摘要:

Heat fluxes at the ice‐ocean interface and ice thickness are investigated by comparing field data from the Coordinated Eastern Arctic Experiment (CEAREX) drift phase with model calculations. The calculations are based on two types of models. The first one is a one‐dimensional ice‐ocean model with high vertical resolution. This model is based on the conservation equations for heat, salt, and momentum and uses turbulence models to achieve closure. A discrete element approach is also introduced to explicitly parameterize the ice roughness. The second model is a simple one‐dimensional bulk heat transfer model. In this version, the interfacial salinity is modelled on the basis of salt conservation at the ice‐ocean interface. The bulk heat transfer model is then calibrated using the former model. The two models predict ocean heat fluxes that are quite variable in time owing to short‐term variations in the ice drift. Both models calculate realistic ice thicknesses. It is demonstrated that the observed time variation in ice thickness from eight different experimental sites with varying initial thicknesses and bottom topographies can be reproduced by applying bulk heat transfer coefficients in the range (2.8 ± 1) × 10−4. Horizontal variation of the thermal state within a single pack ice floe results in simultaneous freezing and melting over relatively small spatial scales. When modeling or averaging ice data in space these aspects need t

ISSN:0148-0227    

DOI:10.1029/92JC00815

年代:1992

数据来源: WILEY

摘要:

Data from an accelerometer package are used to estimate the rates at which interaction events occur between adjacent ice floes in the marginal ice zone. Observations suggest that the events are principally the result of a modified collision process, where brash ice partially filling the gap between two floes acts as an intermediary in the transfer of momentum from one to the other. A model is presented from which the rate at which these interactions occur can be predicted. Comparison of the rates estimated from the data with those predicted by the model shows there to be some agreement, although there is considerable room for improvement. The predicted event rate depends strongly on the ratio of the root‐mean‐square wave height to the amount of open water between pairs of floes. The form of this dependence offers an explanation for grouping of events, a feature that may be readily observed in many extracts of d

ISSN:0148-0227    

DOI:10.1029/92JC00152

年代:1992

数据来源: WILEY

摘要:

A highly simplified analytical model of sea ice growth is presented in which the atmosphere is in thermal radiative equilibrium with the ice. This makes the downwelling longwave radiation reaching the ice surface an internal variable rather than a specified forcing. The ice is characterized by only two variables, thickness and surface temperature. The seasonal cycle is resolved into four steps, in which either the thickness or the temperature can vary. An ocean mixed layer of fixed depth but variable temperature provides a heat sink if the ice melts away during the summer. Analytical results show how the ice state depends on properties of the ice and on the externally specified climate. The main result is that this bare bones physics is enough to produce an annual cycle of temperature and thickness not unlike today's conditions. The results also suggest how much the climate must change to produce qualitatively different cycles. Reducing the atmospheric poleward heat transport by 20 W m−2is enough to make the ice so thick (>12 m) that it never warms up to the melting point in summer. Increasing the poleward flux by 30 W m−2is enough to produce an arctic ocean which stores enough heat in summer to prevent any ice forming in winter. The model predicts that today's climate can support either a perennial ice pack or an ice free Arctic. However, a winter ice pack which melts completely each summer is not possible under any clim

ISSN:0148-0227    

DOI:10.1029/92JC00695

年代:1992

数据来源: WILEY

摘要:

A fetch‐dependent boundary‐layer model, driven by observed temperature sounding data, is used to examine theoretical heights of buoyant convection (H) above open leads in the wintertime pack ice of the central Arctic. Assuming wet adiabatic ascent with no entrainment or friction,His estimated as the height at which the model‐predicted equivalent potential temperature at saturation above a lead (θel) intersects with the same value of equivalent potential temperature at saturation (θe) derived from vertical sounding profiles.Hincreases with increasing lead width. For a 1000‐m lead, the widest which can be reasonably expected for the central Arctic, the median value ofHis approximately 1000 m, slightly below the median top of the low‐level Arctic temperature inversion layer. WhileHshows large variability, events of convection up to 4 km, as recently observed from lidar backscatter data, appear to be fairly rare. First, these events require an open lead of at least 10,000 m. Second, whileHtends to be largest under conditions of low surface wind speed, low surface temperature, and a weak low‐level temperature inversion, this combination appears to be atypical of Arctic conditions. Third, while the meteorological conditions that should favor the development of open leads tend to minimizeH, conditions favoring largeHare also those in which any newly developed leads will qui

ISSN:0148-0227    

DOI:10.1029/92JC00688

年代:1992

数据来源: WILEY

摘要:

Between September 1989 and 1990, twenty‐six current records were collected by instruments on eight moorings located in Pribilof and Zhemchug canyons, and at a site midway between these features. These records provide the first long‐term Eulerian measurements from the slope and mid‐slope of the eastern Bering Sea. Results from the current records, together with water property observations, permit a characterization of the Bering Slope Current. Moderate flow (∼2 to 18 cm s−1) followed the bathymetry toward the northwest and existed primarily in the upper 300 m. Wind and current energy increased in winter, but vector mean current did not increase at all sites. Wind forcing accounted for only a small fraction of the current fluctuations. At one mid‐slope location in Pribilof Canyon, bathymetry resulted in rectification of the strong daily tidal current. Estimates of heat and salt fluxes indicate some significant shoreward transport; however, this flux did not occur preferentially in

ISSN:0148-0227    

DOI:10.1029/92JC00512

年代:1992

数据来源: WILEY

摘要:

A two‐dimensional (latitude‐depth) deep ocean model is presented which is coupled to a sea ice model and an Energy Balance Climate Model (EBCM), the latter having land‐sea and surface‐air resolution. The processes which occur in the ocean model are thermohaline overturning driven by the horizontal density gradient, shallow wind‐driven overturning cells, convective overturning, and vertical and horizontal diffusion of heat and salt. The density field is determined from the temperature and salinity fields using a nonlinear equation of state. Mixed layer salinity is affected by evaporation, precipitation, runoff from continents, and sea ice freezing and melting, as well as by advective, convective, and diffusive exchanges with the deep ocean. The ocean model is first tested in an uncoupled mode, in which hemispherically symmetric mixed layer temperature and salinity, or salinity flux, are specified as upper boundary conditions. An experiment performed with previous models is repeated in which a mixed layer salinity perturbation is introduced in the polar half of one hemisphere after switching from a fixed salinity to a fixed salinity flux boundary condition. For small values of the vertical diffusion coefficientKV, the model undergoes self‐sustained oscillations with a period of about 1500 years. With larger values ofKV, the model locks into either an asymmetric mode with a single overturning cell spanning both hemispheres, or a symmetric quiescent state with downwelling near the equator, upwelling at high latitudes, and a warm deep ocean (depending on the value ofKV). When the ocean model is forced with observed mixed layer temperature and salinity, no oscillations occur. The model successfully simulates the very weak meridional overturning and strong Antarctic Circumpolar Current at the latitudes of the Drake Passage. The coupled EBCM‐deep ocean model displays internal oscillations with a period of 3000 years if the ocean fraction is uniform with latitude andKVand the horizontal diffusion coefficient in the mixed layer are not too large. Globally averaged atmospheric temperature changes of 2 K are driven by oscillations in the heat flux into or out of the deep ocean, with the sudden onset of a heat flux out of the deep ocean associated with the rapid onset of thermohaline overturning after a quiescent period, and the sudden onset of a heat flux into the deep ocean associated with the collapse of thermohaline overturning. When the coupled model is run with prescribed parameters (such as land‐sea fraction and precipitation) varying with latitude based on observations, the model does not oscillate and produces a reasonable deep ocean temperature field but a completely unrealistic salinity field. Resetting the mixed layer salinity to observations on each time step (equivalent to the “flux correction” method used in atmosphere‐ocean general circulation models) is sufficient to give a realistic salinity field throughout the ocean depth, but dramatically alters the flow field and associated heat transport. Although the model is highly idealized, the finding that the maximum perturbation in globally averaged heat flux from the deep ocean to the surface over a 100‐year period is 1.4 W m−2suggests that effect of continuing greenhouse gas increases, which could result in a heating perturbation of 10 W m−2by the end of the next century, will swamp possible surface heating perturbations due to changes in oceanic circulation. On the other hand, the extreme sensitivity of the oceanic flow field to variations in precipitation and evaporation suggests that it will not be possible to produce accurate projections of regional climatic change

ISSN:0148-0227    

DOI:10.1029/92JC00718

年代:1992

数据来源: WILEY

摘要:

This paper provides a detailed hydrographic climatology for the shallow northwestern Arabian Sea prior to and during the southwest monsoon, presented as multiple‐year composite vertical hydrographic sections based on National Oceanographic Data Center historical ocean station data. Temperature and salinity measurements are used to infer the water masses present in the upper 500 m. The hydrographic evolution depicted on bimonthly sections is inferred to result from wind‐driven physical processes. In the northwestern Arabian Sea the water mass in the upper 50 m is the Arabian Sea Surface Water. Waters from 50 to 500 m are formed by mixing of Arabian Sea Surface Water with Antarctic and Indonesian intermediate waters. The inflow of Persian Gulf Water does not significantly influence the hydrography of the northwestern Arabian Sea along the Omani coast. Nitrate has a high inverse correlation with temperature and oxygen in the premonsoon thermocline in the depth interval 50–150 m. During the southwest monsoon, coastal upwelling off Oman and adjacent offshore upward Ekman pumping alter the shallow hydrog

ISSN:0148-0227    

DOI:10.1029/92JC00813

年代:1992

数据来源: WILEY

摘要:

After the Agulhas Current separates from the African coastline, it forms the Agulhas Retroflection and flows back into the Indian Ocean. If the current loop closes onto itself, a warm Agulhas ring is pinched off in the area south of Africa (between 15° and 20°E longitude). These rings can be monitored relatively easily with satellite altimetry when they migrate into the South Atlantic. However, the shedding process itself is difficult to extract from altimeter measurements. Geosat Exact Repeat Mission altimeter data of the period from November 1986 to September 1989 are used to analyze the ring‐shedding process in the Agulhas Current System. The data set has been processed using the collinear method over the area which extends from 30°W to 90°E, and from 15°S to 50°S. The original altimeter data are statistically interpolated to a regular 1° × 1° (spatial) grid using objective analysis. The obtained sea level anomaly maps are computed at regular time steps. Combination of harmonic analysis and principal component analysis enables extraction of typical Agulhas Current frequencies from the 3‐year altimetric data set. The analysis indicates that 11 or 12 dominant events took place over the 3‐year period. Variations occur from year to year. February, March, and April (summer‐autumn) are identified as anomalous months in which the sea level anomaly signal is not as well defined as in the other part of the year. The combined analyses showed that the first three modes of variability (explaining 25%, 20%, and 12% of the variance, respectively) are dominant. The structure of the spatial and temporal scales leads to the hypothesis that these modes can be associated with periodic Agulhas front movements, culminating in the formation of large Agulhas rings. Sharp changes (pulses) in the stability of the sea level pattern then determine the time of pinch off. This indicates a number of 18 (±2) pulses over the 3‐year period in which an anomaly in the average decorrelation time occured. If these pulses can be connected to the formation of Agulhas rings, they are of great importance for the large‐scale circulation because they contribute significantly to the energy and freshwater flux between the Indian Ocean a

ISSN:0148-0227    

DOI:10.1029/92JC00736

年代:1992

数据来源: WILEY

摘要:

Because of their importance in the ocean energetics and general circulation, a proper representation of ring generation mechanisms and evolution in numerical models is crucial for an accurate picture of the heat, salt, and energy budgets. Ring locus, lifetime, propagation speed, radius, and interface displacement statistics are derived for four widely used eddy‐resolving numerical models and compared to the 10‐year statistical analysis of Gulf Stream rings based on time series of satellite infrared determinations performed by Brown et al. (1986). The ring formation process and behavior in ocean numerical models depends upon the governing equations, the vertical coordinate, and the boundary conditions used. It is shown that as more terms are retained in the model equations, the mid‐latitude jet becomes more unstable, and its interior penetration as well as the associated number of rings formed are reduced. Rings in the layer model have slower propagation speeds and longer lifetimes than their level counterparts. Such results illustrate the sensitivity of numerical ocean circulation models to the physical and numerical assumptions

ISSN:0148-0227    

DOI:10.1029/92JC00913

年代:1992

数据来源: WILEY

摘要:

A quantitative analysis of water masses in the Brazil‐Malvinas Confluence zone is performed with a least squares multiple tracer analysis using data from the confluence winter 1989 cruise. The purpose is to find the mixture of source water types that best describes the composition of a given water sample. This method is valuable in regions involving strong mixing among various source water types, as is the Brazil‐Malvinas Confluence zone. Seven main core layers are identified in this region, and all are retained for the analysis: the Thermocline Water (TW), the Subantarctic Surface Water (SASW), the Antarctic Intermediate Water (AAIW), the Upper Circumpolar Deep Water (UCDW), the North Atlantic Deep Water (NADW), the Lower Circumpolar Deep Water (LCDW), and the Weddell Sea Deep Water (WSDW). Tracers selected are temperature, salinity, dissolved nutrients, and oxygen. The results show the proportion of each source water type along four east‐west sections (35.4°S, 36.5°S, 37.9°S, 39°S). They are accurate to within 20% for all sources. The solution presents evidence of local recirculation of AAIW largely influenced by the two strong currents, Brazil and Malvinas. Southward TW and NADW separate from the coast, NADW turning eastward at a higher latitud

ISSN:0148-0227    

DOI:10.1029/92JC00484

年代:1992

数据来源: WILEY