resource information specialists with the global change program. David Wear, project leader of the SRS economics unit located nearby in Research Triangle Park, contributes data on land use change predictions based on timber price fluctuations and population growth.
Overall, the Southeastern United States receives 10 times as much precipitation as needed for human use. If that were spread evenly across the region, year after year, water shortages might be unheard of. But rainfall varies from here to there and from year to year. Equally uneven are the demands placed on water, whether for human consumption, power generation, or crop irrigation. McNulty, Sun, and team built their model layer by layer to reveal both the individual and collective effects of the various factors.
They started with two climate models, each with over 100 years of history and forecasting 100 years into the future. The models differ slightly in how much they predict temperature will increase and whether the future will be wetter or drier than the present. Because the goal is a realistic worst-case scenario—those 3 or 4 years of back-to-back drought that officials actually need to plan for—the differences between the models matter less than the ability to fine tune their forecasts to reflect local conditions. Both of the climate models chosen—one from the Hadley Climate Research Center in Britain and the other from the Canadian Climate Centre—have that capability.
To the climate models, the team added population forecasts. Overall, by 2045, the population of the 13 Southern States is expected to be 90 percent greater than in 1990, but at the watershed level, population change is predicted to vary from a 20-percent decrease to a 500- percent increase.
Then they added changing landcover characteristics, shown in six different classifications from urban to forestland. Forests even got broken down further into two categories. Evergreens, it seems, consume more water than do deciduous trees.
Next, they factored in seven types of water demand, which included how much of the “used” water gets returned to the ecosystem. Because of evaporation, for example, crop irrigation is a far greater drain on available water than residential use, which returns 85 percent of withdrawn water to rivers and streams. “When we started this work, we thought population was going to be the factor driving water stress in the Southern United States,” says McNulty. “It turns out population is very important for water quality—the more people you have, the more likely you are to have reduced water quality—but in the quantity sense there are other factors that have a greater impact.”
Last came ground-water data. Where ground water from aquifers is the major source of water, aquifer levels will drive water stress even in the wettest year on record. Because aquifers can take hundreds or thousands of years to recharge, once that water is gone (and it’s dropping dangerously low in some areas), those communities face nothing short of radical change.
The researchers looked first at the effects of each factor, then put them all together. “You can’t look at a complex issue just by studying the individual components,” says McNulty. “It’s not just population change or climate change, but the combination of the two. And it’s not just average conditions. You have to look at variability to understand vulnerability. Spatial scales matter, too. Some of the biggest changes will occur at the finest scale.”
In a nutshell, here’s what they found: Looking at the South as a whole, climate change is the leading factor in increasing the potential for water shortages. More important locally are population growth—particularly around Miami and in parts of Virginia, Texas, and North Carolina—and ground-water availability.
“That’s why it’s so important to have locally explicit models,” says McNulty. “In Texas, ground water is critically important. In other areas, not so much. Also, on a regional scale, precipitation is important, but the less precipitation you have, the more variability matters. A 20-percent drop makes a much bigger difference in Texas, where water stress is already high, than in western North Carolina and eastern Tennessee, where they get 90 inches of rain a year.”
As they expand the model nationwide, the researchers are adding in a few more factors. They plan to show the effects of seasonal fluctuations in precipitation as well as annual averages and to make more explicit the connections across watersheds. Some of the Nation’s largest metropolitan areas pipe water from distant watersheds. As a result, a light winter snowfall in the Sierra Nevadas will spell trouble for Los Angeles the following summer.
Seasonal variation matters in North Carolina, too. Sun said the team tested its model for rainfall patterns on the southeastern Piedmont in 2002, one of the years Falls Lake and many other reservoirs across North Carolina dropped dangerously low. “That year, the total amount of rainfall was not so bad,” he said, “but it didn’t fall in summer, when demand for water was highest.”
The same thing happened in 2005, leaving me stuck on a mudflat and wondering about the future.
For more information:
Steve McNulty at 919-515-9489 or firstname.lastname@example.org
Ge Sun at 949-515-9498 or email@example.com
Garnet Bass is a freelance writer based in Raleigh, NC, who specializes in science and economic development.