1. Numerical modeling of Atlantic and Pacific waters dynamics Elena Golubeva Institute of Computational Mathematics and Mathematical Geophysics Siberian Branch of the Russian Academy of Sciences 21-23 October, 2009, Woods Hole Oceanographic Institution

2. ICMMG coupled ice-ocean numerical model The ICMMG z-coordinate ocean model started its history in the late 1990-s when it was designed as z-coordinate analogue of the sigma-coordinate model of the World Ocean (Kuzin,[1982], Golubeva et al.[1992], Golubeva,[2001], Golubeva and Platov,[2007]). The model was formulated using some assumptions: the hydrostatic approximation, the Boussinesq approximation, and the rigid-lid approximation at the sea surface, Laplacian viscosity in the horizontal velocity equations. The model dynamics are resolved by separating barotropic and baroclinic modes. The barotropic mode is described in terms of a stream function. The discretization of the model equations is conducted using curvilinear orthogonal coordinates and based on the splitting technique applied for coordinates and physical processes. QUICKEST (Leonard,[1992]) is used in the latest model version for the T-S advection. Mixed layer parameterization: –Vertical adjustment based on the Richardson number The sea ice model – CICE 3.14 ( an elastic viscous-plastic model) W.D.Hibler, [1979], E.C. Hunke and Dukowicz,[1997], C.M.Bitz, W.H.Lipscomb,1999, E.C. Hunke,2004

3. Arctic – North Atlantic model domain Grid The model grid has1x1 degree resolution in the North Atlantic. At 65N, the North Atlantic spherical coordinate grid is merged with the polar grid, which is a result of the spherical coordinate rotation and reprojection of its hemisphere onto the area above 65N (see Murray, [1996] for details). The horizontal spacing in the polar grid varies from 50 to 34 km. Maximum model resolution is in the vicinity of the North Pole. Vertically, the grid has 33 constant layers, with a higher resolution near the surface. The minimum shelf depth is 50 m. Open bondaries In the Bering Strait, an open boundary is introduced with a prescribed transport of 0.8 Sv. The river run-off transport is prescribed for most of the significant Arctic and Atlantic rivers Southern boundary at 20S: linear distribution between zero transport in the west point and the summarized transport value in the western point. T,S data are used for all incoming flows and free- following condition for outcoming flow.

4. The numerical experiments according to AOMIP reveal the effect of the ice cover, water circulation, and thermohaline structure of the Arctic Ocean on variations in the state of the atmosphere. During the second part of XX we found the warming and cooling periods in the Atlantic water layer, the freshwater accumulation regimes in the Canadian Basin and freshwater flow through the Fram Strait and Canadian Archipelago straits, a reconfiguration of the water circulation of the surface and intermediate layers of the ocean, a shift in the boundary between Atlantic and Pacific waters, a significant reduction of the ice area. E. N. Golubeva and G. A. Platov, Numerical Modeling of the Arctic Ocean Ice System Response to Variations in the Atmospheric Circulation from 1948 to 2007 ISSN 0001-4338, Izvestiya, Atmospheric and Oceanic Physics, 2009, Vol. 45, No. 1, pp. 137–151. © Pleiades Publishing, Ltd., 2009.

5. In the Amerasian Basin, the Pacific waters generate upper thermohaline waters occupying the intermediate layer between the relatively fresh Arctic surfacewaters (0–100 m) and the warm and saline waters of the North Atlantic origin, thus being a kind of interlayer opposing the motion of Atlantic waters toward the ice surface. In the past, the existence of these waters was clearly traced up to the Amerasian Basin boundary(the Lomonosov Ridge). According to the 1993 measurement data for the southern part of the Canadian Basin, there was an eastward shift of the boundary between the Atlantic and Pacific water masses [McLaughlin et al.,1996]. The initial position of this boundary near the Lomonosov Ridge was shifted into the Mendeleev Ridge. In the EWG measurement data for the period 1948–1993 an analysis of the database has demonstrated that the sharp disappearance of features of the Pacific water mass in the Makarov Trough and northern part of the Nansen Basin occurred between 1984 and 1985.

6. 200м. 1960 - 1974

7. 200 м. 1995

8. a shift in the boundary between Atlantic and Pacific waters 1970 1990

9. 1965 Pacific water Atlantic water 50m

10. Pacific Water tracer distribution in 0-50 m layer 1970 1980 1990 2000

11. Atlantic water tracer Pacific Water tracer 1970 1980 1994 0m 300m

12. Pacific waterAtlantic water 2000 2005

13. What is the problem? Two regimes of the ice drift –yes Two regimes of the surface circulation????

14. Can we stop warming in Arctic Ocean model? Experiment 1: 1958-1970--->1960  The same result Experiment 1:Repeated 1965

15. 1965 exp. Velocity distribution at the depth of 200m 10 years 20 years

16. 1965 exp. Velocity distribution at the depth of 200m 30 years 40 years

17. 50 m 200 m 55 years

18. initial40years

19. What’s wrong with numerical model? What’s wrong with experimental design? Can we switch our model between different circulation regimes? Thank you!