About 40 million residents in the Western U.S. from Denver to Los Angeles rely on the Colorado River Basin (CRB) for water. Scientists and policy makers routinely use regional or global-scale models with resolutions ranging from 200 km to 25 km to predict the impact of climate change on future stream flows.
An EESA staff geological scientist and his colleagues at Colorado School Mines and Princeton University used a localized, physically based model and advanced computing to study an archetypal mountainous watershed in the Colorado River Basin, the East River Watershed. Their research found a higher evapotranspiration rate for the catchment that they attributed largely to high-elevation trees.
Such catchments occupy 15% of CRB lands in the Rocky Mountains, yet account for about 85% of streamflow into the basin. The researchers estimated that the 10% increased water loss to the atmosphere would produce a 4% reduction in future streamflow into the CRB with increasing global temperatures. The more sensitive, highest-resolution model predicted the increased water loss when compared to a relatively less sensitive, 1-km resolution model that they ran (a 16% versus a 12% streamflow loss).
The finding was made possible by a 100-m resolution computer model developed by Lauren Foster, who conducted the modeling as a Department of Energy-funded graduate student of Kenneth H. Williams, director of Berkeley’s Environmental Remediation and Water Resources Program, and Reed Maxwell, a Princeton University engineering professor who directs its Integrated Groundwater Modeling Center. All of the investigators are performing research as part of a Department of Energy watershed research program focused on understanding how complex, multi-scale interactions can lead to a cascade of effects on downstream water availability, nutrient and metal loading, and carbon cycling.
Foster said, “Many water managers are likely overestimating water supplies for the CRB and other mountain generated sources of water. They could apply adjustment factors to incorporate the new findings into their models.”
To determine the impact of model resolution on hydrological predictions, the researchers studied the representative Rocky Mountain headwater that covers 255 km. Data from three meteorological stations within the headwater were modeled using nine different precipitation scenarios, as well as a global temperature increase by midcentury of 4 degrees Celsius, half that amount, and no temperature change. The results on the East River Basin at 100-m and 1-km resolution were compared to larger-scale modeling results.
Foster then sought the source of the reduced streamflow for the study published September 22 in Environmental Research Letters. They identified that evergreen forests at high elevations were sensitive to climate change and produced much of the heightened water loss with global warming by 2050. In particular, the high-resolution modeling pinpointed evergreen forests growing between 3,250 m to 3,750 m as the primary driver of increased evapotranspiration. That is, these evergreen trees lost much more water in the simulations when a small increase in temperature occurred than did other vegetation types or land features.