Ocean-Sea Ice Dynamics as Simulated by a Circulation Model
3. Model design and experimental set-up
3.1 Ocean-sea ice model and forcing
In order to examine the effects of freshwater discharge on high-la,titude ocean dyna- mies, we utilize a coupled ocean-sea ice model. The ocean model is set up on t h e base of the hydrostatic Geophysical Fluid Dynamics Laboratory (GFDL) primitive equa- tion model MOM-2 (Pacanowski 1995), employing the implicit free-surface inethod by Dukowicz & Smith (1994). T h e model domain Spans the Arctic Mediterranean i . e . the Arctic Ocean proper and the Nordic Seas) and the Atlantic Ocean north of approximately 20's. The model is formulated on a rotated grid t o avoid the singula- rity of geographical coordinates a t the pole (Fig. 3). It has a horizontal resolution of about 100 km and 19 non-equidistant levels in the vertical. Using the flux-corrected transport (FCT) algorithm for tracer advection, explicit diffusion is set t o Zero (Ger- des et al. 1991). The ocean model is coupled t o a dynamic-therrnodynainic sea ice model with viscous-plastic rheology, which is defined on the sa.me horizontal grid (Harder 1996). We emphasize t h a t the ocean-sea ice model is fully prognostic, i.e.
no diagnostic or restoring terms are added t o the conservation eqmtions. A detailed description of the model can be found in Prange (2003).
T h e ocea,n-sea ice rnodel is forced by atmospheric fields, comprising 2 m-temperature, 2 m-dewpoint temperature, cloud Cover, precipitation, wind speed, and wind stress.
Except for daily wind stress, all forcing fields are monthly varying. The a,timospheric fields a.re derived from a validated 15 year (1979--1993) set of assimilated data pro- vided by the reanalysis project of the European Center for Medium-Range Weather Forecasts (ECMWF). The data have been processed t o construct a "typical" year, i.e. a mean annual cycle with daily fluctuations superimposed (Roeske 2001). In addition t o atmospheric forcing, t,he ocean-sea ice system is forced by river runoff and Bering Stra,it inflow. Fourteen Arctic rivers (Fig. 2) are implemented as well as some additional ungauged runoff from the Arctic coastlines (see below). For the Atlantic portion of the model domain, the eight 1a.rgest rivers are included as well as the freshwater supply from Hudson Bay and the Baltic Sea.
3.2 Experiments
Three experiments are performed, differing in freshwater discharge into the Arctic Ocean. T h e discharge distributions refer t o present-day (control run), mid-Holocene ( ~ 7 ka), arid early Holocene ( ~ 1 0 ka). The effect of variable freshwater input is isolated from other processes by applying the same atmospheric forcing in all expe- riments. In this context, it is worth noting that the real early-to-mid Holocene mean wind forcing was probably not too far away from our "typical year"-forcing. A recent analysis of alkenone-derived sea surface tempera,tures in the North Atlantic realm indicates that the mean atmospheric circulation of the early-to-mid Holocene was shifted t o a high index phase of the North Atlantic Oscillation (Rimbu et al. 2003), which bears similarities with the unusual high index phase of the period 1979-1993.
3.2.1 Present-day (control run)
A climatology for rnonthly discharge of the largest Arctic rivers (Fig. 2) has been constructed by Prange (2003). The river water inflow is implemented in the model as mass fluxes with Zero salinity. Based On various estimates (e.g., Plitkin 1978; AANII 1990) some ungauged runoff is added during summer (June-September). Along the coastlines of the Barents, Kara, and Laptev Seas a n additional freshwa.ter inflow of 520 km3 y r l is equally distributed. Ungauged runoff from the eastern Siberian, North American and northern Greenland coasts is smaller: a total of 180 km3 y r l is added in these regions.
Runoff from the Norwegian coast is included as a consta,nt freshwater inflow. Moreo- ver, monthly varying inflow of Pacific water through Bering Strait is implemented based on direct measurements (see Section 2.1). Aside from being a source term in the Arctic Ocean freshwa,ter budget, the Bering Stra,it inflow is associated with a heat supply during the summer months. T h e temperature rises up t o 4'C in Sep- tember, while winter temperatures (December-May) are at, freezing for the salinity
190
(Becker 1995; Roach et al. 1995).
3.2.2 Experiment 10 ka
Even though there is geological evidence for increased freshwater runoff froin the continents into the Arctic Ocean a t the early Holocene, a quantification is difficult and subject t o considerable ui~cert~ainty. Based on geological studies, summarized in Section 2.2. we estimate the 10 ka freshwater discharge t o force the Arctic Ocean.
We presume river water inflow from the Norwegian coast as well as into the Barents and Kara Seas to be 25 % higher than today. Extreme runoff, a,mounting t o three times the modern one, is assumed from North American and northern Greenla.ncl coasts. To implement, these changes in the ocea,n model, we increase the freshwater input t o each coast.al grid cell by the rcspcctive percentage.
Concerning the freshwater budget of the Arctic Ocean, the enhanced runoff is partly con~pensated by a lesser inflow of low-saline Pacific water. We assume t h e 10 ka Bering Strait inflow t o be half of the modern one.
3.2.3 Experiment 7 ka
For the freshwater forcing of the Arctic Ocean a.t 7 ka we apply the following dischar- ge distribution, based on geological evidence presented in Section 2.3. Runoff from the Norwegia,n coast as well as into the Kara Sea is 25 % lower than today. whereas the freshwater flux into the Barents and East Siberkn Seas is enhanced by 25 %. The largest change occurs in the Lapt,ev Sea. Here the total river discharge is doubled.
Table 1 summarizes the freshwater forcings for t h e three experiments.
4. Results
freshwater input t o affect the large-scale Atla,ntic thermohaline circulation (Gerdes& Köberl 1995; Prange & Gerdes 1999).