In my previous post I reported that the United States consumed about 410 billion gallons of water per day, and that almost half of it was used to generate electricity. The first graph at right shows 2005 water consumption by state and by use.

First, the largest users are California, Texas, Idaho, and Florida. Between them, these 4 states use more than 26% of all water consumed in the United States.

Second, the amount used varies tremendously between states: California uses 112 times as much as does Rhode Island.

Third, in most states, the major use is power generation. However, in some states, such as California, Idaho, Montana, and Nebraska, it is for irrigation.

Because the amount of power used in a state is a function of population, and because the amount of irrigation is a function of the land area of the state, it would stand to reason that the population and land area of a state must account for much of the variance between states. It does. In fact, it accounts for about 74% of the variance (r-squared = 0.74).

The second graph at right shows per capita water consumption by state. Idaho, Montana, Wyoming, and Nebraska have much higher per capita water consumption than the others. These are states with lots of irrigation but relatively small populations. Missouri ranks 18th highest in per capita water consumption.

The third graph at right shows water consumption per square mile of land area. New Jersey, Maryland, and Connecticut have much higher consumption per square mile than do the other states. They are small states with dense populations. Missouri ranks 30th highest in gallons consumed per square mile.

One can use linear regression to predict what each state’s water consumption “should be” given its population and land area. The regression accounts for about 74% of the variance between states. That’s pretty good. The fourth chart at right shows how much each state’s actual consumption differed from its predicted consumption, as a percentage of predicted consumption. Using this method, water consumption in Idaho, Nebraska, and Montana exceeded predicted consumption the most. Idaho’s consumption was 5.5 times as much as predicted.

Missouri ranks 18th highest in its actual estimated water consumption, compared to its predicted consumption.

Source:

For water consumption: Estimated Use of Water in the United States in 2005, Circular 1344, U.S. Geological Survey, http://pubs.usgs.gov/circ/1344.

For population: Table 1. Intercensal Estimates of the Resident Population for the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010, (ST-EST00INT-01), U.S. Census Bureau, Population, http://www.census.gov/popest/data/intercensal/national/nat2010.html.

For land area of the states: Profile of the People and Land of the United States, National Atlas, http://nationalatlas.gov/articles/mapping/a_general.html#one.

For statistics: a linear regression was performed using state water withdrawals as the dependent variable with state population and state land area as the independent variables. The regression was performed using StatPlusMac using the default settings.