Impact of Sea-Ice Bottom Topography on the Ekman Pumping - Towards an Interdisciplinary Approach in Earth System Science - page 134

From the results of this calculation we can conclude that the variations in Ekman
pumping associated with variable oceanic drag coef cients is at least in the same

order of magnitude as the variations due to changes in the surface velocity of the
ice. Even though the strong local effect might be damped when averages are taken
over a larger area, we can still assume that the effect will remain of the same order
of the velocity variations shown by Rabe et al. ( 2011 ). Thus we speculate that the
presence of different sea-ice regimes on a large scale may induce a basin-scale
variation in Ekman pumping that then would have consequences for the Ekman
transport and the large scale ocean circulation. Numerical experiments and simu-
lations with large scale sea ice-ocean models could help to investigate the effect on
an Arctic basin scale ocean circulation.

5 Summary and Conclusion

Airborne altimetry and EM-bird observations have been used in the present study to
reconstruct the surface and bottom topography of the sea-ice in the Arctic Ocean.
From the obtained pro

les we detected the keels and calculated keel mean depth
and keel mean separation along pro les of different length. This information is then

used to calculate the oceanic drag coefficients. These coefficients are calculated by
applying a parameterization presented in Lu et al. ( 2011 ) to a hypothetical situation
of 100 % sea-ice cover. The calculated drag coef cients are in the range of values

obtained by in situ observations. The range of variability is large and this suggests
that the choice of a constant drag in sea-ice numerical models might lead to a
misrepresentation of the actual ice-ocean momentum transfer.
The calculated oceanic drag coef

cients have been used for a rough calculation
of Ekman pumping. The results provide an insight into the expected magnitude of
Ekman pumping caused by the variability of the oceanic drag coef cients. The

upper layer vertical velocity generated by variations in drag coef cients is on the

same order of magnitude as for variable ice velocity at the surface. In order to better
understand the importance of the variable drag coef

cients on the large scale
oceanic circulation, Pan Arctic simulations with global circulation models are
required. The results shown here suggest that neglecting the contribution of variable
oceanic drag coef cients in the momentum transfer between ice and ocean can lead

to considerable errors in numerical models or data analysis.

Acknowledgments We would like to thank the crew of the RV Polarstern and the HeliService
International GmbH. Moreover we thank all the people who contributed to collect the laser
altimeter data and EM-bird data, in particular Stefan Hendricks. We are also very thankful to
Michael Karcher for the interesting and constructive discussions. Finally we thank the Earth
System Science Research School (ESSReS) for any support to this study and to the PhD project.