A study of seasonally frozen soil was conducted from October 1985 through April 1986 and from October 1986 through April 1987. Three runoff plots were established. On October 30, 1985, 86 mm (millimeters) of water was applied to plot 1, and 43 mm of water was applied to plot 3. No water was applied to plot 2. The winter of 1985-86 had colder-than-normal air temperatures and greater-thannormal precipitation. Some freezing-induced redistribution was measured within the soil profile at some sites. No measurable upward movement of water from the water table to the freezing front was detected in any of the plots.

Snowmelt runoff occurred on March 21 and 22. Plot 1 had 14.2 mm of runoff, plot 2 had less than 0.1 mm of runoff, and plot 3 had 9.0 mm of runoff. Infiltration was determined as the difference between soil water content on March 3 and on March 24. Infiltration was 64.8 mm for plot 1, 43.0 mm for plot 2, and 34.8 mm for plot 3.

The ground-water level started to rise rapidly 2 days after the start of the major snowmelt. Recharge computed from the change in ground-water levels for March 24-27 was 13.2 mm. The mean change in soil water content for March 24-27 indicates a loss (recharge) of 5.1 mm for plot 1, a loss of 1.9 mm for plot 2, and a gain of 4.4 mm for plot 3. The difference between recharge computed from the change in ground-water levels and recharge computed from the change in soil water content indicates that some of the recharge is from a location other than the plots.

The winter of 1986-87 had warmer-than-normal air temperatures and less-than-normal precipitation. Because water was entering and leaving the soil profile during the mild winter, changes in soil water content caused by freezing-induced redistribution, infiltration, or evaporation could not be quantified. On February 26, rainfall runoff occurred from snow-free frozen soil on plots 2 and 3. Snowmelt runoff occurred on all three plots on March 4 and only on plot 3 on March 5. Between February 9 and March 5, infiltration was 50.1 mm for plot 1, 25.4 mm for plot 2, and 49.6 mm for plot 3.

The ground-water level rose very rapidly for a few days at the beginning of March. This rise corresponded to the snowmelt runoff on March 4 and 5. The ground-water level then stabilized until March 18 when it again started to rise rapidly. This rise continued throughout the month. Recharge computed from the change in groundwater levels for March 5-11 was 5.9 mm, recharge for March 11-26 was 50.4 mm, and recharge for March 26 through April 1 was 18.7 mm.