Appendix 1 -- Climate Context

The hypothesized changes for the Gulf of Alaska and Sub-Arctic Pacific, and for the Bering Sea, expected to accompany greenhouse gas-induced climate change are summarized in Tables 1 and 2, respectively (pp. 28-29). The object of these tables, and their attendant discussions, is to summarize the results of thought experiments using the relatively crude projections now being made by the climate community. A special attempt was made to be at least internally consistent with the speculations itemized below. It is important to note that these speculations are based on the assumption of a secular warming of the atmosphere over the North Pacific, especially at higher latitudes. The present climate is punctuated by decadal-scale variations or regime shifts of substantial amplitude, as is discussed in the following section. It is highly likely that these types of variations will continue, and at any particular time, their effects can swamp the changes associated with global warming.

Gulf of Alaska and Subarctic Climate Change Scenario

With global climate change (for this scenario, a warming), it is predicted that the rate of temperature change at high latitudes will be greater than at lower latitudes. This would serve to decrease the meridional thermal gradient that would result in a more sluggish atmospheric circulation. This is the primary assumption that will be used to describe the subsequent affects on the marine ecosystem.

With a decrease in the atmospheric meridional thermal gradient, storm intensity would decrease and the storm tracks would be shifted northward (Table 1). The zero wind stress curl line that separates the subarctic and subtropical gyres would shift poleward, as would the bifurcation of the Alaska and California Currents. Previous results have suggested that a northward shift of westwind drift could result in an intensification of the currents of the Alaska gyre. A reduction in the wind stress curl would result in a decrease in transport, but the split between the southward flowing California Current and northward flowing Alaska Current is unknown. Any change in northward transport will effect the poleward heat flux.

Decreased storm activity over the GOA is likely, especially in winter. The ramifications for the central Gulf are less upwelling and less wind mixing. In terms of surface heat fluxes, there will likely be decreases in sensible heat flux, but unknown changes in evaporation. The net effect will be a shallower mixed layer depth and warmer mixed layer temperatures.

An increase in absolute humidity in the atmosphere would likely result in increased coastal precipitation. At present maximum freshwater runoff occurs in the fall, since the high precipitation rates of winter are tied up in snow. As air temperatures warm, more of the winter precipitation would fall as rain. The warmer temperature could decrease the size of the coastal glaciers, thus increasing runoff at least in the near future. Maximum runoff would likely occur in winter coinciding with the maximum wind stress. The increase in freshwater runoff and warming would increase the stratification along the coast. The decrease in upwelling and warming in the central North Pacific would have competing effects on stratification, but the increases in temperature would probably dominate.

Decreased wind stress along the coast would result in a weaker Alaska Coastal Current (ACC). The increase in freshwater would add to the baroclinic structure, but the reduction in wind stress would weaken the confinement along the coast and thus decrease the baroclinic gradient. Thus any change in the number and intensity of eddies formed on the shelf is unknown. The effect of climate change on the eddies at the shelf break is also unknown.

Diminished downwelling on the shelf would tend to reduce the cross-shelf flux of nutrient-poor water at upper levels, but this mechanism would be counteracted by less wind mixing, with an unknown net effect on nutrient concentrations over the shelf. The reduction in the strength of the ACC would effect the transport of nutrients on the shelf. The timing of the spring bloom would probably be earlier since the water would be warmer, and the formation of the spring mixed layer could be earlier.

Bering Sea Climate Change Scenario

I. Atmosphere

Surface air temperatures (Increase) -- There is a strong consensus among the climate community that global warming will be accompanied by enhanced surface temperature rise at higher latitudes, largely due to positive feedback effects associated with less snow cover and sea ice and hence lower albedo (Table 2). The Bering Sea region is at a high enough latitude that it is expected to warm significantly.

Storm intensities (Decrease); Storm frequencies (Increase) -- The enhanced warming at high latitudes will have the effect of reducing the meridional gradient in air temperature at mid-latitudes; this reduction in baroclinity will lead to weaker storms (Table 2). The zone of maximum baroclinity will tend to shift poleward. Since the storms now track across the North Pacific in a mean sense south of the Bering Sea (Anderson and Gyakum 1989), a northward shift of this track will tend to cause a higher incidence of storms in the Bering Sea (as suggested by the GCM simulations of Hall et al. 1994).

Sea level pressure (Decrease in N. Bering); Southerly wind (Increase); Wind stress curl (Unknown, competing effects) -- Sea level pressure is expected to be significantly lower in the Arctic, as suggested by GCM simulations and from hydrostatic considerations assuming relatively strong warming in the Arctic (Table 2). This effect will tend to cause lower sea level pressure in the northern portion of the Bering Sea, and more winds from the south, especially in the vicinity of Bering Strait. It is unknown how the mean wind stress curl is likely to change, since it is net effect of the storms that largely determines the curl, and there is likely to be compensation between changes in the frequency and intensity of the storms. It is probable that the location of the maximum in the curl will shift poleward from its present position along the Aleutians (e.g., Bond et al. 1994).

Humidity, Precipitation, Fresh water runoff (Increase) -- Warmer air temperatures allow higher water vapor concentrations and would be expected to lead to greater precipitation amounts, as also indicated by GCM results (Table 2). Greater fresh water runoff is then also expected. These effects are liable to be more pronounced in the northern Bering Sea.

II. Circulation and Transports

Alaskan Stream (Probable decrease) -- The Alaskan Stream (AS) is expected to decrease in intensity because enhanced rises in sea level (Gregory 1993) and reduced wind stress curl is expected in the central North Pacific. These effect may be tempered by greater storminess in the Gulf of Alaska (See Table 1).

Near Strait Inflow (Decrease) -- This primary source of inflow into the Bering Sea is expected to decrease due to a decrease in the intensity and the westward extent of the AS, and perhaps also due to changes in Sverdrup transports associated with a weakening and northward displacement of the wind stress curl.

Bering Slope Current (Decrease) -- This flow is expected to decrease because a reduced AS will supply less mass transport through the Aleutian passes, and because of reduced baroclinity between the deep basin and shelf water masses. As with the Near Strait inflow, it should also tend to weaken due to reduced deep Bering basin Sverdrup transport.

Kamchatka Current (Decrease) -- The Kamchatka current will decrease if the Near Strait inflow is less and the Sverdrup forcing is reduced, as suggested above.

Bering Strait Outflow (Unknown, competing effects) -- The meridional component of the wind, which is expected to be more from the south, would tend to cause more outflow. This effect is liable to be counteracted by a decrease in the steric difference between the Bering Sea and Arctic Ocean.

Unimak Pass inflow (Unknown) -- The portion of the Alaska Coastal Current (ACC) that remains on the shelf flows through Unimak Pass. The ACC is largely due to the set-up associated with the along-shore winds. A change in the storminess of the Gulf of Alaska would result in a change in the total transport by the ACC, but there may also be changes in the fraction of the ACC that stays on the shelf to flow through Unimak Pass.

Shelf coastal current (Unknown) -- This current originates from the inflow through Unimak Pass and therefore its change is uncertain. It may tend to be enhanced by greater freshwater runoff from the west coast of Alaska.

III. Hydrography

Sea level (Increase) -- A significant rise in sea level is expected due to the steric effect and melting of glaciers and ice packs.

Sea surface temperature (Increase) -- The SST is expected to warm substantially due to warmer air temperatures and greater downwelling longwave radiation from a more humid atmosphere. There may be a positive feedback associated with warmer SST's promoting the breakup of stratus cloud decks in summer months and hence greater solar insolation.

Shelf bottom temperature (Increase) -- The cold water that is presently found near the bottom on the shelf is formed by convective cooling due to melting of ice advected from the north. With less ice created and advected in the northern portion of the Bering Sea, this mechanism will be suppressed.

Basin stratification (Increase) -- The ocean warming should be enhanced near the surface, promoting the static stability.

Shelf stratification (Unknown, competing effects) -- The change in static stability over the shelf is unknown because both the near surface and bottom waters are expected to warm.

Mixing energy (Decrease) -- Much of the mixing is accomplished by storms, which are expected to be more frequent but weaker. Since the strongest storms cause a disproportionate share of the mixing (mixing rate varies roughly with the wind speed cubed), a net reduction is expected in the wind energy available for mixing. At least over the deep Bering basin, vertical mixing is also liable to be suppressed by increased upper-level static stability. An important aspect of mixing is its seasonality, especially as it applies to nutrient supply and primary productivity, but changes in this seasonality are uncertain.

Eddy activity (Unknown) -- The mechanism(s) responsible for eddy generation in the Bering Sea are not well understood, especially in its eastern portion where the mean currents are weak and the eddy activity is high. The flow along the eastern slope and shelf break is expected to decrease as noted above, but it is not evident how this change would impact the eddies. Increased river flows may generate and sustain eddies in areas of the Gulf of Alaska and the Bering Sea. This should have the effect of entraining nutrients and increasing local productivity.

Shelfbreak nutrient supply (Decrease) -- This is expected to decrease because of reductions in the slope currents and perhaps the wind-driven upwelling along the shelfbreak during the summer. These effects will tend to be augmented by an increased supply of low nutrient water from estuaries.

IV. Sea Ice

Extent, Thickness, Brine rejection (Decrease) -- The ice in the Bering Sea is generally confined to shelf regions. It is generally formed in coastal regions and the southward march of its leading edge is mostly due to advection. Its extent over the shelf is expected to be substantially reduced for two reasons: warmer air temperatures and less wind from the north, especially in winter. A shorter freezing season will tend to produce thinner ice. Less ice formation implies a decrease in brine rejection.