Date of Award
Doctor of Philosophy
Environmental Resources & Policy
The United States is the world's top virtual water exporting nation, but not much is known about the country's internal virtual water flow patterns and the volumes involved. Previous studies have suggested that the import of virtual water, defined as the volume of water required to produce a commodity or service, can relieve pressure on a region's water resources. This study seeks to quantify virtual water flows among U.S. states using the water footprint and input-output analytical methods, and to compare the quantitative results to actual water use volumes in agriculture. The results showed an overall pattern where virtual water is transferred from sparsely populated states mostly in the Midwest, where the country's most fertile agricultural land is located, to the relatively dry Western states, and to the densely populated, but relatively wet coastal regions in the East of the country. For the year 2008, states used 196 Gm3 of water to produce agricultural commodities (crops and livestock) that were exported for consumption in other states. This total virtual water export volume is equivalent to 35 percent of total water withdrawals for all sectors in the U.S., or 41 percent of total rainfall evapotranspiration volume. Gross annual virtual water import volumes were 191 Gm3, giving a net interstate virtual water flow volume of 5 Gm3 for all states. The total virtual water import volume represent 34 percent of total water withdrawals in the U.S., or 40 percent of total rainfall evapotranspiration volume. The estimates in this study cover virtual water flows as a result of trade in 9 primary crops which represent 95 percent of the cultivated area harvested, and trade in nine primary animals that represent nearly 90 percent of animal establishments, and 97 percent of the total national sales in the U.S. for the year 2008. The estimates do not include virtual water flows as a result of trade in processed crop and livestock products and industrial products, which would have resulted in even higher virtual water flow volumes. Commodities making the greatest call on the nation's water resources were corn for grain, with 20 percent of total water use, and milk cows with 17 percent. The total evapotranspiration volume for the nine primary crops analyzed was 332 Gm3/yr. This consists of 93 Gm3 irrigation water (excluding 25 percent irrigation losses), and 239 Gm3 from rainfall, showing that rainfall contributed 72 percent of the total water volumes required to produce primary crops. If irrigation return flows are considered, the proportion contributed by rainfall becomes 65 percent, compared to 35 percent (128 Gm3) for irrigation water. The nine live animals for all states used 636 Gm3 in 2008, with beef cattle taking up 340 Gm3, or 53 percent of the total volumes used for animal production. Net virtual water exports in absolute terms ranged from 91 Mm3/yr in the state of Washington, to 15 Gm3/yr in Iowa, while the minimum net virtual water import value was 47 Mm3/yr in Vermont, to 11 Gm3/yr in Florida. On a per capita basis, the people of North Dakota were responsible for the largest agricultural net virtual export volume (16,011 m3/yr/ca), although the state has only 0.2 percent of the national population. Washington was responsible for the lowest per capita net virtual water export (375 m3/yr/ca). The people of Delaware (0.3 percent the total population) were responsible for the largest net virtual water imports related to agricultural commodities on a per capita basis (1511 m3/yr/ca), with Nevada ranking lowest. In absolute terms, water footprint values in relation to the 18 primary crops and livestock groups ranged from 1157 Mm3/yr in Rhode Island, to 61,471 Mm3/yr in California. Water footprint per capita values ranged from 1,083 m3/yr/capita in New York, to 4,872 m3/yr/capita in Nebraska. Both water footprint and input-output methodologies showed that virtual water transfer constitutes a substantial portion of the water balance in water scarce states such as California, where imports and exports were found to be 13 and 15 percent of total actual water use. The ratios of net virtual water import to agricultural water use volumes were very high for relatively humid states such as Rhode Island (nearly 5,000 percent) and Connecticut (more than 3,500 percent), partly showing that factors related to economic structure dominate climatic factors (water endowments) in shaping virtual water flow patterns in most U.S. states. These results suggest that rather than being the main reason behind observed virtual water flow patterns, water availability is complimentary to other factors of production, mainly the availability of suitable agricultural land. Similar to Japan or some European countries, most highly states in the eastern part of the country rely heavily on virtual water imports to meet their local agricultural consumption requirements, while their economies focus on sectors that are less land and water intensive, such as the services industry. The study also revealed that the volumes of international virtual water imports and exports are dwarfed by internal (interstate) virtual water volumes in the U.S., showing an overall preference for home consumption to international trade. The productive value of water ($/m3 used) was found to be much higher for industry and domestic sectors, in comparison to more water intensive agricultural use. While input-output analysis appears less prone to estimation errors and is less laborious to implement, it is limited in assessing the virtual water content of individual commodities when compared to water footprint analysis. However, the two alternative methodologies both produced results that are to a large extent consistent with production and consumption patterns in the U.S. The study adds new insights and information to earlier global studies that did not elaborate much on the internal virtual water flow dynamics of the world's largest virtual water exporter. The knowledge is relevant for this large country, where there are wide variations in water and other natural resource endowments between regions.
This dissertation is Open Access and may be downloaded by anyone.