Almost 30% of Earth’s freshwater supply lies hidden from view as groundwater. These waters, though mostly invisible, are vital for us humans. Groundwater provides about half the global supply of drinking water and is used to grow the majority of the world’s irrigated crops.
Groundwater is also an inextricable cog in the global water cycle. In many areas, discharge from groundwater replenishes streams and rivers, helping sustain aquatic ecosystems. Many of these ecosystems are now under threat, according to a new study.
Inge de Graaf, a hydrological environmental systems researcher at the University of Freiburg, and colleagues simulated on a global scale how current rates of groundwater extraction will affect surface streams and rivers and the ecosystems associated with them.
“Almost 20% of the regions where groundwater is pumped currently suffer from a reduction of river flow, putting ecosystems at risk,” de Graaf wrote in a recent blog post. “We expect that by 2050 more than half of the regions with groundwater abstractions will not be able to maintain healthy ecosystems.”
Areas already at risk include regions with relatively dry climates, such as the High Plains of the United States, and places where large amounts of groundwater are used for irrigation, such as the upper Ganges and Indus basins in the Indian subcontinent. But groundwater pumping has also affected river flow in other locations, including parts of the northeastern United States and Argentina.
Technically, groundwater is a renewable resource, but unsustainable rates of groundwater extraction can deplete reserves faster than they can be replenished by rain, snow, or surface waters. As groundwater levels drop, streams, rivers, and the aquatic ecosystems dependent on these waters can suffer tremendous, and sometimes irreversible, losses.
Building a Global Groundwater Model
Several existing hydrological models simulate the flow of groundwater and its interactions with surface water. But these models work at the level of individual catchment areas. “This is the first study I’ve seen that models groundwater–surface water interactions on a global scale over timescales relevant to management or planning,” said Audrey Sawyer, a hydrogeologist at The Ohio State University who was not involved in the study. “The results provide a great road map for identifying areas that need higher-resolution models and more observations.”
To build a global-scale hydrological model that simulates when loss of groundwater contributions will cause streamflows to fall below levels needed to sustain aquatic life, de Graaf leaned on existing models. These included the PCRaster Global Water Balance model 2 (PCR-GLOBWEB 2) developed at Utrecht University, which simulates moisture storage and exchange between atmospheric, surface, and groundwater reservoirs and accounts for water demands from agriculture, animal husbandry, household use, and industry, and the U.S. Geological Survey’s modular hydrologic model (MODFLOW), which predicts groundwater status and groundwater–surface water interactions.
As inputs for the model, de Graaf used historical data on groundwater demand and extraction from 1960 to 2010. After 2010, she assumed that groundwater use would remain mostly constant through 2100, increasing only in response to irrigation needs as a result of climate change. The model also accounted for different scenarios of climate change based on the Representative Concentration Pathway 8.5 scenario from the Intergovernmental Panel on Climate Change to simulate changes in precipitation due to climate change.
Determining When Streamflow Hits Critical Levels
The model incorporates a previously defined standard that to maintain healthy ecosystems, groundwater extraction should not lower the natural monthly flow of a stream by more than 10% over a period of time. Streams naturally ebb and rise over time, but using this standard, de Graaf calculated a value (the low-flow index) that represents the groundwater discharge needed to maintain at least the minimum natural streamflow necessary to sustain aquatic life in different streams. Streamflows were assumed to reach critically low levels if monthly flow was 10% below the low-flow index for more than 3 months of a year for two consecutive years.
However, groundwater levels and streamflows can be affected by more than groundwater extraction. Climate change, for example, can also affect both. To distinguish between alterations to streamflow driven by climate change alone and those caused by climate change and groundwater pumping, de Graaf ran simulations from 1965 through 2099 that either included groundwater and surface water use by humans or were “natural runs” that excluded human activity. Flow limits reached under both conditions were excluded because they could not be attributed solely to groundwater pumping.
Using results from the model, de Graaf estimates that by 2050 streamflows will be affected in the majority of watersheds worldwide, sometimes even before major groundwater loss. “Only a small drop in groundwater levels can cause these critical river flows,” de Graaf wrote. “Moreover, the impact of groundwater pumping will often become noticeable only after years or decades. This means that we cannot detect the future impact of groundwater pumping on rivers from the current levels of groundwater decline. It really behaves like a ticking time bomb.”
Results from de Graaf and her team’s research were published in October in Nature.
The global scale of the model makes it “a great starting point for identifying watersheds and regions where we need more surface water and groundwater data and higher-resolution models,” Sawyer said. But the scale of the model also means that “we need to follow-up with observations and more refined models relevant to the scale of land use planning and ecosystem processes,” she said.
—Adityarup Chakravorty ([email protected]), Science Writer