82
Land-Ocean SYstems in the Siberian Arctic: Dvnamics and History
Oceanographic Tables (1975), if all heat from the intermediate layer is transferred to the
surface, the sea ice core growth is reduced by 68 cm. Is this sufficient to control the ice
conditions to the extent as shown in Figure 4? Observations carried out on the ESARI-92
expedition in the eastern Laptev Sea showed that the thickness of fast ice in the vicinity of the
fast ice edge was about 1.9 m (Dethleff et a!., 1993). This value differs from the thickness of
the first-year drifting ice behind the flaw polynya by a value of the same magnitude.
As follows from Figure 4, by the time the fast ice edge attained a quasi-steady character, its
position was to a great extent governed by the heat content of the intermediate water layer. If
this is the case, then there must have been some mechanisms for an extremely effective heat
exchange with the surface. Let us discuss the entire complex of mechanisms responsible for
heat transfer from the intermediate warm layer to the lower ice boundary.
Theoretical considerations suggest that the heat exchange between the lower ice surface and
the warm intermediate layer, limited by the seasonal pycnocline from above and by the main
pycnocline from below, can via the following:
molecular heat exchange;
turbulent heat exchange;
convective heat exchange caused by salination of the surface layer at ice formation;
double-diffusion convection through the seasonal pycnocline with the surface sub-ice
layer;
intrusion stratification through the lateral boundaries of the thermocline front.
Molecular heat exchange
The molecular heat fluxes calculated during the onset of ice formation in autumn of 1996 using
TRANSDRIFT III data vary within 0.1 - 0.5 J m- 2 ·s-I. For this calculation, a simple formula
(2) has been used (Turner, 1973):
(2)
where DT is the temperature difference through the pycnocline with a thickness DZ, and Ktm is
the coefficient of molecular temperature conductivity (1.3"10- 7 m 2 s-I for sea water). In the
mass form the heat flux can be presented as:
(3)
where r is the sea water density and Cp is the heat capacity of sea water (approximately 4.19
J.g DC-I. In the three months during which the fast ice edge becomes quasi-steady, about
2.26.10-3 to 21.60.10- 3 KJ m- 2 are transferred to the upper layer. This leads to a decrease in ice
thickness of 2 to 7 cm (process 1, Figure 7).
Turbulent heat exchange
The criterion for different regimes of turbulent exchange is the Richardson gradient number.
For polar regions, it can be written in the form
Ri = g.bDS-Hp·DU-2
(4)
where g is gravitational acceleration, DU·Hp-1 is the gradient of the horizontal component of the
current speed in the pycnocline, DS is the salinity difference in the seasonal pycnocline and b is
the salinity compression coefficient. Calculations of the Richardson gradient number for the
seasonal pycnocline presented in Table 1 were made for the stations on transect II (Transdrift
Land-Ocean SYstems in the Siberian Arctic: Dvnamics and History
Oceanographic Tables (1975), if all heat from the intermediate layer is transferred to the
surface, the sea ice core growth is reduced by 68 cm. Is this sufficient to control the ice
conditions to the extent as shown in Figure 4? Observations carried out on the ESARI-92
expedition in the eastern Laptev Sea showed that the thickness of fast ice in the vicinity of the
fast ice edge was about 1.9 m (Dethleff et a!., 1993). This value differs from the thickness of
the first-year drifting ice behind the flaw polynya by a value of the same magnitude.
As follows from Figure 4, by the time the fast ice edge attained a quasi-steady character, its
position was to a great extent governed by the heat content of the intermediate water layer. If
this is the case, then there must have been some mechanisms for an extremely effective heat
exchange with the surface. Let us discuss the entire complex of mechanisms responsible for
heat transfer from the intermediate warm layer to the lower ice boundary.
Theoretical considerations suggest that the heat exchange between the lower ice surface and
the warm intermediate layer, limited by the seasonal pycnocline from above and by the main
pycnocline from below, can via the following:
molecular heat exchange;
turbulent heat exchange;
convective heat exchange caused by salination of the surface layer at ice formation;
double-diffusion convection through the seasonal pycnocline with the surface sub-ice
layer;
intrusion stratification through the lateral boundaries of the thermocline front.
Molecular heat exchange
The molecular heat fluxes calculated during the onset of ice formation in autumn of 1996 using
TRANSDRIFT III data vary within 0.1 - 0.5 J m- 2 ·s-I. For this calculation, a simple formula
(2) has been used (Turner, 1973):
(2)
where DT is the temperature difference through the pycnocline with a thickness DZ, and Ktm is
the coefficient of molecular temperature conductivity (1.3"10- 7 m 2 s-I for sea water). In the
mass form the heat flux can be presented as:
(3)
where r is the sea water density and Cp is the heat capacity of sea water (approximately 4.19
J.g DC-I. In the three months during which the fast ice edge becomes quasi-steady, about
2.26.10-3 to 21.60.10- 3 KJ m- 2 are transferred to the upper layer. This leads to a decrease in ice
thickness of 2 to 7 cm (process 1, Figure 7).
Turbulent heat exchange
The criterion for different regimes of turbulent exchange is the Richardson gradient number.
For polar regions, it can be written in the form
Ri = g.bDS-Hp·DU-2
(4)
where g is gravitational acceleration, DU·Hp-1 is the gradient of the horizontal component of the
current speed in the pycnocline, DS is the salinity difference in the seasonal pycnocline and b is
the salinity compression coefficient. Calculations of the Richardson gradient number for the
seasonal pycnocline presented in Table 1 were made for the stations on transect II (Transdrift