Virtual Water Transfers in U.S. Cities from Domestic Commodity Flows

Abstract:

Cities have imported water into their boundaries for centuries but understanding how cities indirectly affect watersheds through the commodities which they import is fairly unknown. Thus, we present and discuss here a methodology for determining the virtual water transfers to and from U.S. cities associated with domestic commodity flows. For our methodology, we only consider agricultural and industrial commodities and, to represent the commodity flows, we use the Freight Analysis Framework (FAF) dataset from the U.S. Department of Transportation. Accordingly, we determine virtual water transfers for the 123 geographic regions in the FAF, which consists of 17 states, 73 metropolitan statistical areas (MSAs), and 33 remainders of states. Out of the 41 sectors that comprise the FAF data, we consider only the 29 sectors that account for the agricultural and industrial commodities. Using both water use data for macro-sectors and national water use coefficients for different industries, we determine a weighted water use coefficient for each of the 29 sectors considered. Ultimately, we use these weighted coefficients to estimate virtual water transfers and the water footprint for each city. Preliminary comparisons with other water footprint estimates indicate that our methodology yields reasonable results. In terms of the water footprint, we find that cities (i.e. MSAs) are net consumers, can consume a large proportion of their own production, and can have a large agricultural production. We also find that the per capita water footprint of industrial consumption decreases with increasing population in cities, suggesting that large cities may be more efficient.