• What are the characteristic modes of natural variability in the physical and biological processes in the CCS and CGOA?

• What are the most important processes affecting population distribution and abundance of the species listed in Tables 3 and 4 in the CCS and CGOA and how do the varying strengths of these processes affect the response? Physical rocesses might include the intensity, timing and persistence of upwelling, mixing, cross-shelf and alongshore transport, stratification, temperature and the timing of seasonal transitions. Biological processes include zooplankton production and juvenile salmon mortality caused by predators. Other factors, including the genetic composition of populations and how it varies spatially and temporally may also be explored.

• Is there evidence for linkage between processes (both physical and biological) occuring at shorter (event-to-seasonal) and longer (interannual-to-interdecadal) time scales?

Long-term monitoring will provide a link between the intensive, process-oriented studies from the CGOA and CCS sites and the larger-scale, longer period climate variations. Monitoring in the Northeast Pacific GLOBEC program will proceed differently than that done for the U.S. GLOBEC program in the Northwest Atlantic. In the Northeast Pacific program, the frequent sampling of multiple cross-shelf transects (from Prince William Sound to the Monterey region), coupled with observations from moorings, drifters, floats and ships-of-opportunity, will be the analog of the broad-scale shipboard surveys that are used to monitor conditions on Georges Bank in the Northwest Atlantic. The CGOA and CCS ecosystems are highly advective. Following well-defined populations for extended periods, as is done on Georges Bank, will be difficult.

Monitoring of the Northeast Pacific will include the collection of data from satellites, enhanced volunteer observing ship (VOS) programs, coastal stations, selected cross-shelf transects, nearshore and offshore buoys, subsurface moorings, near-surface drifters, and perhaps other technologies. We recommend that specific biological and physical observations be obtained at the basin (gyre) scale. This will enable the connection to be made between the large scale forcing and the regional process studies in the CGOA and CCS. Figure 10 provides a cartoon of the types of observations needed to make that connection. These larger-scale observations of the circulation and biology of the gyre are critical in connecting the coastal regions of the CCS and CGOA to basin-scale forcing and in understanding the covariability between these regions.

Regular occupation of a few selected transects is key to monitoring the ocean conditions and variability in the coastal regions of the Northeast Pacific. Satellite sensing and moored instrumentation are excellent tools for some observations, but many biological quantities require ship sampling. Quarterly or bimonthly sampling with large oceanographic vessels to 100-200 km offshore would be supplemented by more frequent sampling (perhaps monthly, or more frequently during critical times [e.g. spring bloom; spawning events; juvenile entry into coastal waters]) of the more nearshore end (out to perhaps 20-25 km) of these transects by smaller vessels. Frequent cruises on established lines will be needed for calibrating indirect measures from remote-instrumentation and to directly sample ecosystem components such as zooplankton abundance, species composition, abundance of juvenile salmon and their competitors and predators, that cannot be collected remotely. Observations from several transects in the Northeast Pacific will also help to relate the biological and physical observations from moorings to larger regions.

The requirement for frequent sampling of the transects places some constraints on the potential number and location of the transects. Nearby logistical support by marine field stations or laboratories could warrant transects in the following regions (South to North): Monterey (MLML; MBARI; PFEG), Point Reyes/Arena (Bodega Bay Laboratory of UC Davis), Mendocino Region (Humboldt State), Coos Bay (OIMB), Newport (NMFS; Hatfield Marine Science Center), Columbia River (Astoria NMFS Lab), La Perouse Bank (ongoing Canadian site), west coast of Vancouver Island (Bamfield Marine Station), Line P (irregular Canadian occupation by IOS), Auke Bay (University of Alaska; NMFS Lab), Prince William Sound/Seward (University of Alaska GAK line), and Kodiak Island (FOCI line 8 of the NMFS/PMEL). Establishing routine monitoring of physics and biology from any of these stations would be a major improvement over current assessment efforts. For reasonable along-coast coverage, the sites probably most appropriate and able to undertake routine transect monitoring are Monterey, Pt. Reyes/Arena, Coos Bay, Newport, Auke Bay and Seward (not including the Canadian sites).

Detailed U.S. GLOBEC process studies (described in a later section) should be done at several of these regional transects. The highest priority monitoring transects are those tied to the process studies. Thus, we would anticipate that a single cross-shelf transect of stations would be sampled at least quarterly, perhaps bimonthly, throughout the 7 years from large oceanographic vessels, with more frequent sampling nearshore (by smaller vessels) during that region's process-study year, and perhaps during the spring and summer of all years. Observations taken at the transect sites should include hydrography, currents, net sampling of zooplankton, hydroacoustics, purse-seining (perhaps needed only in spring and summer), and ancillary observations. Drifters should be released as frequently as possible to describe the spatial and temporal variability in coastal ocean circulation. Because it is important that the observations be obtained consistently through time and in each of the various regions (e.g., transect sites), we recommend a core (minimum) set of measurements that should be obtained at each monitoring transect (Table 5). Consistency in sampling is crucial to making cross-regional comparisons and to facilitate time-series analysis. With the exception of the sampling of the salmon juveniles and other pelagic fishes (e.g., forage species) by purse-seining , all of the observations in the core program can be obtained from standard oceanographic research vessels. Sampling of the fish will require specially equipped vessels, either from the NOAA research fleet or chartered fishing vessels. Issues that arose in specifying the core monitoring measurements that should be addressed by the individual groups proposing to undertake transect monitoring and some quidelines on priorities for these issues are provide in Table 6.

Observations of solar radiation, wind speed and direction, atmospheric pressure, air and water temperature, humidity, salinity and sea level height should be continued at existing monitoring sites in the North Pacific, and initiated at new locations near the regional process study sites. Some extensive monitoring is ongoing already. The TOGA/TAO array along the equatorical Pacific provides a link to the ENSO, basin-scale variability. The CLIVAR/GOALS program will continue to monitor and predict ocean variability in the central equatorial Pacific, perhaps extending to the whole basin. This activity should feed into the monitoring since we want to understand the basin-scale connections of the regional-scale monitoring.

Nearshore and offshore surface buoys should be used to measure solar radiation, wind speed and direction, atmospheric pressure, sea-surface elevation, air temperature, hydrographic conditions at several depths, upper-ocean velocities, fluorescence and plankton abundance by acoustic backscatter or optical plankton counters at a few nearshore (1-10 km offshore) and offshore (100-200 km offshore) locations.

In order to monitor biological and physical conditions, several (preferably 3) deep water subsurface moorings should be located as a transect within the Alaskan Gyre. They should be located (1) off the shelf in deep water, adjacent to the CGOA study region, (2) near the center of the Alaskan Gyre, and, (3) at an intermediate location. Brodeur and Ware (1992) and Brodeur et al. (in press) observed that zooplankton abundance increased most markedly over decadal scale periods along the margins, rather than in the center, of the Gulf of Alaska. Multiple deep water moorings, spanning the central gyre to the margin, would provide the data necessary to document productivity shifts that might occur during the program. The emphasis of these moorings should be to provide first, long-term biological observations (e.g., using optics, fluorometers and/or acoustics), and second, physical observations. Moreover, the moorings would provide data required to assure that simulations of the circulation and ecosystem of the Northeast Pacific are accurate.

Additional subsurface moorings measuring water temperature, salinity, velocity, fluorescence, light transmission, solar radiation and zooplankton biomass (using acoustics or optics) should be deployed for the full 5-7 year Northeast Pacific program at a few key sites in each of the regional study regions. These moorings would complement additional "mobile" moorings deployed in the three study regions during the period of the process-oriented field investigations. Thus, just as an example, there might be six instrumented subsurface moorings (two off central California; two off Oregon; and two off the CGOA study site) maintained for 5-7 years. One of the latter could be the innermost "deep-water" mooring of the Alaskan Gyre transect.

One challenging aspect of this program is to link regional-scale variability revealed in process-oriented and monitoring studies conducted in the CGOA and CCS to large/basin-scale oceanic variability that likely forced the biological changes that have occurred over the past several decades. Given the size of the Northeast Pacific, the connection between basin and coastal biophysical dynamics must be accomplished by modeling. Judicious monitoring of essential oceanic features, however, is essential to assure that simulations are accurate. Further, such observations can also be assimilated into models to nudge their output closer to observed conditions.

We envision a strategy which includes both Eulerian and Lagrangian measurements and suggest the following approaches to monitor the large-scale oceanography of the Northeast Pacific. First, a series of three deepwater moorings from the margin to the center of the Alaskan Gyre have been described previously. Second, the Alaskan Stream is a pulse-point in the North Pacific circulation, where measurements provide an index of the strength and variability of the subarctic gyre. The Alaskan Stream current is spatially most stable and narrow west of Kodiak Island (Reed et al. 1991); several subsurface moorings located across the stream there could monitor current velocity, vertical structure and water properties. Third, a recently developed technology, Profiling Automated Lagrangian Circulation Explorers (PALACE) floats, may be appropriate for providing temperature and salinity fields for assimilation into models. These floats collect vertical profiles of temperature and salinity, and can perhaps be modified to measure other parameters (e.g., fluorescence). They are programmed to reside on a density surface (perhaps at 800 to 1000 m), from where they periodically ascend to the surface collecting environmental information. By remaining deep most of the time, they are not quickly advected away from their initial position. They remain at the surface (order 16-24 hours) long enough to transmit their profile data and position via ARGOS transmitters; their estimated lifetime is two years. Riser (1995) estimates that ca. 60 floats would provide enough data to produce objective maps of the subsurface thermal and motion fields at ca. 500 km resolution for the entire subarctic Pacific. Fourth, satellite-tracked buoys drogued at appropriate depths could provide information on circulation and mesoscale features. Observations of the entire eastern North Pacific basin using these or similar methods are needed to place the regional process studies within the larger scale context of the circulation and water mass characteristics of the Alaskan Gyre, subtropical Gyre, and the bifurcation of the west wind drift as it nears North America.

It is unclear exactly how a monitoring program of the Northeast Pacific should be done--e.g., frequency of monitoring, state variables and rates to be monitored, and the most appropriate methodologies. To better answer these questions and to provide advice for more extensive monitoring in the future, we recommend that monitoring of the Northeast Pacific begin with a relatively modest pilot study. Despite the modest level of initial support for monitoring anticipated initially, we envision that this pilot monitoring include at least two transect lines: one north of the west-wind drift (CGOA), and one south of the west-wind drift (CCS). Tables 5 and 6 describe the minimum set of core measurements that would constitute an acceptable pilot monitoring program. Potential investigators are encouraged to propose ancillary measurements, in addition to the minimal set, that can be readily collected in an efficient and cost-effective fashion. Efforts should be made to cross-reference monitored quantities to other measurement efforts, past or present. Where feasible, monitoring data should be made available in real time, so that opportunistic studies can be conducted within a known physical and biological context. Further, consideration of the statistical power of the monitoring program, including some measure of the statistical properties of estimators derived from the monitored quantities, is advised.