88
Land-Ocean Systems in the Siberian Arctic: Dynamics and History
insignificant contribution from the turbulent heat transfer, govern heat flux through the seasonal
pycnocline to the surface (Figure 7, left diagram). The amount of heat transferred to the surface
by these processes cannot influence the ice cover formation to the extent shown in Figure 4.
At the periphery of the outflow zone, convection is usually 5 times (and, as a maximum, an
order of magnitude) more efficient than molecular heat exchange. However, the efficiency of
the heat exchange governed by double-diffusion convection is twice as large as the salinity
convection (right diagram, Figure 7). So, even when the fast ice edge becomes quasi-steady, all
heat from the intermediate layer at the periphery of the outflow zone is transferred to the
surface. In reality, however, the heat flux from the intermediate layer is small, on average 5.104
KJ m- 2 , i.e. equivalent to 16 crn of unformed ice. However, one should remember that the
potential heat exchange with the surface is extremely large. Its realization requires heat
advection from the adjacent parts of the river discharge zone.
The existence of such advection was tested by numerous CTD-measurements performed
during the expeditions Transdrift I (Kassens et aI., 1994), Transdrift II (Kassens and
Dmitrenko, 1995) and Transdrift III (Kassens et aI., 1997). They demonstrated that the
temperature distribution at the periphery of the river water outflow zone under the seasonal
pycnocline was characterized by numerous inversions. The thickness of such warm and cold
interlayers reached 10 m, and the temperature gradients at the boundaries of the interlayers
reached 2-2.5 °C m- I (Golovin et aI., 1995). The most pronounced isopycnic temperature
inversions were observed at the periphery of the freshened zone in the region west of Kotel,nyy
Island during the autumn of 1995 (Figures I, 9). Their formation was shown to be caused by
the existence of the secondary thermoclinic front under the baroclinic surface discharge front
(Golovin et aI., 1995; Kassens and Dmitrenko, 1995; Figure 5). The mechanisms of
thermoclinic front formation were considered in Discussion 1. In the absence of horizontal
density stratification, but in the presence of strong temperature stratification, any horizontal
disturbance leads to the formation of intrusions. It is obvious that intrusion stratification of the
thermoclinic secondary hydrofront and further advection of heat transferred by isopycnically
spreading warm intrusions are the processes "pumping" heat from beneath the discharge lens to
its periphery. This process is equally efficient during both the summer-autumn season (Figures
5A, 9A) and the period of young ice growth (Figures 5B, 9B).
The formation of intrusions at the front is a typically ageostrophic process. The upper
estimate of its spatial scale is the Rossby baroclinic deformation radius: A. = NDo , where N is
Jrj
the Vaisala frequency, Do is the fluid layer thickness, and f is the Coriolis parameter. For the
Laptev Sea, we have estimated A= 11.8 km. Actual observations of the spatial scale of warm
intrusions near the thermoclinic front during the Transdrift II expedition (Kassens and
Dmitrenko, 1995) showed that their horizontal dimensions could reach 4 kilometers and more.
This estimate is comparable to the width of the river water outflow zone (about 150 km), since
the process of intrusive destruction goes on continuously until complete degradation of the
warm water layer. The high efficiency of this process permits the assumption that intrusive
destruction of the warm layer is complete by the time the fast ice edge attains a quasi-steady
character. All heat except for the heat already transferred to the surface via convective,
turbulent, and molecular heat exchanges will be redistributed by advection to the periphery of
the outflow zone.
Figure 5 displays the evolution of the outflow front from summer (A) to winter (B).
Salination of the surface layer occurs at ice formation. As a result, some heat from the upper
part of the intermediate layer can be isopycnically exchanged with the surface (Figure 5B,
Figure 7, process 4). In order to estimate the intensity of this type of heat exchange, we used
the ratios (1), (5) - (9) for the time during which the fast ice edge stabilizes. When calculating,
Land-Ocean Systems in the Siberian Arctic: Dynamics and History
insignificant contribution from the turbulent heat transfer, govern heat flux through the seasonal
pycnocline to the surface (Figure 7, left diagram). The amount of heat transferred to the surface
by these processes cannot influence the ice cover formation to the extent shown in Figure 4.
At the periphery of the outflow zone, convection is usually 5 times (and, as a maximum, an
order of magnitude) more efficient than molecular heat exchange. However, the efficiency of
the heat exchange governed by double-diffusion convection is twice as large as the salinity
convection (right diagram, Figure 7). So, even when the fast ice edge becomes quasi-steady, all
heat from the intermediate layer at the periphery of the outflow zone is transferred to the
surface. In reality, however, the heat flux from the intermediate layer is small, on average 5.104
KJ m- 2 , i.e. equivalent to 16 crn of unformed ice. However, one should remember that the
potential heat exchange with the surface is extremely large. Its realization requires heat
advection from the adjacent parts of the river discharge zone.
The existence of such advection was tested by numerous CTD-measurements performed
during the expeditions Transdrift I (Kassens et aI., 1994), Transdrift II (Kassens and
Dmitrenko, 1995) and Transdrift III (Kassens et aI., 1997). They demonstrated that the
temperature distribution at the periphery of the river water outflow zone under the seasonal
pycnocline was characterized by numerous inversions. The thickness of such warm and cold
interlayers reached 10 m, and the temperature gradients at the boundaries of the interlayers
reached 2-2.5 °C m- I (Golovin et aI., 1995). The most pronounced isopycnic temperature
inversions were observed at the periphery of the freshened zone in the region west of Kotel,nyy
Island during the autumn of 1995 (Figures I, 9). Their formation was shown to be caused by
the existence of the secondary thermoclinic front under the baroclinic surface discharge front
(Golovin et aI., 1995; Kassens and Dmitrenko, 1995; Figure 5). The mechanisms of
thermoclinic front formation were considered in Discussion 1. In the absence of horizontal
density stratification, but in the presence of strong temperature stratification, any horizontal
disturbance leads to the formation of intrusions. It is obvious that intrusion stratification of the
thermoclinic secondary hydrofront and further advection of heat transferred by isopycnically
spreading warm intrusions are the processes "pumping" heat from beneath the discharge lens to
its periphery. This process is equally efficient during both the summer-autumn season (Figures
5A, 9A) and the period of young ice growth (Figures 5B, 9B).
The formation of intrusions at the front is a typically ageostrophic process. The upper
estimate of its spatial scale is the Rossby baroclinic deformation radius: A. = NDo , where N is
Jrj
the Vaisala frequency, Do is the fluid layer thickness, and f is the Coriolis parameter. For the
Laptev Sea, we have estimated A= 11.8 km. Actual observations of the spatial scale of warm
intrusions near the thermoclinic front during the Transdrift II expedition (Kassens and
Dmitrenko, 1995) showed that their horizontal dimensions could reach 4 kilometers and more.
This estimate is comparable to the width of the river water outflow zone (about 150 km), since
the process of intrusive destruction goes on continuously until complete degradation of the
warm water layer. The high efficiency of this process permits the assumption that intrusive
destruction of the warm layer is complete by the time the fast ice edge attains a quasi-steady
character. All heat except for the heat already transferred to the surface via convective,
turbulent, and molecular heat exchanges will be redistributed by advection to the periphery of
the outflow zone.
Figure 5 displays the evolution of the outflow front from summer (A) to winter (B).
Salination of the surface layer occurs at ice formation. As a result, some heat from the upper
part of the intermediate layer can be isopycnically exchanged with the surface (Figure 5B,
Figure 7, process 4). In order to estimate the intensity of this type of heat exchange, we used
the ratios (1), (5) - (9) for the time during which the fast ice edge stabilizes. When calculating,