Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 PHYS'CAL ~V'~O~M^T I~T~ODUCT10~ This chapter briefly reviews the meteorological, oceanic, and cryological (ice) characteristics of the environment in three (OCS) lease-sale areas in Alaska: the area in the Navaho Basin portion of the Bering Sea and the two in the Chukchi and Beautort seas (see Figure I-. (Although this report focuses much more on the Chukchi and Beaulort lease areas Han it does on the Navann Basin, He Bering Sea is important because it interacts with the Chukchi Sea and, to a lesser extent, with the Beaufort Sea.) The adequacy and applicability of existing studies are discussed as they apply to environ- mental characterization from the point of view of offshore oil and gas development. Specific recommendations about investigations Hat would contribute useful information about each of regions are given in Chapter 7. The high-latitude locations of the Tree lease-sale areas make Hem fundamentally different from most lower latitude regions. Their environ- ment is quite inhospitable for temperate-zone people for much of the year because of extreme cold, strong winds, darkness, and sea ice. However, the environment is strongly seasonal. The Navarin Basin can be ice- covered in winter but is more temperate in summer, when all the sea ice usually melts. Some portions of the Chukchi and Beaufort seas are affected by sea ice all year, containing both seasonal first-year ice and Licker multiyear Arctic pack ice. The possibility of cross-boundary (or cross- border) impacts on Canadian and Siberian waters always exists. However, there is a stronger possibility of transport into U.S. waters from Siberian 67
68 OCS DECISIONS: ALASKA (Bering Sea and Chukchi Sea) and Canadian (Beaufort Sea) coastal shelf waters because of the mean flow patterns. All three lease areas are remote. Although there are many people and much equipment on the North Slope, offshore locations can be hundreds of miles from help, as are some onshore locations. Even if help is available close by on land or on an island, the harsh environment can prevent transportation to remote drilling or pumping sites or to a ruptured of! line or disabled tanker. Furthermore, the United States has only two open- ocean, polar-class, ice-breaking ships, the Coast Guard's Polar Sea and Polar Star. It is possible, at any given time, for one or both of these ships to be deployed in the Antarctic or in the Atlantic sector of the Arctic and to be unavailable for help in the Alaskan OCS. A possible tempering consid- eration is that the United States and Russia have an agreement for mutual aid in case of an of} spill; the Soviet Union offered assistance after Me Exxon Vaklez spill in Prince William Sound. The background that follows could give the impression that there is an abundance of data about the physical environment of the Bering, Chukchi, and Beaufort lease-sale sites. In reality, the region is enormous, and the data are probably barely adequate to correctly predict which way the air, ice, or ocean win move at any one time or place, say, in the event of an of} spin. The background on ice is found in Chapter 7. For recent reviews of these regions, see Carmack (1990) and Niebauer and Schell (1993~. We note that although the arctic environment is hostile, Alaska Natives have lived there for at least hundreds of years and continue to do so. A great deal of Heir food and materials comes from ice-infested ocean waters. Their knowledge of the physical and biological environment is critical to Weir survival. Under certain circumstances and in specific places, a good hunter might better predict a spill trajectory than a general circulation mode} (GCM). Although most Alaska Natives' knowledge is not written down and is therefore not accessible in a data base, traditional knowledge is an mpor~nt counterpart to Western science and should be taken into account seriously. L~VI~O~M[~TAL STUDIES Physical oceanographic studies (including field studies and modeling in oceanography and meteorology) provide a basis for predicting of! spill transport in OCS regions. The starting point is typically output from an
PHYSICAL ENVIRONMENT 69 ocean circulation model, forced by observations or mode} data on meteorol- ogy that are used to compute possible oil spill trajectories from selected points. In conjunction with the probability of an oil spill for selected launch sites, an oil spill risk analysis (OSRA) mode! is used by the Branch of Environmental Operations and Analysis (BEOA) of MMS to estimate the subsequent probability of the oil's arriving in sensitive areas or on the shoreline within a given time. It is the compilation and distillation of these probabilities that find their way into an Environmental Impact Statement (EIS). OSRA calculates only the movement of the center of mass of each hypothetical spill. Separate estimates are made of the surface area actually covered by of! from a spill of a particular size and of the much larger area over which the of} wood be discontinuously spread. The details of of} transport models are provided in the EISs and references in them. ARCTIC OCI^~I Arctic Ocean water and ice circulation are largely contained and restricted from contact with the temperate oceans by the continents and islands that surround it. The center of the Arctic Ocean is permanently ice- covered; the periphery is seasonally ice-free. It is the seasonally ice- covered periphery that includes the lease-sale sites in the Chukchi and Beautort seas. Should a spill occur, the ice, as well as the currents in the region of these sites, will carry spilled oil. The Chukchi and Beaufort seas are off the north coast of Alaska in the Arctic Ocean proper; the Bering Sea is subarctic, located between Siberia and Alaska soup of the Bering Strait and north of the Aleutian Island chain. The Bering Strait is the only direct connection to the Arctic Ocean from the Pacific Ocean; it is about 50 m deep and 85 hen wide. The net flow Trough the strait is northward and accounts for about 15% of the inflow to the Arctic Ocean (Aagaard and Greisman, 1975~. The Chukchi Sea is underlain by a broad shelf, nearly 900 km from the Bering Strait north to He Pelf break. In comparison, Be Beautort Sea has a relatively narrow shelf. Proceeding east from Point Barrow, considered the boundary between Be Chukchi and Beaufort seas, the continental shelf narrows to about 70 km. The shelf widens again farmer east near Mac- kenz~e Bay to about 160 km and extends eastward into Be Amundsen Gulf. The marine and submarine topography of this Arctic Ocean shelf region includes Wrange} Island, at die approximate western boundary between Be
70 OCS DECISIONS: ALASKA Chukchi and East Siberian seas, and Herald and Hanna shoals in the Chukchi Sea shelf north of Bering Strait. Submarine canyons include He Herald and Hanna canyons in the Chukchi Sea, and the Barrow Canyon at the boundary between the Chukchi and Beautort seas. There are also numerous barrier islands along the north and Chukchi coasts of Alaska. Approximately the northeast half of the Bering Sea (which contains the Navann Basin) overlies the widest continental shelf in the world outside the Arctic; the southwest half overlies an abyssal plain with depths of about 4 km. The eastern Bering Sea shelf is about 500 km wide, but it is shallow (approx~nately 170 m at the shelf break). This shelf borders most of Alas- ka as well as the coast of Russia north of Cape Navarin. Southwest of Cape Navar~n, the shelf narrows by an order of magnitude. The continental slope is indented by several undersea canyons. King, St. Lawrence, St. Matthew, Nun~vak, and the Pribilof islands are in the eastern Bering Sea shelf. The climate and weather of this high-latitude region are strongly related to the presence arm fluctuations of sea ice (Overland, 1981), which are related to the large seasonal variation in insolation. The Arctic is surrounded by the large weather systems of the northern hemisphere. In summer, when He North Pacific high, the Asian continental low, and the North Atlantic high expand to the north, the weather in the Arctic is relatively moderate. In winter, the patterns change drastically so Hat the Asian continent is dominated by high pressure, and He North Pacific and North Atlantic are dominated by He Aleutian and Icelandic low-pressure systems, respectively. Because of the weather pattern change, including the intensification of He Aleutian low, winter is harsh in this part of the Arctic. The mean position Of the Aleutian low is actually a statistical artifact; He low is the average position through which most of He migrating cyclones pass. Three to five storms each month pass along He Aleutians into the Gulf of Alaska or into the Bering Sea-Bristo} Bay region. In comparison, fewer Han two storms a month pass across He northern Bering (Overland, 1981~. Relatively few storms occur norm of Bering Strait. Brower et al. (1977a,b) showed an average of fewer Han one each year for any given monk for He region from Wrange! Island eastward to east of Mackenzie Bay. Although infrequent, the storms that do occur In this region can result in storm surges and severe wave action along He norm coast of Alaska, especially during the late summer and early fall. Intense storm surges occur about once each year (Brower et al., 1977b). To the norm of Alaska, He interaction of different weather systems surrounding the Arctic generates a poorly defined
PHYSICAL ENVIRONMENT 71 relative high-pressure system with clockwise circulation over the Beaufort Gyre, which drives the clockwise flow of the near-surface ocean and ice in the Beautort Gyre. The major rivers that enter the region of interest in- clude the Yukon River in the northern Bering Sea/Norton Sound, the Colville River in the U.S. Beaufort Sea, and the Mackenzie River in Be Canadian Beautort Sea. There are a number of smaller rivers such as Be Kobuk and Noatak that empty into the Chukchi and Beautort seas. As pointed out by Carmack (19903, there has not yet been a synthesis of arctic shelf waters, although he listed many individual studies. Carmack pointed out that annually the shelf waters go through a much larger variation in salinity (-2-4 parts per thousand (ppt)) than do the surface waters (~0.5 ppt) of the Arctic. The shelf waters are less saline in summer due to river inflow and ice melt but are more saline in winter due to river freeze-up and brine rejection from ice formation. In addition, in winter, upwelling may cause higher salinities on the shelves, reaching 34.5 ppt or greater (Aagaard et al., 1981; Melting and Lewis, 1982~. During the summer, Be shelves show net offshore flow at the surface similar to an estuary while in the winter, there is net inflow of more dense water at kept (Macdonald et al. 1989; Carmack et al., 1989; Carmack, 1990~. Thus, Be Arctic Ocean surface layers are freshened by river and glacial input, which are strongly seasonal in nature (Carmack, 1990~. The two major freshwater inputs to Be regions of interest here are the Mackenzie River (340 km3/yr or 0.01 Sverdrup (Sv; ~ Sv = lo6 m3/s), which flows into the Beaufort Sea, and Be Yukon River (214 km3/yr or 0.0068 Sv), which flows into the eastern Bering Sea. The seasonal variability for Be Mackenzie River is about 5-fold and it is about 12-fold for the Yukon River. In comparison, about 1,500-2,000 km3/year (~0.05 Sv) of fresh water enters Be Arctic Ocean as a component of Be inflow Trough Be Bering Strait, which includes some of Be Yukon River. SOUTH~STI ~ BATING S" The physical environment of the Bering Sea shelf is characterized by strong variability in Be air, ice, and ocean climate regimes. This includes periods of days to years, including a pronounced interannual variability (Niebauer, 1988~. Daylight is nearly nonexistent in winter and nearly continuous in summer. The wind stress over He sea also varies by an order
72 OCS DECISIONS: ALASKA of magnitude from summer to winter. The shelf region is ice-covered in winter, although the entire Bering Sea is ice- free in summer. Tidal currents dominate the entire OCS; they contribute 60% of the horizontal kinetic energy in the outer shelf (from water depths of approxi- mately 100 m to the shelf break) to 90% along the coast (Kinder and Schumacher, 1981a). Because the southeastern Bering Sea shelf is so broad, the various sources of energy, such as tides, winds, and freshwater input, are applied over a large area. Over the southeastern shelf, the mean current flow is low, 0.01~.05 m/s, and moves toward the northwest. Seaward of Be shelf break, the Bering Slope current flows at speeds of approximately 0. ~ m/s with numerous eddies (Kinder and Schumacher, 198Ib). Because of Be slow mean flow, the hydrographic structure on the shelf tends to be locally formed by the input from insolation, cooling, melting ice, freezing, and river runoff, as well as lateral exchange with bordering oceanic water masses (Kinder and Schumacher, 198Ib; Coachman, 1986~. The Bering Sea has four distinct hydrographic domains, each with associated current patterns, defined by water depths and boundary fronts that are generally parallel to the isobaths (Kinder and Schumacher, 198la,b; Coachman, 1986~. They are the coastal, middle shelf, outer shelf, and · e oceamc c Domains. MOTH BASING Sly AND I I I Id STRAIT The Bering Strait is narrow (85 kin) and shallow (50 m). Flow averages about 0.25 m/s and about ~ Sv in summer, and I-~.5 Sv in winter (Coach- man and Aagaard, 1988). The northward flow results because the sea surface tilts downward toward He norm. There are strong currents and current shears across the strait. The strongest currents in the upper layers of the east side can be more Han 2 m/s. The flow in the western side can be 0.5-0.6 m/s with little vertical shear. However, reversals in flow of at least a week's duration, especially in winter, are related to atmospheric pressure gradients over this region (Bloom, 1964; Coachman and Aagaard, 1981~. Winds tend to be channeled norm or soup Trough He strait. In the northern Bering Sea (approximately norm of 62° N including Norton Sound), there are three identifiable water masses and related fronts. The water masses are related but not identical to He water masses in the soudleastern Bering Sea (Coachman, 1986; Hansell et al., 1989~. Fronts between these water masses are identified (Paquette and Bourke,
PHYSICAL ENVIRONMENT 73 1974; Wiseman and Rouse, 1980; Muench, 1990) as associated with the boundaries between the water that flows from the Bering Sea through the Bering Strait into the Chukchi and Beautort seas and the Arctic Ocean, driven by the sea-level difference between the Pacific and Arctic oceans. These waters make their way eastward along the Alaskan arctic coast as outlined below in the individual regions. CHU~CHI So The Chukchi Sea widens north of the Bering Strait with Kotzebue Sound immediately to the northeast of the strait. The shelf is relatively shallow (20~0 m). Herald Shoal, about 200 km due north of the strait, is 20-30 m deep. Many of the capes and headlands in the region on the Alaskan side of the Bering Sea are high mountains, which tend to cause "corner-effect" accelerations in winds along the coast, analogous to the high winds caused in cities by tall buildings (Kozo, 1984~. The three water masses that flow northward through Bering Strait cross the Chukchi. Paquette and Bourke (1981) showed Mat flow trajectories crossing Be Chukchi are steered Trough Be troughs and canyons. These flows result in melting Mat causes embayments in Be sea-ice cover (see Chapter 7~. Norm of He strait, He Anadyr and Bering water tend to combine and flow northward and slightly eastward at 0.15~.20 m/s, bifurcating in the region between the Point Hope-Cape Lisburne headlands and Herald Shoal. Some of this water goes Trough the Herald Canyon into He Arctic Ocean to He west of Herald Shoal, and some turns eastward as Bering Sea water that flows eastward along the outer shelf of the Beaufort Sea (Coachman et al., 1975; Aagaard, 1984~. In the eastern Chukchi Sea, Alaska coastal water also flows along He Alaska coast, gaining fresher, cooler water from He large rivers that empty into Kotzebue Sound. The Alaska coastal water flows at speeds of 0.25-0.30 m/s past Point Hope, Cape Lisburne, and Icy Cape toward the Beaufort Sea beyond. Water also "drains" from the Chukchi Sea through He Barrow Canyon off Point Barrow into He Arctic Ocean (Aagaard, 1984~. The maximum amplitude of tides in He Chukchi and Beautort seas is only 5-20 cm, which is less Han ~ % of He total variancein local sea level; ddal speeds are about 5 cm/s, which is 1-2% of He total variance (Kowalik, 1984; Aagaard et al., 1989~. In He coastal lagoons, tidal variation can be up to 2 m (Kowalik and Matthews, 1982~.
74 OCS DECISIONS: ALASKA B"UFO~T S" There are several subscale (kilometers to tens of kilometers) wind phenomena that are important to the shelf circulation of the Beautort Sea (Kozo, 19844. Monsoonlike winds occur during the summer, caused by a semipermanent arctic atmospheric front that results from horizontal thermal contrasts along the coast. Heating of the land causes a deficit in atmo- spheric pressure, leading to onshore winds, which are turned to the right (toward the west) along the coast as a result of Coriolis acceleration. Related to Be monsoon is a breeze (Kozo, 1984~. Along the Beautort Sea coast, sea breezes are characterize by large diurnal sea-to-land temperature contrasts, clockwise rotation of surface winds, and surface upends that oppose offshore gradient winds. At least 25 % of the time the surface-w~nd direction is dominated by the sea breeze. One effect of these winds is the maintenance of coastal currents (0-20 km wider toward the west, causing lagoon flushing. Sea breezes cause a general masking (about 25% of the time, as mentioned above) of synoptic wind conditions in this first 20 km from the coast. During summer, sea breezes along this coast are not folBowed by land breezes because the land stays warmer than the water (the summer sun does not set at this latitude). OroUra The Brooks Range of mountains, which has a mean height of I.5 kin, is some 240 km inland of and parallel to Be general trend of the arctic coast of Alaska. This major mountain range affects winds over much of Be Beaufort Sea coast (Kozo, 1984~. For example, near Barter Island, orographic modification of the winds can cause wind speeds 50% greater than that of the geostrophic wind (the wind resulting from a balance between horizontal pressure gradients and the Earth's rotation) because of the corner effect of the mountains, which are close to the sea. This effect can be felt as much as 350 hen away, and it can influence Be circulation of ice and water in the Beaufort Sea. In addition, mountain barrier baroclinicity (a condition in which surfaces of equal pressure are inclined to surfaces of equal density) occurs when stable air moves toward and up a mountain without heating from below. This causes Be isobaric (equal pressure) and isothermal (equal temperature)
PHYSICAL ENVIRONMENT 75 surfaces to tilt away from Be mountain range, resulting in geostrophic winds parallel to the axis of the mountain range. This is mainly a winter phenomenon and is a major reason for wintertime winds from Be west- sou~west between Prubhoe Bay and Barter Island. This phenomenon re- sults in a nearly Degree difference in wind direction when compared with the winds predominantly from He northeast at Barrow. Because mountain barrier baroclinicity depends on high surface albedo (the fraction of incident electromagnetic radiation reflected by a surface), it disappears in summer, when there is no snow. This effect has a horizontal extent of approximately 120 km and occurs about 25% of Be time in Be coastal zone from Crusoe Bay to east of Barter Island. Finally, wintertime atmospheric temperature inversions are common in the Arctic. They cause strong atmospheric surface stability that leads to diminished vertical turbulent exchange and hence to reduced wind stress on Be surface to drive ice and ocean circulation. Current Along the Beaufort Sea coast, in Be event of a spill, Be ice pack will trap and carry Be oil in addition to He currents. There are also river plumes from He Colville River (and He Mackenzie River to He east in Canada), as well as the smaller rivers along the Beautort coast, that affect the coasts and shelf flow of the He Beaulort Sea. The fresh water flowing out over He salt water causes buoyancy-driven circulation. In He winter, river flow often does not completely stop so there is freshwater flow out onto the open shelf under the land-fast ice. Variable weaker patterns can cause these buoyancy flows to interact with winds, causing transient current jets along He coast, especially in summer when ice may not shield the ocean surface from wind stress. Shoreward of the 50-m isobath, local winds do- minate the shelf flow of He ocean currents and drive Hem mostly toward the west under He generally prevailing easterly winds. However, Here are periods, occasionally prolonged in some summers, when west winds cause easterly flow. Thus, the circulation at the coast is strongly wind~riven, but variable and highly seasonal; it is less energetic in winter because of the ice cover. The result is that significant coastal flow in this shallow inshore re- gion appears to be primarily a summer phenomenon (Aagaard, 1984~. Kowalik and Matthews (1983) showed evidence that salt rejection from
76 OCSDECISlONS: ALASKA growing sea ice in a shallow coastal lagoon induces a two-layer water system. The higher-salinity, higher~ensity, and, hence deeper water moves offshore at a rate of approximately I-2 km/day (~-2 cm/s); the less-saline, less~ense upper layer moves onshore. Drifter studies cited by Kowalik and Matthews (1983) showed shoreward movement for drifters released under the sea ice as far as 10 On seaward of the barrier islands. Tidal and surge currents account for most of the variation in the currents, but these are superimposed on the mean brine-induced currents. Offshore, seaward of 50 m over the OCS and continental slope, there is an organized band of flow toward the east, called the Beaufort Undercurrent (Aagaard, 1984), which is topographically steered but apparently not driven by local wind stress. It is estimated that less Wan 25% of flow variability below 60 m is caused by wind (Aagaard et al., 1989~. The dynamics of the undercurrent are not yet filthy understood, but it is characterized by a temperature maximum associated with eastward flow originating in We Bering Sea. The Bering Sea water can be traced at least as far east as Bar- ter Island. This flow seems to be trapped along the OCS and slope between the Am (~40 On offshore) and Be 2,500-m isobaths (~120 km offshore). Aagaard (1984) suggested Cat Be Beautort Undercurrent extends the full length of the Beaufort shelf and slope. The mean currents are approx'- mately 0.~ m/s toward Be east. Aagaard (1984) also reported frequent cross-shelf flow (mostly offshore flow of up to 5 cm/s for as long as 3 days), which links the nearshore region to the undercurrent. The ocean driving of this circulation includes shelf waves and eddies (Aagaard et al., 19893. Upwelling appears to be connected wig eastward-traveling wavelike disturbances in Be velocity records wig vertical displacements of about 150m. Condoning farther offshore of He Beaufort Undercurrent into Be Arctic Ocean, Be surface current and ice flow are characterized by a mean west- ward movement of ice and water at Be outer edge of Be anticyclonic Arctic Ocean gyre. S" ICE The presence of ice in tile arctic OCS is arguably Be most significant physical condition to be dealt wig in developing OCS oil and gas resources (see Chapter 7~.
PHYSICAL ENVIRONMENT 77 Sea ice in the area shows great seasonal and interannual variability as well as some predictable mesoscale features (Burns et al., 1981~. For example, the seasonal sea ice advance and retreat in the Bering Sea is the largest of any found in the Arctic or in subarctic regions, averaging about 1,700 km (Walsh arm Johnson, 1979~. Interannual variability in the position of the average ice edge in the Bering Sea is as great as 400 An (Niebauer, 1983~. Farther north, the oceanic flow that fans out in the Chukchi Sea after going through Bering Strait causes significant mesoscale embayments in the Chukchi ice cover (Paquette and Bourke, 1981), which are predict- able from summer to summer. Finally, winter conditions in the Chukchi and Beautort seas produce fast ice along Be coast that interacts with the off- shore open Arctic Ocean current and wind-driven free-floating sea ice to cause an extensive, somewhat predictable, system of flaw leads and polyn- yas off the Chukchi and Beautort coasts eastward to Be Canadian Archipel- ago. Maximum sea-ice cover occurs in March or early April, lagging minimum insolation in late December by 3 months because of the heat capacity of the ocean and the cold atmosphere. At this time, essentially all of the Arctic is ice-covered, as is about one-third to one-half of the Bering Sea. The mean ice edge in the Bering Sea is about 900 km south of Bering Strait, and it varies from 700 km to about 1,100 hen south of Bering Strait, a range of 400 fan, or more than 40% of the seasonal ice cycle (Niebauer, 1983). Maximum retreat of Be sea ice occurs in September, again lagging maximum insoladon by about 3 monks. In Be mean, Be ice edge retreats about 1,600 km between March and September, moving into Be Chukchi Sea (~e Bering Sea becomes ice-free). By September, in normal years, Be ice pulls away from Be Arctic coasts of Canada, Alaska, and Siberia, except for an approximately 300-lan-long section of Be Siberian coast; leaving a nearly continuous, relatively ice-free corridor around Be perma- nent ice pack. In most years, this corridor varies in wide from about 300 km at the western end of Be East Siberian Sea to less Man 50 hen off Pru~hoe Bay. The deviations around these locations are large. In warm years, Be ice is 600 km off Be coast of Be East Siberian Sea and 300 km off Be Beautort Sea coast at Prubhoe Bay. In cold years, Be ice sometimes does not pull away from Be coast, although Here is invariably open water In some of He bays along He norm coast of Russia in He Chukchi and East Siberian seas. However, individual storms can cause large changes in ice
78 OCS DECISIONS: ALASKA cover in a short time. For example, off Barrow, northwesterly and nor- ~erly winds can cause compaction of drifting ice, closure of open coastal water, and extensive runup and pileup of ice onto Me coast Leeks and Weller, 1984~. Such events trapped and destroyed many whaling ships in He northeastern Chukchi Sea during Be late nineteenth century. In September, when there is a minimum of ice, Be dis~ibudon of ice and open water reflect space ocean flow patterns into Be Chukchi Sea. Open water generally reaches Trough Be strait into Be Chukchi and Beaufort seas, paralleling the coast of Alaska norm to Point Barrow and Ben extending eastward all Be way to Be Canadian Archipelago. This is a region of extensive nearshore leads and recurrent polynyas, which form as a result of Be ~nteracdon of the fast ice and land with drifting ice driven by ocean currents and winds. To Be west, Be ice sometimes does not pull away from Be Chukchi coast south of Wrangel Island. During years of unusually heavy ice cover the ice margin in September can be as far south as about 66° N on the Siberian side but only 70° N on the Alaskan side. This is the result of warm currents that flow through Bering Strait, hugging the Alaskan coast to Point Barrow and then curving off to the east (Paquette and Bourke, 1981~. Paquette and Bourke (1981) depicted current flow crossing the Chukchi as being steered by the bottom troughs and canyons. This flow of warmer water from Bering Strait results in the melting of sea ice and causes recurring embayments in the ice (Paquette and Bourke, 1981~. In spring, along the northwestern coast of Alaska from Point Hope to Point Barrow, there is a region of leads and polynyas offshore of He land-fast ice (Stringer et al., 1982~. Most of the ice in these regions is open pack ice driven by wind and ocean currents. The pack ice is primarily first-year ice, except in the Beau- fort Sea, which contains some multiyear ice from the Arctic. Limited fast ice is present in the eastern Bering Sea, found mainly in protected bays or along shores facing the prevailing winter winds. In the Beautort Sea, fast ice is more extensive because of protective barrier islands and because the grounded pileups of sea ice on the shelf act like small barrier islands. However, the fast ice in the Beaufort Sea is seasonal, usually lasting from November through June. Ice drift rates are highly variable in the Bering Sea, frequently with rates of 17-22 km/day. Rates as high as 32 km/day have been reported (Shapiro and Burns, 1975; Muench and AhInas, 1976; Weeks and Weller, 1984~. In Bering Strait, ice movements of 50 km/day have been observed, although there are reversals. In tile Chukchi Sea, drift rates are considerably lower,
PHYSICAL ENVIRONMENT 79 0.~4.8 km/day for the mean annual drift, but rates as high as 7.4 km/day have been observed (Weeks and Weller, 1984~. In the Beautort Sea, the ice spews are 2-S ~n/day. The offshore ice generally follows the east-to-west anticyclonic ocean circulation, essentially parallel to the coast. Undeformed ice thickness increases toward the north from about 0.5-~.0 m in the open Bering Sea and ~ m in Bristol Bay to 2 m off the arctic coast (Weeks and Weller 1984~. The Bering Sea ice cover is almost entirely first-year ice; north of Bering Strait, the ice is 25-75% second-year ice or older. The older ice is thicker, with a mean of roughly 4 m. However, pressure ridges of heavily deformed ice have been observed with maximum depths (keels) of 50 m and heights (sails) of 13 m. When they become grounded, these deep keels can cause appreciable gouging of the shelf. In near-coastal zones where grounded ridges can form, sails can be as much as 20 m high. The pack ice over the arctic shelf is commonly highly de- formed, with up to 10 ridgesJlan, because of shearing. Farther offshore, beyond the shelf, 2-3 ridges/hn is more typical. There are fewer large ridges in Be Bering Sea ice because of the less constrained nature of the ice drift (there are fewer immovable barriers Me ice can work against to form ridges as it drifts toward Be open sea). PHYSICAL OC~OGQAPHIC STUDII S Given He state of knowledge as described previously, we have come to He following evaluations of He available data. Models of Circulation and Oil pill Tralectories OSRA is He mode} used by BEOA to generate of} spill trajectories from selected hypothetical spill points, and estimate He number of "hits" on an environmental resource target or a shoreline segment, and the conditional probability of some effect on He resource within a selected time. It is He compilation and distillation of these probabilities Hat finally find Heir way into an EIS. The input requirements of He model include data on air and ocean circulation in an area. These inputs typically are taken from ocean circulation and meteorology models. Trajectories for all OCS waters except
SO OCS DECISIONS: ALASKA those in the Alaska region are calculated by BEOA. In He Alaska region, contractors calculate trajectories and provide them to BEOA for mode! input. An important consideration in Alaskan modeling is the presence of ice. The trajectories, and hence OSRA's results, have been of varying quality over the past 17 years (NRC, 1990a [Physical Oceanography]~. Circulation Models The National Research Council has reviewed Me information available for gas and of} leasing in other OCS areas (NRC, 1989a [California and FIoridal; NRC, 1991a [Georges Bank]) and has concluded that ocean spill trajectory estimates have relied too heavily on GCMs. The committee be- lieves the same conclusion holds for the Tree lease areas involved here. With the relative lack of observations, initial reliance on model predictions is understarK able, but as stated in the previous NRC reviews (NRC, 1990a, 1993a) trajectory predictions must be tied more closely to observations Man has been Be practice. The committee also notes dlat GCMs for Be Alaska region have been used by several different contractors using different models, with little effort to synthesize the various results or to reconcile Be predictions with the limited existing observational data. The committee believes that little progress can be expected from refining existing GCMs for such vast areas. It would be more useful to develop limited-area GCMs for use at sites selected for exploration and develop- ment. These could be used to explore Be possible effects to biologically or ecologically important areas in Be vicinity of Be production zone, to explain how mitigation measures might work, and to predict of! movement in Be event of an accident. It is important Mat such a modeling effort be coordinated wig observational efforts in die same area and take advantage of traditional knowledge. The duration of the existing base of observations, in addition to being geographically sparse, is insufficient to distinguish mean circulation from fluctuations that result from annual or even seasonal processes. MMS has Ken an important step in addressing these issues in other lease-sale areas by deploying meteorological buoys. MMS is to be commended as well for equipping some of these wig acoustic Doppler current profilers, which record Be current profile beneath the buoys as a function of time. Because harsh weather and ice preclude the deployment of similar tools in most of
PHYSICAL ENVIRONMENT 81 the lease-sale areas, the description of long-term variability is an important issue. To maximize the ability to contain spills and protect sensitive resources from them, a mode} must be used quickly enough to provide actual predictions of trajectories. This is different from the requirement in the EIS process for statistical estimates of of} spill trajectories in that it must be run on a time scale commensurate with an actual spill. It must include envi- ronmental data and forecasts for currents, wind, and ice conditions as observed during the spill, although it might contain a statistical component that would allow for uncertainties and predict the most likely or possible paths. Several such models are available and have been useful (Giammona et al., 1992~. Once possible development sites have been selected, one or more of these models should be selected for use in case of an accident, and the appropriate site-specific information required by the model (such as local mean currents, tidal currents, mesoscale eddy energy, and ice conditions) should be measured and made ready for mode} input. Atmospheric Forcing of the Ocean Clrcula~on The atmosphere is important in determining the local effect of wind stress and heat transport and also, remotely, as conditions in distant areas are propagated through the coastal ocean by pressure gradients. In OCS areas off the continental United States and in the Bering Sea, MMS has estab- lished a network of meteorological buoys to monitor the lower atmosphere over long periods (10 years). I,here is no comparable set of observations for the Chukchi and Beautort lease-sale areas, although the committee believes that sufficient information is available from standard numerical weather forecasting products and from fixed-station observations to describe atmospheric variability on scales larger than the mesoscale (100 kind. The absence of information related to mesoscale variability in Me lease-sale areas is not seen as a flaw in the existing EIS, but that information will be required once specific production areas are identified. Studies should be designed to determine the spatial structure of Me wind field on scales of a few ldlometers in areas of production, with emphasis on the mechanics of the marine boundary layer and the interaction between the layer and coastal topography. Such observational studies should involve aircraft surveys and fixed-station observations. To take advantage of the
82 OCS DECISIONS: ALASKA long time series of weaker on larger scales, Here should also be studies that define Be relationship between mesoscale variability and fluctuations in larger scale weather. AvallaDilitv and Sultanillt Of the Onse~ational Base Considerable efforts have been made by MMS and other federal agencies, by Be state of Alaska, and by local organizations such as the North Slope Borough to acquire observations on ocean circulation in the areas considered here. Although these efforts have in many cases been quite successful, the number of observations available to describe the circulation in this vast area is much smaller than it is for other OCS areas bordering Be continental United States. Important distinctions also exist between the three lease areas: The Bering Sea, in which Be Navarin Basin is located, is the area for which the richest base of observations is available; less information is available for the Chukchi Sea, and Be greatest uncer- tainties concern Be U.S. portion of the Beautort Sea, although Canadian research on Be eastern portion of Be Beautort is helpful. Ironically, it is the region with Be least data Mat is currently of most interest to Be of} industry. Given the size of Be regions involved and Be relative paucity of physical observations, Be information available is marginally adequate for Be preparation of EISs. The information base is not adequate to build any de- tailed predictive circulation model, as would be required to effectively manage an of} spill in Be Chukchi and Beaufort lease-sale areas. Studies of circulation in specific sites identified for production are recommended. This is consistent wad the strategy followed by MMS in over OCS areas such as studies of ocean circulation on Be Texas-Lowsiana shelf and in the Santa Barbara Channel. Data that have been certified by investigators are available from several federal archives (e.g., Be National Oceanographic Data Center (NODC) and the National Snow and Ice Center). In the case of NODC, data gathered on oceanographic cruises funded by the National Science Foun- dation are required to be report to and eventually archived at the NODC. Data are also available from researchers in their published results as well as data reports. For example, MMS makes dlese research and data reports available. An enormous amount of data is available Trough tile EISs.
PHYSICAL ENVIRONMENT 83 PhYslcal ProoerIles of All and Water The high viscosity of oil at low temperatures, combined wig He influ- ence of surface waves and mesoscale eddies, will lead to spilled of} forming a highly discontinuous pattern of patches, windrows, and sheens rawer than a more-or-less continuous slick. Although models for this have been devel- oped and Weir resets are reported in EISs, this topic needs to be kept under review as new information becomes available about the physical properties of crude of! Cat is discovered in Be region and as understanding of Me physical oceanic conditions improves. T~SMISS10~ OF POISE 1~ THE M~121~ ~VI~O~MI AT The effect of noise on marine mammals especially bowhead whales is a critical issue on the Norm Slope because Alaska Nadves depend on whales for subsistence. Although noise transmission and attenuation in Be sea are physical phenomena, this issue is discussed in detail in Chapter 5, where the committee argues that additional physical information alone is unlikely to be instrumental in resolving Be issue. CONCLUSIONS AND 121 COMMENDATIONS The specific conclusions and recommendations for tills chapter follow. For Be general and overall conclusions, see Chapter 8. Conclusion I: The oceanographic model Cat has been used is elaborate, but inevitably inaccurate because of major uncertainties about the physical processes and the mathematical conditions that are applied to mode} boundaries in the water. Nevertheless, the model's output is sdI! a useful rough guide to the path and fate of spilled oil, although a simpler mode} might have been just as useful and credible. Recommendation I: Improve mode! predictions based on existing knowledge. Blend in observations of factors, such as the mean circu- lation in an area, rather than trusting the model to generate this information accurately (which it generally will not). There are
84 OCS DECISIONS: ALASKA numerous ways in which improvement of the model could come from further research on sea ice physics, on interactions of the currents with topographic features, and on the representation or analysis of small-scale turbulence and mesoscale (tens of kilometers and smaller) eddies. Ideally, there would be continuing evolution of the mode! through comparison of its predictions with new data. More attention should be paid to the tremendous interannual variability in oceanic and ice conditions, and there should be more analysis of extreme events and worst-case scenarios. Further improvements in predicting the fate of spilled oil could come from a better understanding of its spreading and weathering. Any study of environmental impact is likely to require a mode} that can predict the path of an of! spill, but after the initial use of such a model it is important to identify important problems about which uncertainty is unacceptably high and for which reduction in uncertainty in the physical oceanographic parts of the problem would be worth- while. The current approach is "bottom-up," where physical ocean- ographers strive to produce the best model possible. A "top-down" approach driven by specific environmental concerns and spill-response operations might better focus the research. There is little evidence that this focusing has been attempted, and until it has occurred it is difficult to say with confidence whether the existing environmental information and modeling are adequate. Alterrlative: None recommended. Conclusion 2: Trajectory estimates have relied too heavily on general circulation models (GCMs); they have not been closely tied to observations in the past. The committee also notes that the GCMs for the Alaska region have been prepared by different contractors using different models, with little effort to synthesize the various results or to reconcile the predictions with the limited existing observational data. Recommendation 2a: Conduct site-speci~c circulation st~ies In areas identified for production and in "hot spots" - reedling, feeding, arm aggregation areasidentifiedl by biologists and Alaska Natives with long experience in debug with the physical environment. Large-scale studies of He circulation are unlikely to be effective. (The committee estimates that these studies will require at least 2-5 years; Key should
PHYSICAL ENVIRONMENT 85 be subject to continued review. The time required will depend on what the key problems turn out to be and the degree to which inter- annual variability is a consideration.) In particular, the committee recommends five studies: (a) Further work on the exchange of water between shallow lagoons and He open sea is required to develop the basic data and un- ders~ing Mat would be necessary to protect them in the event of an approaching oil spill and to facilitate their restoration should Key be damaged. (b) Boundaries between different water masses are frequently zones of significant biological activity and accumulation, and they also can be places where spilled of} converges. A survey of fronts in the Beaufort and Chukchi seas and studies of their causes and variability would be useful. (c) Interpretation of the ice gouges on the seafloor is hindered by a lack of information about the rate at which they are filled in by the transport of sediment (see Chapter 7~. Determination of near-bottom currents (on all time scales) would provide, in combination with mea- surements of sediment properties, some guidance in interpreting the bottom topography as observed at a specific time. (~) Long time series of ocean currents and related physical characteristics are required to quantify the interannual and seasonal variability in the lease areas. (e) Field work and modeling would clarify the extent to which causeways from the shore that have been built or that are proposed in support of industrial operations need to incorporate breaches to permit Be continued alongshore flow of water and to permit fish passage. Such studies need not delay leasing decisions. Alternative: None recommended. Recommendation 2b: Refine existing models for use at sites selected for exploration Carl development ares at biological "hot spots." Coordinate modeling with observations in the same area. Little progress can be expected from refining existing large-scale models. MMS should develop limited-area models for use at sites selected for exploration and development dial could be used to explore possible effects on biologically, ecologically, and socially important zones in Be vicinity of He production region, to explain how mitigation mea
86 OCS DECISIONS: ALASKA sures might work, and to predict oil movement should an accident occur. Trajectory models should be used to provide focus and esti- mates of probability, but they should not be used as the sole deter- m~nant of where biological studies wall be conducted. If Me models are not refined for specific sites, the results will be of dubious value. See detailed recommendations a, b, 1, and e in Recommendation 2. Altemative: None recommended. Conclusion 3: It is well understood that the atmosphere influences ocean circulation in OCS areas, in part because of the local effect of wind stress and heat transport, but also farther away as conditions in distant areas are propagated through the coastal ocean by pressure gradients. In OCS areas off Me continental United States and in the Bering Sea, MMS has estab- lished a network of meteorological buoys to monitor the lower atmosphere over periods (10 years). There is no comparable set of observations for the Chukchi and Beautort lease-sale areas, although the committee believes that sufficient information is available from standard numerical weaker forecast- ing products and from fixed-station observations to describe atmospheric variability on scales larger than the mesoscale. The absence of information related to mesoscale variability in He lease-sale areas is not seen as a deficiency at Me leasing stage in tile existing EIS, but we note that tills information will be required once specific production areas are identified. Recommendation 3: In production areas and biological "hot-spots, " design studies of the spatial structure of the wir~fie~ on scales of a few kilometers that emphasize marine boundary layer mechanics and the interaction between the boundary layer And coastal topography interaction. Such observational sh~ies should involve aircraft surveys and fixed-station observations. To take advantage of Be long time series of weather on larger scales, studies also will have to be sponsored Mat define Be relation between Be mesoscale variability and fluctuations in Be larger-scale weather. The committee under- stands that Be Beaufort and Arctic Storms Experiment (BASE) pro- gram of the Canadian Atmospheric Environment Service, which is relevant to this recommendation, is planned for 1994. See ~ in Recommendation 2. Alternative: None recommended.