Golovin et al.: F razil Ice Formation during the Spring Flood
137
Based on this classification, one see can from Table 2 that during the period prior to the
flood, at mean velocity in the freshened layer of only about 5 cm/s, the regime of the molecularturbulent mass-heat exchange already appears in the pycnocline, and at 15-20 cmls there is a
purely turbulent exchange regime. During the flood outflow of river water, when the upper
layer becomes strongly freshened (Figures 2, 3), the molecular-turbulent exchange regime
through the pycnocline contributing to intensification of frazil ice formation, occurs only at
U"" 15-20 cmls and the turbulent regime at U > 40-50 cm/s.
Table 2: Estimates of the local Richardson number (Ri*) for the pycnocline between river and sea water for
stations LN961 0, LN961 Oa and LN961 Ob at different velocities of river water transport in the upper layer
Station (date)
LN96 10 (22.05)
LN9610a (06.06)
LN9610b (11.06)
U = 5 cm/s
106
1360
1167
U = 10 cm/s
26
340
292
U = IS cm/s
12
151
130
U = 20 cm/s
6.6
85
73
U = 40 cm/s
21
IH
U = 50 cm/s
13.6
11.7
U = 60 cm/s
9.4
8.1
Using Ri. values from Table 2 and determining U* from (II), let us estimate by means of
(14) the real rate of frazil ice formation Vi for different values of the typical motion velocity of
river water U in the upper freshened layer. Figure 5 presents the estimates of the rates of frazil
ice formation in the contact zone between river and sea water at stations st. LN961 0, LN 961 Oa
and LN961 Ob.
Analysis of the calculated rates of frazil ice formation Vi shows that before the flood the rates
are insignificant though sometimes reach 20 cm/day.This rate is typical for the outflow regions
of the Trofimovskaya and Bykovskaya branches. Approaching peak river discharge, the rate of
frazil ice production increases considerably and can reach 1.7 m a day (Figure 5). This
equivals to ",,34 cm of regular sea ice formation if porosity of frazil ice layer is 80 % (Weeks
and Ackley, 1982). The maximum rate of frazil ice formation is observed at U - 40-50 cm/s.
With the increase of U the value Vi is sharply reduced (Figure 5). This is governed by the fact
that at certain actual density stability of the pycnocline, a large increase in U leads to the
turbulent (Ri. < I 0, see Table 2) and then to the unsteady regime (Ri. <2). Buoyancy is no
longer influence the character of the heat -mass exchange through the pycnocline. As a result,
the rate of frazil ice formation decreases as the values of the effective salt and heat exchange
coefficients become equal (Kt "" Ks). This reduces the efficiency of double-diffusion
supercooling or makes it impossible. The conditions (3) and (6) are not fulfilled. A similar
conclusion also follows from laboratory experiments (Krylov and Zatsepin, 1992; Voropayev
et aI., 1995 J.
A frazil-ice production rate of 1.5 m/day seems to be high, but the observed supercooling in
the pycnocline at this time is also very large (Figures 2b, c). Such high ice-production rates will
lead to speculations that fast ice thickness could increase accordingly. However, much of this
new ice probably also melts locally, since toward the conclusion of our field work, river water
with above zero temperature was already spreading below fast ice (Figure 2). A portion of the
137
Based on this classification, one see can from Table 2 that during the period prior to the
flood, at mean velocity in the freshened layer of only about 5 cm/s, the regime of the molecularturbulent mass-heat exchange already appears in the pycnocline, and at 15-20 cmls there is a
purely turbulent exchange regime. During the flood outflow of river water, when the upper
layer becomes strongly freshened (Figures 2, 3), the molecular-turbulent exchange regime
through the pycnocline contributing to intensification of frazil ice formation, occurs only at
U"" 15-20 cmls and the turbulent regime at U > 40-50 cm/s.
Table 2: Estimates of the local Richardson number (Ri*) for the pycnocline between river and sea water for
stations LN961 0, LN961 Oa and LN961 Ob at different velocities of river water transport in the upper layer
Station (date)
LN96 10 (22.05)
LN9610a (06.06)
LN9610b (11.06)
U = 5 cm/s
106
1360
1167
U = 10 cm/s
26
340
292
U = IS cm/s
12
151
130
U = 20 cm/s
6.6
85
73
U = 40 cm/s
21
IH
U = 50 cm/s
13.6
11.7
U = 60 cm/s
9.4
8.1
Using Ri. values from Table 2 and determining U* from (II), let us estimate by means of
(14) the real rate of frazil ice formation Vi for different values of the typical motion velocity of
river water U in the upper freshened layer. Figure 5 presents the estimates of the rates of frazil
ice formation in the contact zone between river and sea water at stations st. LN961 0, LN 961 Oa
and LN961 Ob.
Analysis of the calculated rates of frazil ice formation Vi shows that before the flood the rates
are insignificant though sometimes reach 20 cm/day.This rate is typical for the outflow regions
of the Trofimovskaya and Bykovskaya branches. Approaching peak river discharge, the rate of
frazil ice production increases considerably and can reach 1.7 m a day (Figure 5). This
equivals to ",,34 cm of regular sea ice formation if porosity of frazil ice layer is 80 % (Weeks
and Ackley, 1982). The maximum rate of frazil ice formation is observed at U - 40-50 cm/s.
With the increase of U the value Vi is sharply reduced (Figure 5). This is governed by the fact
that at certain actual density stability of the pycnocline, a large increase in U leads to the
turbulent (Ri. < I 0, see Table 2) and then to the unsteady regime (Ri. <2). Buoyancy is no
longer influence the character of the heat -mass exchange through the pycnocline. As a result,
the rate of frazil ice formation decreases as the values of the effective salt and heat exchange
coefficients become equal (Kt "" Ks). This reduces the efficiency of double-diffusion
supercooling or makes it impossible. The conditions (3) and (6) are not fulfilled. A similar
conclusion also follows from laboratory experiments (Krylov and Zatsepin, 1992; Voropayev
et aI., 1995 J.
A frazil-ice production rate of 1.5 m/day seems to be high, but the observed supercooling in
the pycnocline at this time is also very large (Figures 2b, c). Such high ice-production rates will
lead to speculations that fast ice thickness could increase accordingly. However, much of this
new ice probably also melts locally, since toward the conclusion of our field work, river water
with above zero temperature was already spreading below fast ice (Figure 2). A portion of the