Table 1 
Prudhoe Bay, Alaska, maximum and minimum 
temperatures °C May 1972 (See Fig. 8) 
Date Maximum Minimum 
1 - 6.1 -16.1 
2 1 -14.4 
3 10.0 -17.8 
4 - 7.8 -16.7 
s) 0.6 -14.4 
6 1.7 - 6.1 
z 20 - 12 
8 - 0.6 - 5.6 
9 “del -12.2 
10 - 5.6 -16.7 
11 5.0 -15.6 
12 - 3.3 12.2 
1S - 3.9 SER 
14 = bY -13.3 
15 -10.6 -15.6 
16 10.6 20.0 
V7 -11.1 16.1 
18 SORRY! -18.9 
19 -10.6 15.6 
20 = 3:9 14.4 
21 72 12.8 
22 - 3.9 15.6 
23 - 3.9 TAS7. 
24 28 10.0 
25 - 17 - 9.4 
26 = 22 - 7.8 
27 1.1 - 6.7 
28 5.6 So 
29 1.1 edd 
30 0 2.2 
31 0.6 Ss 
The snow temperature profiles in Fig. 9 
contain information on processes operating in 
the snow. In Figs. 9b and 9c the temperature 
decreases from -3+1°C at the snow surface to 
about -10°C at the bottom. However, we know 
that the surface temperature was at O°C 
during 6-7 May because of the melt crust and 
the evidence of percolation and_ refreez- 
ing of meltwater in the snow. These profiles also 
have anomalously high temperatures adjacent to 
the ice lenses in the snow. These temperature 
anomalies, caused by the release of latent heat as 
percolating meltwater refroze to form the ice 
lenses, have already been smoothed consider- 
ably. Initially, they can be quite pronounced, as 
has been seen in polar glaciers (Benson 1962, pp. 
22-23 and Fig. 25). The amount of percolating 
meltwater in these examples may be estimated 
from the thickness of the ice lenses and the 
increase in density involved in forming them. 
The total amount of ice in the stratigraphic 
columns was about 1 cm, varying from 0.5 to 
2 cm. Most of the ice lenses occurred at the base 
of or within the fine-grained hard layer. The 
increase in density was in the range of 0.5 to 0.6 
gcm’*. Thus, in these cases the amount of latent 
heat added at the surface but released at depth 
within the snow was about 40 to 50 cal cm. 
This is a significant part of the total amount of 
heat which was required on 14 April to bring the 
snow to the melting point. 
During the time when the percolation 
process operates, the tundra snow cover may 
undergo significant daily temperature variations 
in addition to the longer sort of variations 
summarized in Table 1 and Fig. 8. Indeed, the 
tundra snow is so shallow that temperature 
variations imposed at the surface are transmitted 
rapidly to the bottom even without the action 
of percolating meltwater. This is especially clear 
when one compares the temperature profiles of 
Figs. 9b and 9c with those of Figs. 9d and Ye, 
which were measured only 2 days later. The 
system remained below the melting point during 
this time, so there was no latent heat transfer by 
percolation and refreezing. To facilitate the 
comparison, we have plotted five of the temper- 
ature profiles* from Figs. 7 and 9 on a single 
diagram (Fig. 10). The mid-May cooling trend is 
clearly apparent. Indeed, the amount of heat 
required to raise the snow to its melting point 
(the “cold content’’), was only 34 cal cm? in 
both Figs. 9b and 9c. Two days later, on 16 
May, it had increased to 63 cal cm*? (Fig. 9d). It 
was 87 cal cm in Fig. 9e, but that profile has a 
slightly larger mass. The water equivalent of the 
snow, profiled in Figs. 7 and 9, and its cold 
content are summarized in Table 2. 
“The profile in Fig. 9a was omitted because the depth was anomalous. 
