U22A-01 INVITED 10:30h
Ocean Eddy Resolving Simulation On The Earth Simulator
A high physical and computational performance ocean circulation model is being developed. This model is intended for use with studying mesoscale eddy physics of the world ocean on the Earth Simulator. In order to resolve the eddies of the world ocean, high computational performance is required. We have, therefore, decided to employ a quasi-uniform cubic grid system, by which long time-step-width can be utilized for high resolution calculations compared with the traditional longitude-latitude grid system. Arakawa-B gird method is employed for horizontal spatial descritization and the $ \sigma \hspace{-1mm} - \hspace{-1mm} z$ hybrid coordinate is used for the vertical one. A split-explicit method is used for the time integration method. The development of the dynamical core of the model has been completed and calculations for several test cases are now carried out. In parallel with this development, high resolution simulations using a traditional ocean model is carried out to investigate physics of mesoscale eddies in a limited area. The model used for this study is the Center for Climate System Research (CCSR) Ocean Component model (COCO). The simulation area is the Southern Ocean from $20^\circ S $ to $75^\circ S$. The model is forced by monthly climatology of the surface wind stress, the sea surface temperature, and the sea surface salinity. At the north boundary, the temperature and salinity are restored to the annual-mean climatology of Levitus. In order to study the sensitivity of the resolution to the physics of eddies, experiments are performed for four different horizontal resolutions; $1/2^\circ (in \ longitude ) \times 1/3^\circ (in \ latitude), 3/8^\circ \times 1/4^\circ, 1/4^\circ \times 1/6^\circ, 1/8^\circ \times 1/12^\circ $. There are 85 levels in the vertical with grid spacing of 50 m at the surface and 200 m near the bottom. The simulation results show that the root-mean-square value of the sea surface height variability approaches to the observation one as increasing the resolution. The volume transport of the Drake Passage decreases as increasing the resolution and approaches to the observation value.
U22A-02 INVITED 11:00h
An Eddy-Resolving Ocean Simulation Driven by 1950-2003 Atmospheric Forcing - OFES (OGCM for the Earth Simulator) Project -
A series of challenging eddy-resolving ocean simulations on the global domain have been performed by using OFES (OGCM for the Earth Simulator) based on GFDL MOM3. Our first attempt was a 50 years spin-up simulation driven by climatological forcing, which turned out to be quite satisfactory in capturing a number of realistic features of the world oceans both in the mean state and eddy activity statistics. The successful spin-up encouraged us to move on to the second phase, in which a hindcast experiment covering the period 1950-2003 was attempted to study the variety of intriguing interannual and decadal oceanic phenomena that have occurred during the second half of 20th century. In our simulations, the employed computational domain is near-global region extending from 75S to 75N with the horizontal grid spacing of 0.1 degree and 54 vertical levels. The hindcast run starts from the oceanic field simulated by our 50 years spin-up. Difference in the computational setting of the two runs lies in the form of surface forcing. In the hindcast run, the surface fluxes are specified from daily mean NCEP/NCAR re-analyses data with an additional surface salinity restoring to monthly climatological field, while the spin-up run was forced by monthly mean climatological NCEP/NCAR re-analysis data. Comparison of the preliminary spin-up and the hindcast simulation outcomes clearly shows that meso-scale eddy activities in the latter simulation become more realistic than those in the former. In particular, we notice that the global distribution of sea surface height variability and the frequency of eddy shedding at the loop current are much improved by the realistic high-frequency forcing. The degree of the coincidence of the timing and amplitude of Nino 3 SST index is more than expected and encourages us to compare carefully the simulated big events in the 1990s with the observations. Dipole mode events in the Indian Ocean are also simulated. Our model results show that the subsurface variability and wave propagation in the Indian Ocean play the important role in the evolution and the phase reversal of the dipole mode events. We also looked into the SST variability associated with the Pacific and Pan-Atlantic Decadal Oscillations and found that they are simulated well, namely, the spatial patterns of the decadal SST differences are quite similar to the observation and detailed analyses are going on now. In the presentation, we will report the outline of our hindcast simulation and the present status of our data analyses.
http://www.es.jamstec.go.jp/esc/research/AtmOcn/ofes/ofes.en.html
U22A-03 11:30h
Depth Variability of Indonesian Throughflow Inflow and Outflow Transport From OfES
The ocean model for the earth simulator (OfES) is a global model of the ocean at approximately 11 km resolution in the horizontal and 54 levels in the vertical. With OfES, a more detailed picture of the ocean pathways through the Indonesian seas can now be made. This study is aimed at understanding the partitioning of transport through the various straits in the Indonesian seas. Volume transport was computed at several inflow and outflow straits from the depth integral of velocity from a OfES climatologically forced run. The three main outflow straits, Lombok, Ombai and Timor, show surface enhanced transport as well as a subsurface maximum. Reverse flow (into the Banda Sea) is observed at depth at the Ombai and Timor sections. This reverse flow, observed at some of the outflow straits, provides Indian Ocean water to the Banda Sea. On the inflow side, flow between the Philippines and Sulawesi was divided into three sections; the northern-most section shows westward transport in the surface layer (to about 500 m), while the two other sections show eastward transport. Therefore, in the OfES, the Mindanao Current enters the Celebes Sea just south of the Philippines, circulates within the sea, and a portion exits back to the Pacific north of Sulawesi. The flow at depth is also opposite in the northern and southern sections. South Pacific water appears to enter the Banda Sea from the east and west sides of Halmahera. Output from the OfES is in general agreement with recent observations in the Maluku and Halmahera Seas, and shows substantial flow at depth. The exact pathway of this deeper flow is still unknown. Generally speaking, the OfES has extrema in transport in winter and summer, when the monsoons are strong. During northern summer the net transport is maximum towards the Indian Ocean. Flow is strong from the Mindanao Current around the south of Mindanao into the Celebes Sea, supplying North Pacific water to the Indonesian seas. Flow is also maximum southward between Celebes and Halmahera, and this water comes from the New Guinea Coastal Current and brings South Pacific water into the region. In the OfES the N. Pacific water continues through the Celebes Sea into the Makassar Strait, which also has peak southward flow during the northern summer. The S. Pacific water passes through the Lifamatola Passage into the Banda Sea. Flow exits the southern Indonesian seas through a variety of gaps in the Indonesian archipelago. In the OfES, some flow also exits through the Karimata Strait between Borneo and Malaysia. Like the inflow, flow is maximum southward through the southern straits in June through August, during the southeast monsoon. Some of the flow from Makassar passes into the Java Sea and north into the Karimata Strait. Flow through Karimata is perhaps more important during the northern winter months when the southward flow is reduced, and in some cases, reversed. In OfES, flow is to the south in Karimata Strait during January. This could represent a significant fresh water source to the ITF outflow, since precipitation is large throughout the Java Sea.
U22A-04 11:45h
Pathway and formation of Antarctic Bottom Water using simulated CFCs distribution -OFES (OGCM for the Earth Simulator) Project-
We have investigated the pathway of Antarctic Bottom Water (AABW) using simulated chlorofluorocarbons (CFCs) distribution in a global eddy-resolving OGCM. Our goal is to improve our understanding of the processes and pathways determining the distribution of CFCs in the Southern Ocean, where much of this tracer is entrained by formation of the deep and bottom water. The model exhibits considerable skill in reproducing the section along the Greenwich Meridian (AJAX) and 170W (WOCE P15S). In the South Atlantic, the model indicates two pathways transporting CFC-11 from the Weddell Sea to the AJAX section: an upper pathway through the Scotia Sea, and a lower pathway south along the South Scotia Ridge. The simulated CFC-11 inventory along the two sections is much like that observed.