Content uploaded by Gordon A. Mcbean
Author content
All content in this area was uploaded by Gordon A. Mcbean on Apr 19, 2021
Content may be subject to copyright.
A preview of the PDF is not available
Estimation of the Pacific Ocean meridional heat flux at 35°N
Abstract
The meridional heat flux in the North Pacific Ocean at 35°N is estimated primarily using hydrographic section data, following the method of Bryan (1962) and Bennett (1978). The meridional heat flux in the Kuroshio, computed using the Worthington and Kawai section across the current, was 1.76 PW (positive northward), with over 80% of the flux occurring in the upper 400 m. The large‐scale baroclinic heat flux across the rest of the section (145°? to North America) was —1.0 PW for the indopac (1976) section and —0.5 PW for the IOS‐72 section. The fluxes across the sections were also concentrated in the upper ocean with the upper 300 m accounting for over 75% of the flux. The mean horizontal barotropic gyre circulation results in little (0.1 PW) net heat flux because the northward‐moving water is only about 0.5°C warmer than the southward‐moving water. The contributions due to Ekman heat flux (—0.16 PW) and flow through the Japan Sea (0.13 PW) are also relatively small. The eddy heat flux is quite uncertain, but estimated to be about 0.3 PW. The total meridional heat flux, for the 1976 section, is calculated to be about 1.0 PW. The total is very dependent on the baroclinic heat flux in the highly variable Kuroshio region. The northward heat flux found in this study is more consistent with large‐scale atmospheric estimates and with Bryden et al. ‘s (1990) estimate for 24°? in the Pacific.
... e meridional transport of heat by the atmosphere, concluding that it amounts to a bit less than 2 PW in the Northern Hemisphere. Bryden et al. (1991) has calculated the heat transport in the ocean, arriving at a figure of about 1.2 PW in the Atlantic and about 0.8 PW in the Pacific. That adds to 4 PW, 1 PW short of what the satellite data suggests. McBean (1991) has also calculated the Pacific transport, concluding that it is about 1.0 PW. Clearly the uncertainty greatly exceeds the target! The largest contributor to ocean fluxes is the geostrophic flows. It is also the largest contributor to present uncertainty. However, the geostrophic flows are the most obvious ones to study and are receivin ...
... ed by the wind stress. This Ekman heat transport depends on the fact that the Ekman current is considered to be largely confined to uppermost layers of the ocean, while the compensating return flow of water is considered to be of deeper water. The Ekman current thus carries water considerably warmer than the return flow, with a resulting heat flux. McBean (1991) quotes a number of studies with a wide range of values, finally himself settling for a mean southward flux of 0.16 PW across 35°N in the Pacific. Bryden (1991), looking at 24°N, finds a northward heat transport of 0.37 PW (almost half of his total Pacific northward transport of 0.76 PW). The uncertainty seen by McBean is largely associa ...
... t to compression. (Such is the mechanism by which non-turbulent mixing takes place, for example, in stirring a cake batter.) Various efforts have been made to infer the heat transport effected by eddy fluxes (e.g. Holloway, 1986), or to measure it by determining the correlation between temperature and velocity fields (e.g. Bennett and White, 1986). McBean (1991), preferring Holloway (1986), selects a value of 0.3 PW for the North Pacific at 35°N. Evidently not negligible, and in line with the a priori expectation suggested above! (Although Bryden et al. (1981) do not include eddy transport.) The very nature of eddy mixing in the ocean has not yet been clearly elucidated and deserves a good deal ...
Concern over the future of the climate of the earth has reached a very high level. Based on the observed inexorable increase in the concentration of greenhouse gases in the atmosphere, and on computer models of the possible effects of future concentrations of these gases, initiatives are being undertaken by many governments to implement policies which would restrain the increase. Proposals are widespread for much more drastic action, involving substantial distortions of the world economy. With decisions of this magnitude resting on the advice of scientists, it is incumbent on the scientists to make that advice as well founded as possible. The climate system is an immensely complex one, with the atmosphere and the ocean each playing roles of comparable importance. However, the atmosphere is both better observed and better modelled than is the ocean, so many of the more difficult problems in dealing with the climate system concern the ocean, both in its interface with the atmosphere and internally. It is a duty of ocean scientists to identify these problems and to set about solving them. Among the more difficult, if perhaps less obvious, of these problems are those presented by fluxes to and within the ocean.
Direct estimates of ocean heat transport indicate-that the meridional
circulation across 24°N carries 1.2 PW of heat northward in the
Atlantic Ocean and 0.8 PW of heat northward in the Pacific Ocean. The
total ocean heat transport across 24°N of 2.0 PW is substantially
less than recent indirect estimates of ocean transport based on
satellite radiation measurements and atmospheric energy transport
estimates. Heat transports in the Atlantic and Pacific Oceans are
accomplished by two different mechanisms. In the Atlantic, a deep
vertical-meridional circulation cell with northward flowing warm surface
waters and southward flowing cold deep waters effects the northward heat
transport across 24°N. In the Pacific, a horizontal circulation cell
in the upper waters with northward flowing warm waters on the western
side of the ocean and southward flowing colder waters in the central and
eastern parts of the ocean effect the northward heat transport across
24°N. Prospects for direct determination of the ocean heat transport
as a function of latitude are described with emphasis on the need to
identify the principal mechanisms of ocean heat transport in each ocean
basin.
A dynamically consistent steady state circulation pattern of the North Pacific (NP) ocean is presented. The fields describing oceanic state and atmospheric forcing are obtained by fitting a steady state large-scale circulation model to the 1.0 release of World Ocean Circulation Experiment hydrology, da Silva atmospheric climatologies, TOPEX-Poseidon altimetry, and Marine Environmental Data Service drifter data. Heat, salt, and mass transports of the major circulation features of the NP are presented. All the estimates are supplied with error bars obtained via the approximate inversion of the Hessian matrix. The optimized pattern shows, in particular, that NP is heated at an average rate of 11+/-7Wm-2 and loses 26+/-18cm of fresh water from its surface per year. Zonal mean transport north of 20°N is characterized by an upwelling of 7+/-2Sv (1Sv=106m3s-1) in the intermediate layers (500-3000 m) within the latitudes 30°-50°N. A weaker overturning cell characterized by the downwelling of 3+/-1Sv at 55°-60°N contributes to the meridional circulation pattern north of 51°N. The net meridional advective heat transport in the midlatitudes is statistically undistinguishable from zero and amounts to -0.1+/-0.4PW (1PW=1015W). At 35°N, fresh water is transported southward at the rate equivalent to the average rainfall to the north of 35°N of 18+/-14cmyr-1. Analysis of the water mass transformation rates reveals the major NP intermediate water formation site to be east of Hokkaido and South Kuril Islands. A much weaker source is detected in the Alaskan Gyre. The net production rate of the NPIW source water east of Kuril Islands is estimated as 2.5 Sv.
The heat transported meridionally in the Pacific Ocean is calculated from the surface heat budgets of Clark and Weare and others; both budgets were based on Bunker's method with different radiation formulas. The meridional heat transport is also calculated from the surface heat budget of Esbensen and Kushnir, who used Budyko's method. The heat transport is southward at most latitudes if the numbers of Clark and of Weare are used. It is northward in the North Pacific and southward in the South Pacific if Eshensen and Kushnir's numbers are used. Systematic errors in both calculations appear to be so large that confident determination of even the sign of the heat transport in the North Pacific is not possible. The amount of heat transported poleward by all oceans is obtained from the Pacific Ocean calculation and transports in the Atlantic and Indian Oceans based on Bunker's surface heat fluxes.
By using a rectangular basin of uniform depth with inflow and outflow openings, the circulation in the Japan Sea is investigated numerically. Heat flux through the sea surface is determined from the annual mean atmospheric conditions for the Japan Sea, but no wind stress is considered. In the transient state, the warm water supplied through an inflow opening travels cyclonically along the coast as a density-driven boundary current in a rotating system. In the quasi-steady state, the warm water flows northward as a western boundary current which corresponds to the East Korean Warm Current and gradually separates from the coast as it flows northward. No strong boundary current corresponding to the nearshore branch of the Tsushima Current exists. Under annual mean atmospheric conditions, formation of the deep water characteristic of the Japan Sea and of the thermal front corresponding to the Polar Front do not take place.
Recent current measurements are incorporated into new heat and mass budgets for the Arctic Ocean. These budgets demonstrate the overwhelming importance of advection by the West Spitsbergen Current, which in 1971–1972 transported 16.3×109 kcal cm−2 s−1 into the Arctic Ocean. Partially on the basis of the heat budget, we suggest that there is a large heat loss to the atmosphere in the southwestern Eurasian basin, in excess of 20 kcal cm−2 yr−1. Annual mean transport in the West Spitsbergen Current could easily vary by 35% and might constitute a significant perturbation on the Arctic heat budget. The mass budgets point toward the general ineffectiveness of the Arctic Ocean in transforming subsurface water masses, and they also indicate a large shear between the ice and the upper layer of the East Greenland Current.
Ocean heat transport across 24°N in the North Pacific is estimated to be 0.76 × 1015 W northward from the 1985 transpacific hydrographic section. This northward heat transport is due half to a zonally averaged, vertical meridional circulation cell and half to a horizontal circulation cell. The vertical meridional cell is a shallow one, in which the northward Ekman transport of warm surface waters returns southward only slightly deeper and colder, all within the upper 700 m of the water column. In terms of its meridional heat transport, the horizontal circulation cell is also shallow with effectively all of its northward heat transport in the upper 700 m of the water column. Previous estimates of North Pacific heat transport at subtropical latitudes had ranged between 1.14 × 1015 W northward and 1.17 × 1015 southward. The error in this new direct estimate of Pacific heat transport is approximately 0.3 × 1015 W. In addition, it is suggested that the annual variation in poleward heat transport across 24°N in the Pacific is of order 0.2 × 1015 W, as long as the deep circulation below 1000 m exhibits little variation in water mass transport. Together, the Pacific and Atlantic transoceanic sections essentially close off the global ocean north of 24°N so that the total ocean heat transport across 24°N is estimated to be 2.0 × 1015 W northward. This ocean heat transport is larger than the northward atmospheric energy transport across 24°N of 1.7 × 1015 W. The ocean and atmospheric together transport 3.7 × 1015 W of heat across 24°N, which is in reasonable agreement with classic values of 4.0 × 1015 W derived from consideration of the Earth's radiation budget but which is markedly less than the 5.3 × 1015 W required by recent satellite radiation budget determinations.
The poleward energy transports in the atmosphere–ocean system are estimated for the annual mean and the four seasons based on satellite measurements of the net radiation balance at the top of the atmosphere, atmospheric transports of energy at the north or south poles various types of corrections had to be made, so that the global balances are maintained. This also enabled us to estimate the uncertainties in the procedures used. The uncertainties found are similar to those reported by Hastenrath based on different satellite data sets but using the same correction method. Finally, oceanic heat transports are computed for the annual mean and seasons. One of the crucial terms in the heat budget, the interseasonal storage of energy in the oceans, is estimated for three different layers 0–112, 0–552 and 0–550 m, enabling a further error estimate in the inferred oceanic heat transports. The present results confirm the presence of a strong annual cycle in the transport of energy by the oceans.
Monthly climatological estimates of wind stress and sea surface temperature are used to compute meridional Ekman heat fluxes in the world ocean. Qualitatively the annual cycles of the Atlantic and Pacific oceans are quite similar, but quantitatively, the Pacific estimates are up to several times larger than the Atlantic estimates. The Indian Ocean exhibits an annual mean southward flux over nearly all of the 24.5°N–31.5°S latitude belt, which qualitatively supports an annual mean net southward heat flux for this region as determined by surface heat balance requirements. A large southward heat flux in the Indian Ocean centered at about 7.5°N during Northern Hemisphere summer is responsible for a global Ekman heat flux distribution, with an annual cycle in the tropics that qualitatively resembles the results of Oort and Vonder Haar.
Estimates of oceanic meridional energy transport for the Pacific, Atlantic and Indian oceans are made using calculated annual mean surface energy fluxes. The surface flux calculations use bulk formulations based on 31-year (1949–79) surface marine observations from the Consolidated Data Set of the U.S. Navy Fleet Numerical Oceanography Center. The results show a northward transport at all latitudes in the Atlantic Ocean, a poleward transport in both hemispheres centered around 10°S in the Pacific Ocean with a northward transport south of 25°S, and a southward transport at all latitudes in the Indian Ocean. When all oceans are taken together, the transport is poleward centered around 5°S with a maximum northward transport at 25°N and a maximum southward transport at 15°S. A sensitivity experiment carried out on the possible errors associated with the surface energy flux shows that the transport in the Pacific Ocean—both in direction and in magnitude—is most uncertain. The transport directions of t...
A current meter mooring maintained for 2 years at 35°N, 152°E as part of the ``Wespac'' array (Schmitz et al., 1987) sampled the Kuroshio Extension during the last 390 days of deployment. Based on a method developed by Hall and Bryden (1985) for the Gulf Stream ast 68°W, velocity and temperature time series at five depths from 350 to 6000 m have been used to construct the representative horizontal velocity structure of the current. Cross-section position is quantified in terms of temperature at 350 dbar, and an alongstream flow direction is defined on the basis of the current shear. Integrated transport of the Kuroshio is 87+/-21 Sv, compared to 94+/-26 Sv for the Gulf Stream at 68°W based on analogous data sets and analysis techniques. The two currents are compared in terms of horizontal and veritical and distributions of transport as well as relative strength of barotropic and baroclinic components. The Kuroshio has a larger relative barotropic component than the Gulf Stream, while the latter has a stronger, deeper thermocline expression. The effects of rotating velocity measurements to a continuously changing alongstream direction are examined for both currents; the major impact is to cut the apparent velocity variance in half, demonstrating that much of the variance is due to the meandering of a quasi-fixed velocity profile rather than the passage of eddies. Results from these somewhat unconventional analysis methods are placed in broader geographical context and compare favorably with more traditional methods of determining these currents' transports.