THE "HEAT ISLAND" ASSOCIATED
(Tm - Ta )
dhi
1
=
WITH A FREEZING RIVER
dt ρiL hi
h
1
(1)
+
+ s
k
i Hia ks
The measurement series described in this work
began during the winter of 198990. The sam-
where the i subscripts refer to the river ice, the s
pling point network was altered several times
subscripts refer to snow on the ice, and a sub-
during that winter, and most data reported have
scripts refer to air above the ice or snow. The
been from the homogeneous measurement se-
terms are defined in Nomenclature, and the tem-
quence used during 199093. Temperature mea-
perature of the water beneath the ice is fixed at
surements in the vicinity of the Piermont bridge,
the melting point.
P, detected an interesting heat island during early
Integrating eq 1 over a finite time increment
December 1989. The temperature measurement
results in a quadratic equation. The positive root
points that defined this heat island are those east
of this quadratic equation can be solved for the
of the Piermont bridge on Figure 3.
final ice thickness hif at the end of the time incre-
December 1989 was an unusually cold month
ment:
in the experiment area, and this month was cli-
matologically one of the coldest Decembers of
1/ 2
2 2k T ∆t
hif = (hii + hr ) - i a
record in the northeastern United States. The air
- hr
(2)
ρiL
temperature remained continuously below freez-
ing at the recording site at L in Figure 3 through-
where
out the first 30 days of the month. Morning
temperatures at L were less than 15C on all but
1
h
hr = ki
+ s.
(3)
four days. Ice cover formed on the Connecticut
Hia ks
River on 4 December. A light snowfall on 6 De-
cember allowed visual verification that the ice
where hii is the initial ice thickness and hr acts as
cover was continuous throughout the study area.
an equivalent ice thickness corresponding to the
A few additional centimeters of snowfall provided
snow exchanging heat with the air. The squared
thin snow cover through early December.
term in eq 2 is much larger than the quotient
A significant snowfall of 2535 cm occurred in
term, allowing the braced square root term to be
the study area on 16 December 1989. We com-
replaced by the first two terms of a converging
pared (Hogan and Ferrick 1990) the morning air
binomial series. This allows eq 2 to be rewritten
temperatures near the river on the five days prior
as
to this snowfall with those on the days following.
kiTa ∆t
The temperatures observed prior to the snowfall,
hif = hii -
ρiL (hii + hr )
(4)
on the bridge P and at h adjacent to the bridge,
were 2C greater than those observed east of h,
in which the hr term acts as an equivalent ice
along basin IV. The temperatures on the bridge,
thickness corresponding to the snow exchanging
and at h, were less than those east of h following
heat with the air.
the snow of 16 December.
This can now be used to estimate ice growth on
We propose that the air over and adjacent to
the Connecticut River during early December
the river was warmed by the heat rejected during
1989. Ice cover was observed on the morning of 4
river ice growth prior to 16 December, and that
December, indicating ice growth began sometime
the insulating effect of the additional snow halted
on 3 December. The mean daily temperature in
this warming process. We will examine this hy-
the plane intersecting the river bank was 14C
pothesis using the Ashton (1989) ice growth equa-
during the period 315 December. The snowfall
tion.
of 6 December was quite light, and can be ne-
The heat released due to ice growth is propor-
glected in using eq 2 to calculate a mean ice growth
tional to the change in ice thickness, over a length
of 3 cm/day (0.03 m/day), yielding 0.36 m (36
of river of a given width. Temperature index mod-
cm) of ice for the period.
els provide good estimates of the growth of river
We now assume that the 2535 cm of snow
ice on a daily scale. The equation for ice growth of
which fell on 16 December settled to a rather
Ashton (1989) can be rewritten to include the ther-
uniform 20 cm during the following days. Ashton
mal resistance of snow cover:
31
Previous Page
