Winter

Usually air temperatures are not cold enough

The crest of a bank may also be snow-free more

for sufficiently long periods during the fall to

often than the lower parts of the bank (Fig. 5),

cause substantial soil water migration to the freez-

depending on wind conditions, which would en-

ing soil to form significant ground ice. Thus, the

hance heat loss at the crest as well. Snow insulates

amount of ground ice that thaws in the fall during

soil and thus inhibits frost penetration and re-

diurnal thaw periods would be much less than

tards soil thaw in the spring (Harlan and Nixon

that which thaws in the spring after an entire

1978), and its insulating effect is influenced by

winter. I conclude that the amount of bank soil

snow type, thickness, and density.

loss or instability that results from FTC in the fall

Reid (1985) found a difference in the frost depth

would be less than that which occurs due to spring

along north- and south-facing banks, as well: frost

thaw (Table 2).

was thicker in north-facing banks. Harlan and

Nixon (1978) concluded that south-facing banks

receive more insolation and thus do not freeze as

Winter

frequently, deeply, or completely as north-facing

Thickening of frozen soil

banks. Wuebben* points out that more diurnal

Once the average daily air temperature falls

FTC may occur in south-facing banks during that

below 32F and solar insolation is sufficiently low,

portion of the season when north-facing banks

the soil continues to lose heat to the atmosphere,

may remain frozen.

most soil thawing stops, and the zone of frozen soil

thickens (Table 2). A special condition exists along

Ground ice growth and frost heaving

an eroding bank (which is usually unvegetated)

Three conditions must exist for ground ice to

versus a stable, vegetated bank. Vegetation insu-

grow and become a substantial component of a

lates soil and reduces heat loss to the air, thereby

soil mass: a soil-moisture supply, sufficiently cold

reducing frost depth (McRoberts 1978) in a stable

air temperatures to cause soil heat loss and subse-

bank, while an eroding bank is likely to have greater

quent freezing, and a frost-susceptible soil, which

heat loss and deeper frost. The deeper frost makes

is usually a silty soil (Anderson et al. 1978). (Needle

more of the surface soil susceptible to erosion and

ice, however, will form in almost any type soil.†)

instability than would be in a vegetated bank.

Silty soils absorb moisture rapidly because they

Heat loss from soils at the crest of an eroding

have particles small enough to provide compara-

bank may even be greater than occurs from lower

tively high capillary rise and large enough to fur-

parts of a bank because heat is lost through the

nish voids of adequate size to allow quick flow of

upland land surface near the bank crest and

moisture through the silt, which leads to rapid

saturation of the voids of the soil (Jumikis 1962).

through the upper bank (Fig. 4). Reid's (1985) mea-

Coarse- and fine-grained soils do not absorb mois-

surements showed that the frozen zone of the

ture rapidly.

upland land surface landward of the crest of an

In addition to soil texture, soil frost susceptibil-

eroding bank was thicker near the bank face along

ity depends on, and varies with, vegetative cover

Orwell Reservoir in Minnesota, and thinned

and depth, thickness and density of snow cover,

landward away from the bank.

initial soil temperature, air temperature

regime, exposure to sun, the tempera-

Bank

Upland

ture gradient within the soil, the rate of

heat removal, the mobility of soil water,

the depth to the water table, overbur-

Zone of Enhanced Heat Loss

Through Upland Land Surface

den stress, and soil density (Chamber-

and Upper Bank Near Crest

lain 1981, Jumikis 1962).

Frozen Zone

* Personal communication with James L.

Unfrozen Zone

Wuebben, Research Hydraulic Engineer, Ice

Engineering Research Division, CRREL,

1994.

† Personal communication with Edwin J.

Figure 4. Bank crest zone with thickest frozen zone along a bank

Chamberlain, Jr., Research Civil Engineer,

profile.

Applied Research Division, CRREL, 1995.