1. Intercomparison of ocean circulation in regional Arctic Ocean models at increasing spatial resolution – Preliminary Results Robert Osinski, Wieslaw Maslowski Naval Postgraduate School 13-th AOMIP Meeting, Woods Hole, 21-23 October 2009

2. Outline Model characteristics Model grid configurations Comparison of circulation and EKE -Barents Sea -Nordic Seas -Fram Strait -Labrador Sea Conclusions

3. Atmosphere Coupler Sea Ice Ocean Ocean sea-ice coupled model. Parallel Ocean Program (POP) Sea Ice model (CICE) Forcing data (ERA15) Fluxes exchange between each component

4. Model configurations: 18-km (1/6°), 9-km (1/12°) and 2-km (1/48°). Parameter18-km Model9-km Model2-km Model Ocean Model Ice Model LANL POP Zhang & Hibler (1997) LANL POP CICE LANL POP CICE Horizontal grid368x3041280x7205520x2880 Vertical levels304548 BathymetryETOPO5+IBCAOModified IBCAO+ETOPO5 Modified IBCAO+GINA+AOOS Ocean timestep20 min.8 min.2 min. Ice timestep120 min.24 min.6 min.

5. Pan-Arctic models domains

6. 1980 annual mean velocity [cm/s] at 0-223m in the Barents Sea. [cm/s] More realistic representation of NCC and WSC Limited interaction between AW and NCC in the 18km model Circulation around Central Bank, west of Novaya Zemlya and outflow into Santa Anna Trough Maslowski et al.. (2008)

7. 20 [cm/s] 1983 annual mean velocity at 0-223 m in the Barents Sea Norway Novaya Zemlya Svalbard 75°N 30°E Stronger and extended NCC and inflow into the Kara Sea in the 2km model Increased complexity of the circulation in the southern Barents Sea Improved WSC and boundary currents 9-km 2-km Norway Novaya Zemlya Svalbard 20 [cm/s] 30°E 75°N [cm/s]

8. Surface (0-5m) annual mean Eddy Kinetic Energy (EKE) in the Barents Sea Three areas of activities: NCC WSC Kara Gate and southern Kara Sea [cm 2 /s 2 ] 9-km 2-km

9. Surface (0-5m) seasonal mean EKE: winter (top) and summer (bottom) from 9-km (left) and 2-km (right) models Both models show the largest EKE in winter (JFM) and weakest in summer (JAS) EKE in NCC 2-3 times larger in the 2-km model inflow through Kara Gate [cm 2 /s 2 ]

10. Annual mean EKE at 0-223m in the Barents Sea: 9-km (left) and 2-km (right). WSC better defined over larger depth range along the slope [cm 2 /s 2 ]

11. 1983 Annual mean velocity at 0-223 m in the Nordic Seas Main differences: EGCC and NCC Irminger Current inflow into the Iceland Sea better defined circulation across the Iceland-Scotland Ridge western branch of Norwegian Atlantic Current [cm/s] 20 [cm/s] 9-km2-km

12. Main differences: Norwegian Atlantic Current convergence along the northern Norway EGCC and NCC (connection to the Baltic outflow) Irminger Current inflow into the Iceland Sea better defined circulation across the Iceland-Scotland Ridge eddies in the central Norwegian Sea Surface (0-5 m) annual mean EKE in the Nordic Seas [cm 2 /s 2 ] 9-km2-km

13. Main differences: current recirculation / shear around 80°N boundary current inflow over the Yermak Plateau circulation over the Greenland shelf / EGCC Annual mean velocity at 0-223 m - Fram Strait region 20 [cm/s] [cm/s]

14. Annual mean EKE in the upper 223 m – Fram Strait region Complex recirculation / eddy activity at Fram Strait - challenging for measuring / oomodeling mass and property exchanges Long-lived eddies present in the region [cm 2 /s 2 ]

15. 1980 annual mean velocity at 0-183m in the Labrador Sea Maslowski et al. (2008) significantly stronger and better defined currents in the 9-km model (note the odifferent vector scales) recirculation AW inflow into the southern Labrador Sea – challenging for models [cm/s]

16. 1983 annual mean velocity at 0-183 m in the Labrador Sea stronger and reaching further north coastal flow along west Greenland main recirculation further north better defined Labrador Current well-resolved coastal currents in the western Labrador Sea [cm/s]

17. [cm 2 /s 2 ] (C) Horzontal distribution of EKE deduced from the surface drifters data relased in North Atlantic Ocean and Labrador Sea during 1993-1997 (Cuny et al. 2002) (A) Surface (0-5m) annual mean Eddy Kinetic Energy (EKE) 2-km model (B) Surface (0-5m) annual mean Eddy Kinetic Energy (EKE) 9-km model (A)(B)(C)

18. Conclusions Significant improvement in resolving coastal and boundary current flow at high resolution EKE increases with resolution but mainly along coasts and slopes Total mean kinetic energy 2-3 times higher in 2-km vs. 9-km model Eddy-resolving model – a tool for synthesis and guiding data collection