Shinzou Fujio

Diagnostic calculation for circulation and water mass movement in the deep Pacific

The steady circulation of the deep Pacific is estimated with a robust diagnostic model, which is internally constrained by hydrographic data. It is shown that the input data should be modified to fit the model in inverse proportion to the Coriolis parameter because a density field inconsistent with the model generates unrealistic geostrophic flows. The model reproduces most of the deep currents previously reported, such as the deep western boundary current east of New Zealand. In addition, as a new feature, the present model diagnoses an anticyclonic circulation around the East Pacific Rise. This circulation is discovered to be associated with a rise of isopycnals at middepth. Tracking of many particles in the diagnosed velocity field reveals that two water masses enter the Southwest Pacific Basin. One is the deep water of the South Indian Basin which enters through a gap to the south of New Zealand. The other is the upper water of the Antarctic Circumpolar Current; this water becomes dense near the Ross Sea and sinks into the deep Southeastern Pacific Basin. The anticyclonic circulation around the East Pacific Rise transports it to the Southwest Pacific Basin. These waters supply comparable volumes to the Southwest Pacific Basin; the residence time is estimated to be 86 years. The deep water in the Southwest Pacific Basin is brought northward rapidly by the deep western boundary current east of New Zealand; it takes only a few decades to move from the east of New Zealand to the North Pacific.

Deep current structure above the Izu‐Ogasawara Trench

Above the Izu-Ogasawara Trench south of Japan, direct current measurements were made at 34°N from 1987 to 1996, and hydrographic observations were carried out at 34° and 30°N in 1995. The geostrophic shears calculated from the conductivity-temperature-depth data were consistent with the shears calculated from the current data. It is found that there are opposing currents that flow southward on the western flank and northward on the eastern flank along the isobaths. In the cross-trench direction the magnitude of the mean velocity tended to increase with the distance from the deepest point of the trench and exceeded 10 cm s−1 on the eastern flank. In the vertical direction the mean velocity increased with depth, but it decreased just above the bottom, probably because of friction. The southward transport above the western flank was estimated to be 5–8 Sv both at 34° and 30°N. However, the northward transport above the eastern flank increased from 5 Sv at 30°N to 22 Sv at 34°N, suggesting a large inflow from the west.

World ocean circulation diagnostically derived from hydrographic and wind stress fields: 2. The water movement

The water movement in the world ocean is examined using a velocity field diagnostically derived from hydrographic and wind stress data. The movement of water is calculated numerically by tracking its constituent particles, and the attention is focused on where the water originates. The particle tracking indicates that deep water is formed not only in the North Atlantic and the Weddell Sea but also in the Ross Sea. The North Atlantic Deep Water contributes considerable volume to all the deep basins of the world ocean, while the Weddell Sea Deep Water (WSDW) tends to be confined to the southern region of the southern ocean. Similarly, the Ross Sea Deep Water (RSDW) does not flow northward in the present model because the Antarctic circumpolar current (ACC) penetrates sufficiently deep to carry the RSDW away to the South Atlantic. It is also found that the WSDW does not mix directly with the circumpolar Water because of the South Scotia Ridge, but rather mixes with the water recirculating in the Weddell Gyre, which is supplied by the middepth water of the ACC in the Atlantic-Indian Basin. Water composition in 16 basins is summarized in the form of box diagrams of deep water exchange.

Three-dimensional numerical experiments on tidal exchange through a narrow strait in a homogeneous and a stratified sea

We have investigated the three-dimensional Lagrangian motion of water particles related with tidal exchange between two basins with a constant depth connected through a narrow strait and the effects of density stratification on the exchange processes by tracking a number of the labeled particles. Tide-induced transient eddies (TITEs), which are similar to those in two-dimensional basin, are generated behind the headlands. Upwelling appears around the center of the eddy and sinking around the boundary. When the basins are filled with homogeneous water, a pair of vortices are produced in the vertical cross section of the strait due to bottom stress, with upwellings along the side walls of the strait and sinking in the center of the strait. These circulations form the horizontally convergent field in the cross-strait direction in the upper layers while the horizontal divergence takes place in the bottom layer. These vertical water-motions produce the three-dimensional distribution of velocity shear and phase lag of the tidal current around the strait, and the Lagrangian drifts of water particles become large. As a result, water exchange through the strait is greatly enhanced: The water exchange rate reaches 94.1% which is much larger than that obtained in the vertically integrated two-dimensional model. When the basins are stratified, the stable stratification suppresses the vertical motion so that a pair of vertical vortices are confined in the lower layers. This leads to a decrease in the exchange rate, down to 88.6%. Our numerical results show that the three-dimensional structure of tidal currents should be taken into account in tidal exchange through a narrow strait.

Study of Water Motion at the Dissolved Oxygen Minimum Layer and Local Oxygen Consumption Rate from the Lagrangian Viewpoint

By using the Euler-Lagrangian method, we examine water movements within the layer of minimum oxygen concentration and estimate local oxygen consumption rates for 15 regions of the global ocean. To do this, a number of labeled particles (which represent water parcels) are deployed at the center of a grid with 15 depth levels and tracked backward in time for 50 years in a three-dimensional velocity field. We assume that a particle picks up oxygen when it encounters the point of maximum oxygen concentration along the 50 years segment of its path. We introduce a contribution rate from waters distributed throughout the global ocean to the oxygen concentration of a local layer under consideration. Water parcels which are assumed to pick up oxygen within the oxygen minimum layer of an oceanic region under consideration make a very small contribution to the overall oxygen concentration of this layer. In addition, these parcels move out of the layer and water parcels from the upper layers take their place. The averaged Lagrangian local oxygen consumption rate is 0.033 ml/l/yr for the depth of the oxygen minimum layer, 0.20 ml/l/yr at 100 m depth (euphotic layer), 0.043 ml/l/yr for layers from 150 m to 800 m depth and 0.012 ml/l/yr for deep layers from 800 m to 3000 m. The present Lagrangian numerical experiment produces a maximum difference between observed and calculated concentrations of oxygen and, therefore, a maximum oxygen consumption rate. Although the present method has an ambiguity as to how oxygen is picked up, we nevertheless were able to identify regions in which the water parcels pick up oxygen of maximum concentration. We found that the South Equatorial Current (SEC) transports oxygen of higher concentration to the middle latitude regions of both the North Atlantic and the North Pacific across the equator.

Effect of bottom slope in northeastern North Pacific on deep-water upwelling and overturning circulation

Hydrographic data show that the meridional deep current at 47°N is weak and southward in northeastern North Pacific; the strong northward current expected for an upwelling in a flat-bottom ocean is absent. This may imply that the eastward-rising bottom slope in the Northeast Pacific Basin contributes to the overturning circulation. After analysis of observational data, we examine the bottom-slope effect using models in which deep water enters the lower deep layer, upwells to the upper deep layer, and exits laterally. The analytical model is based on geostrophic hydrostatic balance, Sverdrup relation, and vertical advection–diffusion balance of density, and incorporates a small bottom slope and an eastward-increasing upwelling. Due to the sloping bottom, current in the lower deep layer intensifies bottomward, and the intensification is weaker for larger vertical eddy diffusivity (KV), weaker stratification, and smaller eastward increase in upwelling. Varying the value of KV changes the vertical structure and direction of the current; the current is more barotropic and flows further eastward as KV increases. The eastward current is reproduced with the numerical model that incorporates the realistic bottom-slope gradient and includes boundary currents. The interior current flows eastward primarily, runs up the bottom slope, and produces an upwelling. The eastward current has a realistic volume transport that is similar to the net inflow, unlike the large northward current for a flat bottom. The upwelling water in the upper deep layer flows southward and then westward in the southern region, although it may partly upwell further into the intermediate layer.

Pathway and variability of deep circulation around 40°N in the northwest Pacific Ocean

To clarify the global deep-water circulation in the northwest Pacific, we conducted current observations with seven moorings at 40°N east of Japan from May 2007 to October 2008, together with hydrographic observations. By analyzing the data, while taking into consideration that the deep circulation has a northward component in this region and carries low-silica, high-dissolved-oxygen water, we clarified that the deep circulation flows within the region between 144°30′ and 146°10′E at 40°N on and east of the eastern slope of the Japan Trench with marked variability; the deep circulation flows partly on the eastern slope of the trench and mainly to the east during P1 (10 May–24 November 2007), is confined to the eastern slope of the trench during P2 (25 November 2007–20 May 2008), and flows on and to the immediate east of the eastern slope of the trench during P3 (21 May–15 October 2008). Previous studies have identified two branches of the deep circulation at lower latitudes in the western North Pacific; one flows off the western trenches and the other detours near the Shatsky Rise. It was thus concluded that the eastern branch flows westward at 38°N and then northward to the east of the trench, finally joining the western branch around 40°N during P1 and P3, whereas the eastern branch passes westward south of 38°N, joins the western branch around 38°N, and flows northward on the eastern slope of the trench during P2.

Diagnostic Approaches on Deep Ocean Circulation

Abstract Three different approaches on the diagnostic study of deep ocean circulation are discussed. Firstly, the time scale of the circulation and the age of the water are studied by using a reservoir model. The results demonstrate that the multi-layered structure of the deep layer is essential to produce the middepth age maximum which is found in the North Pacific Ocean. Secondly, the diagnostic method to obtain reasonable flow field from limited numbers of direct current observations is proposed. The method is applied to the deep layer in the Philippine Sea, and the obtained flow field appears very reasonable. Lastly, by using climatological data of density field, the current field in the deep Pacific Ocean is obtained by following the diagnostic method given by Sarmiento and Bryan (1982), and the exchanges of water masses among sub-basins in the deep Pacific are discussed by Lagrangian tracking method of water particles. Generally speaking, the results are consistent with those obtained by various investigators.