Wintertime Expansion and Contraction of the Terra Nova Bay Polynya
153
downward longwave radiation emitted by the atmosphère, and Qu is the upward
longwave radiation emitted by the sea surface.
Priestly [19] and Bromwich and Kurtz [38] assert that it is the strong offshore
winds that are responsible for maintaining wintertime open water at Terra Nova
Bay. Figure 3 shows the individual heat flux terms based on an assumed 50%
cloud fraction, 70% relative humidity, and meteorological data from the Manuela
AWS station assuming bulk sensible and latent heat parameterizations developed
by Andréas and Murphy [39] and a longwave parameterization developed by
Maykut and Church [40]. The dominant term in the heat budget is clearly the sensible heat flux (Fig. 3a), with average wintertime values for the 3 years of -770 W m’2
(Table 2). To test the hypothesis that it is the strong katabatic winds that are
responsible for the fluctuations in polynya extent, the polynya model was stepped
forward in time using only the sensible heat flux as input. The daily averaged estimâtes of polynya extent were then regressed against the daily averaged open
water fraction.
Figure 4a shows the scatter plot and least-squares régression between the modeled polynya width and the open water fraction estimated from the satellite data.
The corrélation coefficient is 0.39, explaining 15% of the variance (Table 3).
Figures 4b-d show the open water fraction derived from the satellite data and the
open water fraction estimated by the polynya model and régression analysis. The
observed open water fraction shows significantly more variability than the modeled results and the model misses several major fluctuations in the open water
fraction. In particular, the model dramatically underestimates the observed open
water fraction during three major polynya events during May and June 1988 (Fig.
4b). In addition, the model significantly overestimates the open water fraction
during the later half of July and the First half of August. Similarly, a major polynya
expansion during July 1989 is missed entirely by the model. During other periods,
there is no obvious correspondence between increases or decreases in open water
fraction and the modeled estimâtes.
The mean modeled polynya extent for this case is 21 ±4 km (Table 3). This is
only slightly larger than the 20-km polynya extent derived from AVHRR data [6]
assuming a polynya size of 1300 km2 and a fixed north/south polynya dimension
of 65 km. Note, however, one would expect the AVHRR to underestimate polynya
extent because the instrument is only capable of estimating polynya size during
cloud-free periods. These periods are associated with increased radiative cooling
and hence should lead to smaller polynyas. In contrast, the microwave analysis is
expected to overestimate the open water fraction because the analysis région for
the microwave data includes both the polynya région proper and some fraction of
the offshore loosely Consolidated ice.
Assuming a constant relative humidity of 70%, the second largest term in the
heat budget is the latent heat flux (Fig. 3b) with a 3-year wintertime mean loss
from the polynya of -250 W m'2 (Table 2). The overall modeled structure (not
shown) looks nearly identical to the structure computed from the sensible heat
flux alone, and the corrélation coefficient remains close to 0.39 (Table 3). It is the
close correspondence between the sensible and latent heat fluxes that motivated
Pease [10] to increase the sensible heat exchange coefficient and to ignore the
latent heat flux term in her heat budget.
153
downward longwave radiation emitted by the atmosphère, and Qu is the upward
longwave radiation emitted by the sea surface.
Priestly [19] and Bromwich and Kurtz [38] assert that it is the strong offshore
winds that are responsible for maintaining wintertime open water at Terra Nova
Bay. Figure 3 shows the individual heat flux terms based on an assumed 50%
cloud fraction, 70% relative humidity, and meteorological data from the Manuela
AWS station assuming bulk sensible and latent heat parameterizations developed
by Andréas and Murphy [39] and a longwave parameterization developed by
Maykut and Church [40]. The dominant term in the heat budget is clearly the sensible heat flux (Fig. 3a), with average wintertime values for the 3 years of -770 W m’2
(Table 2). To test the hypothesis that it is the strong katabatic winds that are
responsible for the fluctuations in polynya extent, the polynya model was stepped
forward in time using only the sensible heat flux as input. The daily averaged estimâtes of polynya extent were then regressed against the daily averaged open
water fraction.
Figure 4a shows the scatter plot and least-squares régression between the modeled polynya width and the open water fraction estimated from the satellite data.
The corrélation coefficient is 0.39, explaining 15% of the variance (Table 3).
Figures 4b-d show the open water fraction derived from the satellite data and the
open water fraction estimated by the polynya model and régression analysis. The
observed open water fraction shows significantly more variability than the modeled results and the model misses several major fluctuations in the open water
fraction. In particular, the model dramatically underestimates the observed open
water fraction during three major polynya events during May and June 1988 (Fig.
4b). In addition, the model significantly overestimates the open water fraction
during the later half of July and the First half of August. Similarly, a major polynya
expansion during July 1989 is missed entirely by the model. During other periods,
there is no obvious correspondence between increases or decreases in open water
fraction and the modeled estimâtes.
The mean modeled polynya extent for this case is 21 ±4 km (Table 3). This is
only slightly larger than the 20-km polynya extent derived from AVHRR data [6]
assuming a polynya size of 1300 km2 and a fixed north/south polynya dimension
of 65 km. Note, however, one would expect the AVHRR to underestimate polynya
extent because the instrument is only capable of estimating polynya size during
cloud-free periods. These periods are associated with increased radiative cooling
and hence should lead to smaller polynyas. In contrast, the microwave analysis is
expected to overestimate the open water fraction because the analysis région for
the microwave data includes both the polynya région proper and some fraction of
the offshore loosely Consolidated ice.
Assuming a constant relative humidity of 70%, the second largest term in the
heat budget is the latent heat flux (Fig. 3b) with a 3-year wintertime mean loss
from the polynya of -250 W m'2 (Table 2). The overall modeled structure (not
shown) looks nearly identical to the structure computed from the sensible heat
flux alone, and the corrélation coefficient remains close to 0.39 (Table 3). It is the
close correspondence between the sensible and latent heat fluxes that motivated
Pease [10] to increase the sensible heat exchange coefficient and to ignore the
latent heat flux term in her heat budget.