Posted 27 August 2014. PMN Crop News.
New Tool Eases Task of Simulating Aquifer Refill
Source: American Society of Agronomy Press Release. www.agronomy.org
Madison, Wisconsin (August 1, 2014)--How quickly a lake fills after water is drawn for irrigation or drinking is easily measured, but that’s not true for underground water reserves, called aquifers. Because it takes place belowground, groundwater replenishment—or recharge—can’t be directly observed. Scientists must estimate it, often by using complex mathematical models.
A new screening tool may now ease the task. Writing in the Vadose Zone Journal, scientists describe a method for identifying timeframes and regions where the seepage of water into an aquifer is likely constant, rather than fluctuating with rainfall patterns or climate. By locating these areas upfront and excluding them, modelers can then focus their computational might where it’s truly needed, says Jesse Dickinson, a U.S. Geological Survey scientist who led the research.
“There are a lot of complicated processes that go on between the land and the aquifer, and many models now include features that can simulate variability in infiltration from the land surface to the water table,” he says. But these models also require much more data and computing time, and can add unnecessary complexity, he adds. “So what we did was create a mathematical tool so that you can find certain areas where maybe all that [detail] isn’t needed.”
Hydrologists are adding more detailed data to groundwater models today for several reasons, Dickinson explains. For one, aquifers once contained so much water that whatever happened short-term at the surface made little difference to the volume stored belowground. But as aquifers are increasingly tapped—and, in some cases, tapped out—for irrigation and urban use, the amount of water they contain starts depending more heavily on seasonal rainfall or even individual storms. Meanwhile, precipitation patterns are expected to become even more variable with climate change.
“So, people are getting more and more interested in modeling these shorter-term cycles in variability and recharge in order to forecast how much water is left in these really depleted systems,” Dickinson says.
Scientists are also adding more complexity to models today because more data are available than ever before, along with sophisticated software that can handle them. But the screening tool shows this can be overkill in some cases. When the researchers applied it to California’s Central Valley, for example, it predicted that recharge in most of the underlying aquifer doesn’t fluctuate much on time scales of 30 days or less. That is, any daily variation in, say, soil moisture levels or storm events, can likely be excluded from recharge models without affecting their accuracy.
Over periods of 90 days up to a year, in contrast, the tool predicted that changing conditions at the surface do lead to varying recharge rates below, suggesting that seasonal and yearly data are important to include. But whether recharge was judged to be variable or steady depended greatly on the depth of the water table. Even over the course of a year, recharge was still constant in some portions of the Central Valley; specifically in spots where the aquifer is so deep that any fluctuations in precipitation or irrigation are damped by the time surface water reaches it.
“But where the water table is shallow, that’s where we found the variability becomes more important to include,” Dickinson says. “And that’s actually where most of the farming is done and most of the water resources models are being constructed by various groups.”
Although the tool is probably most relevant to groundwater modelers at this point, he says, anyone can download it at az.water.usgs.gov/software/. All that’s needed to use it is information on soil type, an estimate of depth to the water table, and a sense of the long-term trends in recharge: whether the system overall is wet or dry, for example, and when the bulk of precipitation occurs during the year. The tool runs in the MATLAB programming environment.
The team was also able to make maps of the entire Central Valley because of the wealth of available information, but the screening tool can be applied to much smaller areas, as well. “We designed it to be pretty simple and to be used immediately,” Dickinson says.
The research was funded by the Southwest Climate Science Center.