Hydrologic effects of highway-deicing chemicals in Massachusetts

Methods of estimating annual mean and annual maximum chloride concentrations and sodium concentrations in streams were developed using multiple and simple linear regression techniques and data collected during the 1972-77 water years. Independent variables are easily obtainable parameters such as total salt application within a basin, annual precipitation, and drainage basin characteristics. Methods for obtaining gross estimates of chloride loads and sodium loads from salt-application data and estimates of nonhighway-related chloride and sodium were suggested. A chloride budget was calculated for a small drainage basin containing a section of interstate highway. The chloride budget was described in terms of the percentages in direct runoff to the stream, in ground-water discharges to the stream, in storage in the ground, and in the amount unaccounted for. Attempts were made to relate chloride concentrations in ground water adjacent to highways to quantities of salt applied to the highways on an annual basis, annual precipitation, depth of the wells below land surface, depth of wells below the water table, and distance of well from edge of pavement. Little correlation was observed between annual salt-application values and annual mean chloride concentrations in ground water near highways. The irregular distribution of highway runoff, due to topographic differences between sites, and variations in salting and runoff during individual storm events seem to affect correlation between quantities of salt applied and chloride concentrations in ground water near highways. (USGS)