The Bering Sea in 1998: The Second Consecutive Year of Extreme Weather-forced Anomalies

In 1998, anomalous conditions in the Bering Sea included elevated heat content of the water, cross-shelf advection of zooplankton and larval fish, major changes in the structure of the zooplankton community, and an unprecedented second observation of a large-scale bloom of the coccolithophorid phytoplankton, Emiliania huxleyi. Some of these anomalies appear to be related to the unusual weather patterns of 1997 and 1998, while the causes of others remain unknown. 
 
The Bering Sea is located in the northernmost part of the North Pacific Ocean, and its broad eastern continental shelf constitutes approximately 44%of its area. Because Pacific water must pass through the Bering Sea before entering the Arctic, climatic events in the Bering affect heat and biogeochemical transport to the Arctic. The Bering Sea, in particular its broad eastern shelf region, is also the site of some of the worlds major fisheries. It contributes over half of the U.S. fishery production, with a commercial catch worth one billion dollars in 1997.

PAGES 561,565-566 In 1998, anomalous conditions in the Bering Sea included elevated heat content of the water, cross-shelf advection of zooplankton and larval fish, major changes in the structure of the zooplankton community, and an unprecedented sec ond observation of a large-scale bloom of the coccolithophorid phytoplankton,£>77///an/a huxleyi.Some of these anomalies appear to be relat ed to the unusual weather patterns of 1997 and 1998, while the causes of others remain unknown.
The Bering Sea is located in the northernmost part of the North Pacific Ocean, and its broad eastern continental shelf constitutes approxi mately 44% of its area. Because Pacific water must pass through the Bering Sea before enter ing the Arctic, climatic events in the Bering affect heat and biogeochemical transport to the Arctic. The Bering Sea, in particular its broad eastern shelf region, is also the site of some of the worlds major fisheries. It contributes over half of the U.S. fishery production, with a commercial catch worth one billion dollars in 1997.

Two Major Investigations
During 1998, two major scientific programs were active on the southeastern Bering Sea shelf. One, the NSF-sponsored Inner Front Project, focused on a series of sites near the 50 m isobath (the inner front) from Slime Bank to Nunivak Island.The other, the NOAA-sponsored Southeast ern Bering Sea Carrying Capacity (SEBSCC) proj ect, concentrated on conditions conducive to fisheries production over the middle and outer shelves. In 1998, these two studies used a combi nation of shipboard measurements (9 cruises between February and October), moorings (11 sites), and satellite imagery to measure physical properties, nutrient concentrations, and biologi cal parameters, including standing stocks and production of phytoplankton,and the distribu tion and abundance of zooplankton, fish and seabirds (Napp & Hunt,accepted).
Variation in the extent, duration, and timing of sea-ice cover over the continental shelf of the eastern Bering Sea is important in determining both physical properties and biological process es subsequent to ice melt [Stabeno et ai, 1999;D. A.Stockwell et al.,submitted manuscript, 1999]. Events in 1998 indicated that weather-driven interannual differences in winter sea ice cover can influence the heat content of water exported from the eastern Bering Sea shelf to the Arctic. Usually cold northerly winds and ice melt remove residual heat from water column over much of the Bering Sea shelf. However, in 1998, northerly winds advected ice over the southeast ern shelf for only a brief period in February In addition, weak winter winds failed to mix these melt-cooled waters to the bottom ( Figure 1). Thus, spring 1998 began with unusually warm bottom water.This residual heat, coupled with summer warming, resulted in the water column over the middle domain having higher heat con tent in summer 1998 than has been observed in recent years [Stabeno et ai, 1999].
The northward movement of this warm wa ter could be traced in the Sea-viewing Wide Field-of-view Sensor (SeaWIFS) images by its association with a bloom of minute phytoplankton cells, the coccolithophorids, that pro-

12
Temperature at M2 duce calcium carbonate plates.These plates re flect light in a characteristic way that allows easy detection by satellite remote sensing. SeaWiFS images from July show the bloom being advected northward east of St. Lawrence Island and through Bering Strait into the Arctic ( Figure 2).

Temperatures 2°C above Normal
Export of heat to the surface waters of the Arc tic Ocean can increase ice melt and sea surfaceatmosphere heat exchange in the Arctic. Recent studies have shown that the ice in the Arctic Basin is thinning [Wadhams, 1995] .Although we have no measurements in 1998 of the tempera ture of the water flowing from the eastern Bering Sea into the Arctic, average temperatures on the shelf were ~2°C above normal.
When the average heat content of eastern Bering Sea shelf waters is elevated and the typi cal northward flow of this water occurs, there will be changes in the surface heat content of the Arctic Ocean. A change of this nature could have important implications for Arctic Ocean cir culation and heat budgets.
The early retreat of the ice in 1998 was fol lowed by an unusually stormy spring that lasted into mid-June. A variety of biological tracers sug gested that strong southwest winds associated with the spring storms resulted in transport of warm surface waters from the shelf break near Unimak Pass to the inner shelf, then northward ' to the Arctic.
For instance, in spring, we observed substantial quantities of bull kelp (Nereocystis luetkeand) floating near the inner front southwest of Cape Newenham and off Nunivak Island (Figure 2). This species has its northernmost and western most distribution in the Gulf of Alaska and east ern Aleutian Islands.

Unusual Observations
During June, oceanic and outer shelf species of zooplankton, particularly copepods (Neocalanus spp.and Metridia pacified) and euphausiids (Thysanoessa inermis) were found along the 50 m isobath at Port Moller, Cape Newenham and Nunivak Island (D. A.Stockwell et al.,submitted manuscript 1999).These obser vations of onshore and northward transport of organisms are unusual in that previous studies have found cross shelf advection and northward flow in the southeastern Bering Sea to be weak [Stabeno etal, 1999].
This advection also affected the distribution of early life stages of walleye pollock (Theragra chalcogrammd),\he most important component of the Bering Sea commercial fisheries. In June 1998, we found larval pollock in abundance off Port Moller, and in August, young-of-the-year pol lock were abundant off Cape Newenham.These observations suggest transport of larvae along the 50 m isobath from spawning areas near Uni mak Island. Surveys in July and September, 1998, in the southeastern Bering Sea and around the Pribilof Islands encountered far fewer pol lock than in previous years (R.Brodeur,unpubl. data), possibly because their distribution had shifted to the northeast.
A model of wind-generated surface currents has shown that the strongest year classes of wall eye pollock occurred when surface winds favor ed advection of larvae inshore, away from canni balistic adults and older juveniles [Wespestad et al., 1999] .The conditions in 1998 were similar to those hypothesized for strong survival of youngof-the-year pollock.The effects of transport of lar val and juvenile pollock on the strength of the 1998 year class will not be known until surveys of the distribution and abundance of these fish are concluded in 1999 and subsequent years.

More Anomalies
Coincident with the warm sea temperatures, other anomalies were observed in both 1997 and 1998, but their connection to the unusual weather is yet to be determined. For example, images from SeaWiFS from February through October 1998 revealed that cells from the previ ous year's extensive coccolithophorid bloom [Vance et al., 1998] had over-wintered ( Figure 2) and bloomed again.The duration of the 1998 bloom, on the order of 9 months, was much greater than the more usual duration of a few weeks to 3 months seen in the North Atlantic Ocean [Holligan et al, 1993].
Coccolithophorids are some of the smallest of phytoplankton species (about 10 microns),and they are not grazed as efficiently as larger phyto plankton cells are by large species of zooplank ton [Nejstgaard et al., 1995] .Thus, coccolitho phorids are likely to sustain a food web that is initially cropped by microzooplankton, thereby creating a longer, less efficient food chain for the transfer of energy from plant cells to upper trophic level organisms such as fish,seabirds, and marine mammals.
Also found in 1997 and 1998 were changes in the relative abundance of zooplankton compo nents of the inner shelf ecosystem.This suggests significant changes since the 1970s and early 1980s (D. A.Stockwell et al.,submitted manu script, 1999).In particular,the euphausiid Thysa noessa raschii,a krill-like organism, was one to two orders of magnitude less abundant in 1997 and 1998 than earlier, and copepods were an order of magnitude more abundant.

Many Shearwaters Died
The decrease in the abundance of euphausiids in the surface waters of the Bering Sea may have affected at least one species of top predator.The short-tailed shearwater (Puffinus tenuirostris~),a species of seabird that migrates from Australia, has in the past foraged almost exclusively on euphausiids in the southeastern Bering Sea (Baduini et al.,accepted). In 1997,high numbers of these birds died in the southeastern Bering Sea, apparently of starvation.
In 1998, though no unusual mortality was seen, birds were underweight. In both years, shearwa ter diets were broader than in previous years, with fish becoming a dominant prey in 1998. Major changes in the components of the zoo plankton community will be likely to affect other higher trophic level species including fish and whales.
Other ecosystem anomalies in 1997 and 1998 included a surprisingly low return of sockeye salmon (Oncorhynchus nerkd) to Bristol Bay the world's largest wild sockeye salmon fishery The catch was approximately half of that forecasted in each year, and both years were well below the average over the last 25 years. In addition, salmon runs to the Yukon River, Kuskokwim River, and other Bering Sea rivers were well below the lev els that were predicted for 1998.The Yukon River chum salmon (O. ketd) catch in 1998 was only about 4% of the mean catch since 1974.The fish that did return were later and smaller than normal.The decrease in abundance and size was noted across several age groups, which suggests that the causative factor was likely to be of marine origin [Kruse, 1998].

FALL MEETING PREVIEW AGU Session Explores Digital Earth and Digital Libraries for Earth System Education PAGES 561-562
Imagine hyperlinking on your personal com puter's Internet connection to delve into an interoperable, global mosaic of geospatial information systems that spouts seamless fountains of data from what otherwise might be a crazy quilt of sources: geoscientific stud ies, Earth and space-based monitoring instru ments, local zoning details, and other geodata relating to the past, present, or future. Imagine, too, the educational opportunities and social implications when this data is manipulated. Perhaps this digital system could help predict climate change, improve community planning, increase agricultural productivity, conduct emergency rescue mis sions, facilitate virtual diplomacy or even help to plan a vacation.
It would be a kind of "digital library of Earth information" or maybe a more encompassing "Digital Earth": frameworks for geo-referenced data as well as intriguing models for helping to better understand the complex nature of Earth systems-including the atmosphere, hydrosphere, biosphere, and solid Earth-and their interactions.
With the convergence of advances in the geosciences and observational and informa tion technologies, a growing group of educa tors, computer experts, and policy makers is pursuing the goal of a national digital library for science education and a Digital Earth. Along the way to creating these models, the visionaries, bureaucrats, and technocrats are also trying to work through some technologi cal and cultural issues such as ensuring inter operability between data systems, access, privacy, and e-commerce.
The upcoming AGU Fall Meeting in San Francisco, California, features a session on "digital libraries for Earth system education." Creating a Digital Earth is also the topic of a series of meetings of the ad hoc U.S. Feder al Digital Earth Interagency Working Group -which includes NASA, the U.S. Geological Survey (USGS),the National Oceanic and Atmospheric Administration (NOAA), the National Science Foundation (NSF), and other agencies.
At a November 10 strategic planning ses sion at USGS headquarters in Reston,Virginia, the group discussed a smorgasbord of proj ects already underway and possibilities relat ed to developing a Digital Earth.Those pre senting updates on their efforts included experts from federal agencies and from pub lic and private organizations.
Dan Dubno, a producer with CBS News, urged the group to integrate the news media in developing a Digital Earth, in part be cause the media is where so many people turn to for their information. Dubno said media concerns about a Digital Earth in clude the reliability of information, freedom of the press issues about access to uncensored geospatial data, and the need for ex perts to help interpret the material."You are storytellers," he told the group."You are going to have to contextualize this." Jay Feuquay of the USGS Earth Resources Observation Systems (EROS) Data Center in South Dakota, which archives images from a number of satellites including Landsat 7, approaches the Digital Earth concept from the perspective of a data provider."I am looking for Digital Earth to solve my pro blem," he said."I have a satellite beaming down a terabyte-and-a-half of information every week. I need to get it out to people. I need to make it available." Joel Halvorson of the Science Museum of Minnesota added that he muses about Digital Earth concepts constantly from his center's new site on the banks of the Mississippi."We will be involved" with creating a Digital