Much of the variability in the physical environment in the Gulf of
Alaska results from large scale atmospheric phenomena. Global patterns
in the upper level atmospheric pressure generate climatic conditions
that include an annual cycle in the number of low pressure centers
traversing the region [ Niebauer, 1988]. The consistent passage
of storms along the Aleutian Island chain (the Aleutian Low) dominates
wintertime atmospheric circulation. Interaction of frequent storms
with the mountainous coastline results in a high precipitation rate
(>200 cm yr
) along the coastal region. The discharge rate of
freshwater reflects seasonal variations in air temperature,
precipitation, runoff and storage from the previous winter [
Royer, 1982].
Along the Alaska Peninsula, along with a deep (>250 m) sea valley, a
high, nearly continuous mountain chain exists (Figure 1). The
mountains perturb regional winds so that in Shelikof Strait proper
down-gradient winds are common [ Schumacher et al., 1989], and
the winds over coastal waters west of Kodiak Island are altered for
60 km offshore [ Macklin et al., 1993].
The dominant circulation feature is the Alaska Coastal Current (ACC),
a distinct flow that only 20 years ago was unknown. FOCI research has
elucidated many of the characteristics of the ACC, which extends for
>1500 km along the coast of Alaska [ Reed and Schumacher, 1987].
This is one of the most vigorous coastal currents in the world with
speeds typically between 25 and 100 cm s
[ Stabeno et al.,
1995]. Volume transport results from the addition of freshwater along
the entire coastline and is perturbed by the alongshore wind through
both confinement of the freshwater and alteration of coastal sea level
[ Schumacher and Reed, 1980; Royer, 1981; Reed and
Schumacher, 1981]. The observed mean transport in Shelikof Strait is
0.80
10
m
s
; wind forced pulses
exceed 3.0
10
m
s
[ Schumacher et
al., 1989; Stabeno et al., 1995]. Wind-driven fluctuations
within the strait proper are greater than those over coastal waters
east of Kodiak Island due to the topographic effects on the winds
[ Stabeno et al., 1995]. Differential Ekman pumping may amplify
this mechanism within the strait proper [ Reed and Schumacher,
1989]. Estimates of net volume transport computed from water property
observations collected between 1985 and 1992 have a mean of
0.66
10
m
s
[ Reed and Bograd, 1995].
In Shelikof Strait, horizontal density gradients and vertical shear in
the mean flow create the baroclinic instability, evident in satellite
images [ Vastano et al., 1992; Schumacher et al., 1991] and
in analysis of current records [ Mysak et al., 1981], which
dominates flow patterns and generates eddies [ Schumacher et al.,
1993]. The ACC does not span the sea valley, and estimates of
coherence become insignificant for separations >10 km [ Reed and
Schumacher, 1989b; Bograd et al., 1994]. Estuarine-like flow
also exists, with warmer more saline water from the continental slope
entering on the southeastern side of the valley [ Reed et al.,
1987]. The ACC bifurcates east of Sutwik Island; one branch continues
along the Alaska Peninsula and the other flows seaward through the sea
valley [ Schumacher et al., 1989]. Since 1986, 51 satellite
tracked buoys (drogued at 40 m) were deployed in the study area during
spring near Cape Kekurnoi. To date, 25% of the buoys continued along
the Peninsula. The remainder moved seaward past the Semidi Islands,
most, however, traveled shoreward (between
157
and 158
W)
and joined the flow along the Peninsula. Only 25% of the buoys left
the shelf permanently and became incorporated in the Alaskan Stream
[ Stabeno and Reed, 1990].
FOCI research has found and elucidated the dominant circulation and mesoscale features: the ACC, eddies generated by baroclinic instability, and an estuarine-like flow of slope waters into the sea valley. These features must be simulated in any numerical model of the region. Further, results show that the timing and location of hatching determines whether larvae enter an eddy, or are transported with either the slow-moving coastal flow or the rapid ACC. Modeling studies [ Stabeno et al., 1995] suggest that the location of late larvae varies greatly year to year depending on advection. The phasing between biological and physical processes determines transport of larvae and presumably their eventual recruitment.