Strait region to date. Studies since tiien. notably the Inner siielf 

 Transfer and Recycling (ISHTAR) program iiave expanded 

 our understanding of the regional oceanography yet further 

 (Coachman, 1986: Walsh ^/ a/., 1989). 



Complete, integrated studies of the region have been 

 restricted by its strategic significance, national boundaries, and 

 Exclusive Economic Zones. The Third Joint US-USSR Bering 

 & Chukchi Seas Expedition on the Soviet research vessel 

 Akademik Korolev (Korolev) in the summer of 1988 (AK-47) 

 afforded an opportunity for US and Soviet scientists to study 

 the oceanography of the northern Bering/Chukchi Seas without 

 limitations imposed by territorial boundaries. The cruise took 

 place from 26 July to 2 September 1 988 and occupied 1 02 CTD 

 stations in the Gulf of Anadyr, Chirikov basin, and southern 

 Chukchi Sea (see Frontispiece). (An additional 1 1 stations 

 were occupied near the Aleutian Islands and in deep parts of the 

 Bering Sea, but are not discussed here.) 



There are three primary water masses in the northern 

 Bering Sea, and the basis for their identification is salinity 

 (Coachman et ai, 1975). The most saline is water from the 

 continental slope of the eastern Bering Sea Shelf edge, which 

 enters the region via the Gulf of Anadyr and Anadyr Strait to 

 the west of St. Lawrence Island. This is the most important 

 water source to the extremely productive northern 

 Bering/Chukchi Sea ecosystem because of its high nutrient 

 loading. The least saline water lies in the east, the Alaskan 

 Coastal water, which flows parallel to the Alaskan coast 

 northward through Shpanberg Strait to the east of St. Lawrence 

 Island. The water mass of intermediate salinity, which is also 

 the coldest, is Bering Shelf water in residence over the extensive 

 shelf area south of St. Lawrence Island. It is advected northward 

 around both ends of St. Lawrence, through both Anadyr and 

 Shpanberg Straits, and northward between the other two; part 

 of the water mass modification occuring in the system is com- 

 mingling of these three water masses as they are advected 

 northward. 



Salinity valuesof the water masses are not only space, but 

 time-variable, as much as 0.5 ppt seasonally and interannually. 

 Thus, in the absence of quasi-synoptic data from the whole 

 system, the precise changes of water mass properties as they 

 transit the various basins and straits have never been observed. 

 As the Korolev data provide the first-ever quasi-synoptic 

 picture of the regional water masses, this paper describes their 

 modification as they are advected north from the shelf break of 

 the Bering Sea through the Gulf of Anadyr, Chirikov basin, and 

 the southern Chukchi Sea. 



Methods 



Conductivity-temperature-depth (CTD) casts were made 

 surface-to-bottom using a Sea-Bird Electronics model SBE-9 

 system with a General Oceanics RMS 1 2 rosette water sampler. 

 The rosette held 1 2, 2.5-liter "GO-FLO" water sampling bottles. 

 These provided water samples for many other projects as well 

 as samples for salinity analysis to compare with the CTD 

 values. The salinity measurements were made usinga Beckman 

 RS7-C laboratory salinometer. 



The Sea-Bird was delivered new, just two weeks before 

 AK-47 began. It has a rated accuracy of 0.004°C/year over the 

 range -5 to H-35°C, 0.0003 S/m/month over the range to 

 7 S/m, and 0.02% of full scale over the depth range 0-3,500 m. 

 The instrument was calibrated by the US Northwest Calibration 

 Center in Seattle before and immediately following the Korolev 

 cruise, with nearly no changes in output. Later, the same CTD 

 was used in the Antarctic, where salinities from some 

 1 ,000 points were compared with samples run on AGE and 

 Guildline salinometers — differences were less than 0.01 ppt. 

 Subsequent calibrations have shown this instrument to be very 

 stable and its accuracy is well within the tolerances acceptable 

 for modem physical oceanographic research. 



Methods of CTD deployment and data reduction are 

 pertinent to data quality, so they are outlined briefly. The CTD 

 operator prepared the rosette and set up the computer about 

 15 min before each station. On station, the instrument was 

 lowered to the sea surface (or up to 5 m below surface, 

 depending on sea state) and held while the program to record 

 data was started. It was then lowered at a rate between 1 5 and 

 30 m/min until it was about 5 m above the sea floor. When the 

 instrument's attitude in the water column was seen to be stable, 

 it was then lowered another 2 or 3 m. The computer was then 

 reset for the uptrace, and the rosette bottles were tripped at 

 predetermined depths on the upcast. 



Data is acquired by the Sea-Bird at a rate of 24 scans of 

 pressure, temperature, and conductivity per second. For 

 AK-47, scans were averaged in groups of six, giving four data 

 groups per second to be recorded. At a drop rate of 30 m/min, 

 CTD values were thus acquired approximately every 0. 1 25 m. 

 The data are averaged internally, digitized, and transmitted to 

 the ship via the center conductor in the sea cable through winch 

 slip rings into the deck laboratory. The deck unit (Sea-Bird 

 model 1 1) converts the data to computer-compatible signals, 

 which are fed into a Packard-Bell AT-type computer via an 

 IEEE 488 (GPIB) bus. 



Using Sea-Bird supplied software, the CTD data were 

 displayed on the CRT monitor in real time as X-Y plots as the 

 instrument was being lowered. As the rosette bottles were 

 tripped on the upcast, the usual problems in calibrating the 

 CTD conductivity sensor were encountered. Because of water 

 disturbance on the upcast by the rosette and CTD housings, 

 salinity readings by the CTD are suspect. Thus, comparison of 

 salinity samples with the CTD output does not necessarily give 

 valid in situ calibration data. Also, comparison with downcast 

 values in shallow, highly variable shelf waters is likewise 

 suspect. Nevertheless, at least two samples from each station 

 were collected for checking the CTD calibration. An ancient 

 Beckman salinometer was used to run these salinities, which 

 presented problems with drift. In spite of all the difficulties, 

 the results show /. consistency in Sea-Bird CTD 

 output station-to-station; 2. close agreement with SEACAT 

 data when the two instruments were run together; and 3. close 

 agreement between CTD values and laboratory 

 determinations, providing confidence in the accuracy of the 

 data from AK-47. 



28 



