Of concern for public health officials and biologists are models of air pollution as a function of environmental characteristics. Using a meta-analysis of government publications Sokal and Rolf (1995) compiled an interesting dataset which investigates air pollution (measured as annual mean SO\(_2\) concentration per m\(^3\)) as a function of six environmental variables for 32 cities in the United States. Whenever the data were available they are based on averages of three years 1969, 1970, and 1971.
data(so2.us.cities)
The dataset contains 8 variables:
City
US city.
Y
Average annual SO\(_2\) concentration per m\(^3\).
X1
Average annual temperature (degrees Celsius).
X2
Number of industrial companies with more than 20 employees.
X3
Population size (1970 census) in thousands.
X4
Average Annual average wind speed.
X5
Average Annual precipitation (cm).
X6
Average number of days with precipitation.