Food, Energy, and Water Production Within Watersheds of the United States

Abstract

The production of food, electricity, and treated water is often tracked and managed along political or infrastructure boundaries. Yet, water resources, a critical input in the production of these goods, are delineated along natural landscape features (i.e., watersheds). The boundary mismatch between water resources and the associated production of economic goods conceals hydrologic dependencies and vulnerabilities in the provisioning of Food‐Energy‐Water (FEW) resources. In this study, we pair economic, infrastructure, and hydrologic data to evaluate the production of food, electricity, and treated water within watersheds of the conterminous United States. The US FEW sectors produced 950 million tonnes of crops, 3,973 million MWh of electricity, and supplied water to 263 million people in 2017. FEW production consumed 128 km3 of blue water (18%) and 583 km3 of green water (82%). Watersheds in central and southern California, the Midwest, and the Southwest have the largest FEW blue water consumption and the greatest exposure to water stress. Nearly three‐fifths of FEW production occurs in regularly water‐stressed watersheds. FEW production in watersheds in the Great Plains and Midwest relies heavily on groundwater to buffer against intra‐ and inter‐annual streamflow variability, while surface reservoir storage buffers against water shortages in all watersheds. We show where FEW production may be susceptible to curtailments due to ongoing groundwater depletion or known infrastructure deficiencies. This study adds to our understanding of how a nation's water resources and associated infrastructure support economic activity, as well as areas where economic activity is exposed to hydrological and infrastructure risks.

Watersheds and Infrastructure Providing Food, Energy, and Water to US Cities

“…We used sub-county, sub-watershed data from Ao et al (2023b) and described by Ao et al (2023a) to resolve the mismatch between the county boundaries used for commodity production/flows and watershed boundaries. Briefly, we used high-resolution (30-250 m grid cells) crop area (USDA, 2021) and irrigation data (Pervez & Brown, 2010) to downscale crop production and flow data to sub-county, sub-watershed pairings.…”

“…While we calculated green water footprints, our results focus on blue water footprints and virtual water flows given watersheds' role in aggregating and provisioning blue water to cities. We used existing calculations from Ao et al (2023b), as described by Ao et al (2023a), to estimate the water footprint of agricultural production and flows. Crop-specific water footprints and virtual water transfers were aggregated to their corresponding SCTG category.…”

“…Instead, we use surface water and groundwater availability to partition the served population to each source watershed. Our approach follows that of Ao et al (2023a). Briefly, we determine surface water and groundwater availability using the most recent years of publicly available data (2010-2015) from the online Water Supply Stress Index (WaSSI) model developed by the US Forest Service (Caldwell et al, 2011;Caldwell et al, 2012;McNulty, S., & Sun, G. (n.d.)).…”

“…The amount of water each city consumed from each supplying watershed was quantified by multiplying a watershed-specific water consumption coefficient (m 3 consumed per capita; from Ao et al, 2023a;Ao et al, 2023b) by the population served by each watershed. The product of the population served and the water consumption coefficient provided the amount of a city's water consumption attributed to each supplying watershed.…”

“…Nonetheless, cities do not maintain the same level of spatial diversity in their direct water supplies as their indirect water dependencies. In 2017, cities consumed a little over 13 billion m 3 of water delivered from the nation's watersheds, delivered by an array of civil infrastructure (for reference, the public supply sector consumed approximately 22 billion m 3 in 2017 across all CONUS (Ao et al, 2023a)).…”