Now showing items 1-9 of 9

• Decadal Variability In The Arctic Ocean: Greenland-Iceland-Norwegian Seas Ice-Ocean-Atmosphere Climate System

This study investigates the decadal variability of the Arctic Ocean---Greenland, Iceland, Norwegian Seas (GIN Sea) system and possible mechanisms driving variability. The theoretical foundation of this work is the theory of Proshutinsky & Johnson [1997] that two major climate states of the Arctic---Anticyclonic Circulation Regime (ACCR) and Cyclonic Circulation Regime (CCR)---are driven by variations in the freshwater contents of the Arctic Ocean and the GIN Sea. It is hypothesized that the Arctic Ocean and the GIN Sea form an auto-oscillatory ice-ocean-atmosphere climate system with a quasi-decadal period of interannual variability. The system is characterized by two stages: (1) cold Arctic (ACCR)---warm GIN Sea with weak interaction between the basins; (2) warm Arctic (CCR)---cold GIN Sea with intense interaction between the basins. Surface air temperature and dynamic height gradients between the basins drive the auto-oscillations. This study investigates interactions between the Arctic Ocean and the GIN Sea. To test the hypothesis, a simple model of the Arctic Ocean and Greenland Sea has been developed. The Arctic shelf processes have been parameterized in a box model coupled with an Arctic Ocean module. Both the Arctic Ocean and Greenland Sea modules are coupled with a thermodynamic ice model and atmospheric models. Several model experiments have been conducted to adjust the model and to reproduce the auto-oscillatory behavior of the climate system. One of the major results of this work is the simulation of auto-oscillatory behavior of the Arctic Ocean---GIN Sea climate system. Periodical solutions obtained with seasonally varying forcing for scenarios with high and low interaction between the regions reproduce major anomalies in the ocean thermohaline structure, sea ice volume, and fresh water fluxes attributed to ACCR and CCR regimes. According to the simulation results, the characteristic time scale of the Arctic Ocean---GIN Sea system variability reproduced in the model is about 10--15 years. This outcome is consistent with theory of Proshutinsky and Johnson [1997] and shows that the Arctic Ocean---GIN Sea can be viewed as a unique auto-oscillating system.
• Heat And Freshwater Controlling Processes On The Northern Gulf Of Alaska Shelf

We examined conditions and processes that control the distribution of heat and freshwater on the northern Gulf of Alaska (GOA) shelf. Cross-shelf heat gradients are weak throughout the year, while salinity gradients are substantial due to the impact of coastal freshwater runoff. Outer shelf water properties are influenced by large anticyclonic eddies, while the inner and middle shelves may be regulated by wind and freshwater runoff dynamics around the Alaska Coastal Current (ACC). On the outer shelf, anticyclonic eddies propagate from the eastern GOA southwestward along the continental slope, where they favor on-shelf (off-shelf) transport of saline and nutrient-rich (fresh and iron-rich) waters Certain along-shelf locations are identified where low-salinity coastal waters are found near the shelfbreak within reach of eddies and may be regions of enhanced cross-shelf freshwater transport. The eddies have lifetimes of ~5 years and increase in size and sea level anomaly west of the Seward Line, which implies more vigorous eddy cross-shelf exchange in the northwestern GOA. By comparison, on the inner shelf the heat and freshwater distribution is dominated by large coastal river runoff, which forces the ACC and controls the vertical distribution of temperatures through stratification. In May 2007, the coastal GOA revealed some of the lowest ocean temperatures since the early 1970s, initiated by strong atmospheric cooling and reduced coastal runoff in November 2006. Stepwise regression shows that 81% of the variability of deep temperatures is explained by salinity stratification and air-sea heat fluxes. Weak baroclinic flow in May 2007 likely aided the cooling through reduced along-shore heat transport. A more detailed examination of heat transport indicated that along-shore heat flux convergence in the ACC may re-supply 10-35% of the heat removed by air-sea fluxes throughout the coastal GOA cooling season, while the annual mean cross-shore heat flux convergence is insignificant. Spatial gradients show increasing heat fluxes from off- to on-shore and from east to west. The cross-shore gradients result from wind speed gradients due to ageostrophic near-shore wind jets near coastal mountains, while the along-shore gradients result from larger-scale pressure systems. While the ACC advects coastal freshwater around the GOA shelf its waters are subjected to disproportional heat loss west of the Seward Line.
• Idealized Modeling Of Circulation Under Landfast Ice

Idealized analytical and numerical models are used to elucidate the effects of a spatially variable landfast ice cover on under-ice circulation. Three separate forcing mechanisms are investigated; lateral inflow onto an ice-covered shelf (an elevated sea level at the western boundary), a spatially uniform upwelling wind blowing along the seaward landfast ice edge and a buoyant inflow under the ice cover that enters the domain through the southern coastal wall. The idealized models are configured to resemble the shallow Alaskan Beaufort Sea shelf. Models show that the inclusion of landfast ice means shelf response is substantially different from an ice-free shelf. In the case of a lateral inflow, landfast ice spreads the inflow offshore (in a manner similar to bottom friction) but the change in surface stress across the ice edge (from ice-covered to ice-free) limits the offshore spreading. In the case of an upwelling wind along the ice edge, the low sea level at the ice edge (due to ice edge upwelling) leads to a cross-shore sea level slope between the coast (high sea level) and the ice edge (low sea level), which drives a geostrophically balanced flow upwind. In the absence of along-shore changes in wind or ice the circulation does not vary along the shelf and currents near the coast are O(10 -3) m s-1. Along- and cross-shore variations in the ice-ocean friction coefficient introduce differences in the response time of the under-ice flow and can lead to along-shore sea level slopes, which drive along-shore flows near the coast (< 0.06 m s-1). In the case of a time dependent buoyant inflow, the landfast ice spreads the buoyant inflow much farther offshore (~ 9 times the local baroclinic Rossby radius, ~ 45 km) than in the ice-free case (< 30 km). When the ice width is finite, the change in surface across the ice edge acts to restrict offshore flow (in the anti-cyclonic bulge) and inhibits onshore flow farther downstream.
• Idealized Modeling Of Seasonal Variation In The Alaska Coastal Current

Analytical and idealized-numerical models were used to understand the physical processes that govern the seasonal variation and fate of the freshwater in the Alaska Coastal Current (ACC). The ACC is forced by freshwater inflow and by mean easterly winds that cause downwelling over the shelf. Two-dimensional modeling using a line-source buoyant inflow gives the coastal current depth $<f> H=<fr><nu>3<sup>2/3</sup></nu><de>2</de></fr><fen lp="par"><fr><nu> f<sup>2</sup>Q<sup>2</sup></nu><de>g<sup>'</sup></de></fr> <rp post="par"></fen>t<sup>2/3</sup></f>$ and coastal current width $<f> Y<inf>2D</inf>=3<sup>1/3</sup><fen lp="par"><fr><nu>g<sup>' </sup>Q</nu><de>f<sup>2</sup></de></fr><rp post="par"></fen><sup> 1/3</sup>t<sup>1/3</sup></f>$, where f is the Coriolis frequency, g ' is reduced gravity, Q is inflow rate and t is time since inflow began. Addition of downwelling wind-stress causes a steep coastal current front that intersects the bottom and is either convecting, stable and steady, or stable and oscillatory depending on $<f> <fr><nu>D</nu><de><g>d</g><inf>*</inf></de></fr></f>$ and $<f> <fr><nu>b<inf>y</inf></nu><de>f<sup>2</sup></de></fr></f>$, where D is bottom depth, delta* is an Ekman depth and by is the cross-shelf buoyancy gradient. Three-dimensional modeling of a half-line source initially develops two-dimensionally but becomes three-dimensional from a balance between coastal influx of buoyancy and its downstream transport. This balance results in a coastal current depth limit $<f> H<inf><rf>max</rf></inf>=<fen lp="par"><fr><nu>2Qf</nu><de>g<sup> '</sup></de></fr><rp post="par"></fen><sup>1/2</sup>x<sup> 1/2</sup></f>$, where x is along-shelf distance. This limit is unchanged under downwelling wind-stress and is reached on time scales of less than 1 month for the ACC. The half-line source coastal current width develops as $<f> Y<inf>2D</inf></f>$ away from the beginning of the line source. Imposition of a downwelling wind-stress tau results in an approximate balance among wind-stress and along- and cross-shelf momentum advection so that the current width is reduced to $<f> Y<inf>wind</inf>&ap;L<inf>D</inf><fen lp="par"><fr><nu>Qf</nu> <de><g>t</g>/<g>r</g><inf>0</inf></de></fr><rp post="par"></fen><sup> 1/2</sup></f>$, where LD is the Rossby radius of deformation. Waves and eddying motions eventually grow in the half-line source coastal current with wavelengths proportional to the coastal current width and with a downstream phase speed slower than the maximum current speed. These features cause an offshore flux of buoyant water, a broader coastal current and accumulation of buoyancy on the shelf. Increasing downwelling wind stress reduces the effects of the instabilities. Although buoyancy accumulates on the shelf during most model runs, there is little accumulation under maximum winter downwelling wind-stress. This suggests that freshwater accumulates on the shelf from spring through fall, but is then transported downstream during winter.
• Numerical Method For Tsunami Calculation Using Full Navier -Stokes Equations And The Volume Of Fluid Method

A two-dimensional numerical model was developed to study tsunami wave generation, propagation and runup. The model is based on solving the Navier-Stokes (NS) equations. The free-surface motion is tracked using the Volume of Fluid technique. The finite difference two-step projection method is used to solve NS equations and the forward time difference method to discretize the time derivative. A structured mesh is used to discretize the spatial domain. The model has been conceived as a versatile, efficient and practical numerical tool for tsunami computation, which could address a comprehensive understanding of tsunami physics with the ultimate aim of mitigating tsunami hazards. The prediction capability of tsunami generation, propagation and runup is improved by including more accurately the effects of vertical velocity/acceleration, dispersion and wave breaking. The model has the capability to represent complex curved boundaries within a Cartesian grid system and to deal with arbitrary transient-deformed moving boundaries. The numerical model was validated using laboratory experiments and analytical solutions. The model was used as a tool to determine the adequacy of the shallow water (SW) approximation in the application of tsunami simulations. Numerical results were compared with experimental data, analytical solutions and SW results in terms of the time-history free surface elevations and velocity. Reasonable agreements were observed based on the spatial and temporal distributions of the free surface and velocity.
• Nutrient Dynamics In The Northern Gulf Of Alaska And Prince William Sound: 1998--2001

The northern Gulf of Alaska (GOA) shelf is a productive coastal region that supports several commercially important fisheries. The mechanisms supporting such high levels of productivity over this shelf are not understood, however, since it is a downwelling-dominated shelf. In an effort to understand the mechanisms underlying such high biological productivity, nutrient distributions were determined 25 times throughout 1998, 1999, 2000, and 2001 from over the northern GOA shelf and in Prince William Sound (PWS). Deep water (>75 m) nitrate, silicate and phosphate concentrations were positively correlated with salinity indicating an offshore nutrient source. The average annual cycle was established, in which nitrate, silicate and phosphate responded seasonally to physical and biological processes. Ammonium concentrations were generally low and uniform (<1.2 muM) with occasional patches of higher concentrations. During each summer, an onshore flux of dense nutrient-rich bottom water onto the shelf was evident when the downwelling relaxed. This seasonal flux created nutrient reservoirs over the deeper shelf regions that were eventually mixed throughout the water column during the winter months. This annual evolution may be vital to the productivity of this shelf. A large degree of interannual variability was found during the study, which included El Nino (1998) and La Nina (1999) years. Spring phytoplankton biomass over the shelf was highest in 2000 when the upper waters were nutrient enriched and strongly stratified. The highest phytoplankton biomass was measured in May 1999 during the passage of a slope eddy, which demonstrated the potential of these phenomena to greatly enhance primary productivity. A large degree of spatial variability was also found, both cross-shelf and along-shelf. Hinchinbrook Canyon was found to consistently have high salinity, nutrient-enriched bottom waters suggesting it plays an important role in the transport of slope waters onto the shelf and probably into PWS. Along-shelf trends were found in the upper coastal waters in the winter and spring, with higher salinities, temperatures, and nutrient concentrations upstream of PWS. The nutrient dynamics were similar in PWS and over the shelf/slope in 2001; however, nutrient drawdown, followed by depletion, and the spring bloom appeared earlier and stronger in PWS.
• Processes Controlling Radon-222 And Radium-226 On The Southeastern Bering Sea Shelf (Chemical Oceanography, Two-Dimensional Model, Continental, Gas-Exchange, Sediment Flux)

An investigation was made into the use of ('222)Rn and ('226)Ra as tracers of air-sea gas exchange, water column mixing and sediment-water exchange on the southeastern Bering Sea shelf. Furthermore, a two-dimensional model was developed to unify these three processes into a coherent picture of ('222)Rn flux out of the sediments, through the water column and into the atmosphere. The best time period to average wind speeds when regressing them against gas transfer coefficients was found to be 3.3 days by a linear regression optimization, approximately the synoptic time scale of storms in the southeastern Bering Sea. A statistically significant relationship between averaged wind speed and transfer coefficients was found at the 80% confidence level. Gas transfer coefficients were found to be obscured in shallow waters by radon flux from the sediments. Two-dimensional mixing in these continental shelf waters rendered the traditional one-dimensional vertical mixing model of excess ('222)Rn unable to obtain reliable vertical eddy diffusivities. Exchange across the sediment-water interface was calculated from the deficiency of ('222)Rn measured in sediment cores, the standing crop of excess ('222)Rn in the overlying water column and the ('222)Rn production rate of sediment surface grab samples. The flux of radon out of the sediments was found to increase in the onshore direction. Biological irrigation appears to be the primary exchange mechanism between the sediment and water columns on this shelf. Distributions in the water column show finestructure reported previously as well as biological removal of ('226)Ra. A (chi)('2) hypersurface search found the optimal horizontal and vertical eddy diffusivities that explained the two-dimensional distribution of ('222)Rn provided from a kriging estimation exercise on the data measured in this study. This model was essentially a hybrid of a least squares surface fit and a numerical integration of the governing differential equation of ('222)Rn. When considered as a two-dimensional system in the cross-shelf direction, the rates of gas exchange, water column mixing and sediment-water exchange agree with each other to an acceptable degree.
• The carbon cycle in an anoxic marine sediment: Concentrations, rates, isotope ratios, and diagenetic models

The carbon cycle in the anoxic sediments of Skan Bay, Alaska, was investigated in order to better understand the processes that control biogeochemical transformations in an organic-rich sediment environment. Depth distributions of concentration and $\delta\sp{13}$C were determined for five major carbon reservoirs: methane (CH$\sb4$), dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), particulate inorganic carbon (PIC), and particulate organic carbon (POC). In addition, methane oxidation and sulfate reduction rates were measured under quasi-in situ conditions using radio-tracer techniques. Diagenetic models were applied to concentration, reaction rate, and isotope ratio depth distributions and the results were integrated into a comprehensive, depth-dependent model of the Skan Bay carbon cycle that considered advective, diffusive, and biological and chemical reactive fluxes for the five major carbon reservoirs. The Skan Bay carbon cycle is fuelled by POC, which is deposited at the sediment surface at a rate of 2290 $\pm$ 480 umol $\cdot$ cm$\sp{-2}$ $\cdot$ yr$\sp{-1}$. Isotope mass-balance calculations indicate that about 60% of this material is derived from kelp while the remainder originates as phytoplankton. About 60% of the organic matter is consumed in the upper 40 cm of the sediment column. The $\delta\sp{13}$C-POC and $\delta\sp{13}$C-DOC depth distributions suggest that the material derived from kelp is more labile, accounting for greater than 60% of the total POC consumption. The products of anaerobic metabolism of POC accumulate in the DOC reservoir creating a large DOC concentration gradient at the sediment-water interface. Flux and stable carbon isotope mass-balance calculations suggest that a sizable portion (30 to 80%) of the DOC produced by degradation of POC diffuses from the sediment prior to oxidation to dissolved inorganic carbon. Methane production appears to occur primarily at depths greater than 40 cm. The CH$\sb4$ diffuses upward and is almost quantitatively oxidized to DIC in a narrow subsurface zone. Methane oxidation accounts for only 20% of the DIC production, but exerts a profound influence on the $\delta\sp{13}$C-DIC profile, contributing to the distinct mid-depth minimum. Pore waters are supersaturated with respect to calcite at depths greater than 10 cm, but isotope mass-balance considerations indicate that carbonate mineral formation is not occurring in these sediments.
• Variability In The Circulation, Temperature, And Salinity Fields Of The Eastern Bering Sea Shelf In Response To Atomospheric Forcing

Although the Bering Sea shelf plays a critical role in mediating the global climate and supports one of the world's largest fisheries, fundamental questions remain about the role of advection on its salt, fresh water, heat and nutrient budgets. I quantify seasonal and inter-annual variability in the temperature, salinity and circulation fields. Shipboard survey temperature and salinity data from summer's end reveal that advection affects the inter-annual variability of fresh water and heat content: heat content anomalies are set by along-shelf summer Ekman transport anomalies whereas fresh water content anomalies are determined by wind direction anomalies averaged over the previous fall, winter and early spring. The latter is consistent with an inverse relationship between coastal and mid-shelf salinity anomalies and late summer -- winter cross-shelf motion of satellite-tracked drifters. These advection anomalies result from the position and strength of the Aleutian Low pressure system. Mooring data applied to the vertically integrated equations of motion show that the momentum balance is primarily geostrophic within at least one external deformation radius of the coast. Local accelerations, wind stress and bottom friction account for < 20% (up to 40%) of the along- (cross-) isobath momentum balance, depending on location and season. Wind-forced surface Ekman divergence is primarily responsible for flow variations. The shelf changes abruptly from strong coastal convergence conditions to strong coastal divergence conditions for winds directed to the south and for winds directed to the west, respectively, and substantial portions of the shelf's currents reorganize between these two modes of wind forcing. Based on the above observations and supporting numerical model integrations, I propose a simple framework for considering the shelf-wide circulation response to variations in wind forcing. Under southeasterly winds, northward transport increases and onshore cross-isobath transport is relatively large. Under northwesterly winds, onshore transport decreases or reverses and nutrient-rich waters flow toward the central shelf from the north and northwest, replacing dilute coastal waters that are carried south and west. These results have implications for the advection of heat, salt, fresh water, nutrients, plankton, eggs and larvae across the entire shelf.