TOMS Description

Historical Background

At Florida State University (FSU), James O'Brien and his students and researchers have been working with layered models for many years. These include Harley Hurlburt, Dana Thompson, John Kindle, George Heburn and Tony Busalacchi, who were among the first to use and develop these models. TOMS follows their tradition.

Mark Luther programmed a non-linear reduced gravity model (i.e. a single active layer over an infinite abyss) in spherical coordinates for the Arabian Sea (Luther and O'Brien, 1985). It was programmed for the CDC Cyber 205 and a later Indian Ocean version (north of 25S) had a horizontal resolution of 1/5 deg (Woodberry et al., 1989). Tommy Jensen generalized the model to a multi-layer model, including entrainment, velocity based friction and normal-mode open boundary conditions. It was applied to the Indian Ocean with 3 active layers (3.5 layer model), and further optimized for the ETA-10 computer (Jensen 1990; 1991; 1993).

When ETA/CDC closed their business, the code was ported to the CRAY-YMP by Jorge Capella. He later added some enhancements and made a new 1/6 deg geometry to 30S (Barnier et al., 1994). This code is the FSU Indian Ocean Model.

It was brought to he University of South Florida, where Mark Luther and his group have used it to give nowcasts of currents in the Indian Ocean, and Jorge Capella brought it to the University to Puerto Rico and have set up a model for the Atlantic Ocean.

Tommy Jensen brought it to Colorado State University and started development of a new thermodynamic version. The result is TOMS. It is a new code, but is based on the hydrodynamical FSU model and many of the ideas used there, for instance, Jorge's segmentation has not been changed. At IPRC it is being further developed and has been used for studies of the Indian and the Pacific Oceans.

References

• Barnier, B., J. Capella and J. O'Brien, 1994: The use of satellite scatterometer winds to drive a primitive equation model
  of the Indian Ocean: The impact of bandlike sampling. Journal of the Geophysical Research, 99 , 14,187-14,196.

• Jensen, T. G. 1990: A numerical study of the seasonal variability of the Somali Current.
  Ph.D. dissertation, Fla. State University, Tallahassee, 118 pp.

• Jensen, T. G. 1991: Modeling the Seasonal Undercurrents in the Somali Current System.
  Journal of the Geophysical Research, 96 , 22,151-22,167.

• Jensen, T. G. 1993: Equatorial Variability and Resonance in a Wind-Driven Indian Ocean Model.
  Journal of the Geophysical Research, 98 , 22,533-22,552.

• Jensen, T. G. 1996:
  Artificial Retardation of Barotropic Waves in Layered Ocean Models.
  Monthly Weather Review, 124 , 1272-1283.

• Jensen, T. G. 1998:
  Open Boundary Conditions in Stratified Ocean Models.
  Journal of Marine Systems, 16 , 297-322.

• Jensen, T. G. 1998: Description of a Thermodynamic Ocean Modelling System (TOMS).
  Atmospheric Science Paper , 670 , Colorado State University, Fort Collins, 50 pp.

• Jensen, T. G. 2001:
  Application of the GWR method to the Tropical Indian Ocean.
  Monthly Weather Review, 129 , 470-485.

• Jensen, T. G. 2003:
  Barotropic mode errors in an Indian Ocean model associated with the GWR method.
  Global and Planetary Change, 37, 1-18

• Jensen, T. G. 2003:
  Cross-equatorial pathways of salt and tracers from the northern Indian Ocean: Modelling results.
  Deep-Sea Research II, 50 , 2111-2128

• Luther, M. E. and J. J. O'Brien, 1985: A model of the seasonal circulation in the Arabian Sea forced by observed winds.
  Progress in Oceanography, 14, 353-385.

• Woodberry, K. E., M. E. Luther, and J. J. O'Brien, 1989: The wind-driven seasonal circulation in the southern tropical Indian Ocean.
  Journal of the Geophysical Research, 94 , 17,985-18,002.

Basic Model Features

• Prognostic volume transport and layer depth
• Prognostic T, S and any number of tracers
• Diagnostic equation for free surface
• Finite depth or reduced gravity mode
• Material layers
• Forced by wind stress, surface heat flux and net precipitation
  and/or restoring to observations

Model Physics

• General bulk mixed layer model
• Convection by partial or full overturning
• Biharmonic and/or spatial varying harmonic horizontal friction
• Flow dependent harmonic horizontal friction (Smagorinsky)
• 3-D varying linear and non-linear friction
• Harmonic horizontal diffusion
• Ri dependent vertical friction/diffusion or explicit entrainment

Geometry and Numerics

• Arbitrary Lagrangian Eulerian vertical coordinate
• Box model or irregular land boundaries with islands
• Periodic or open boundary conditions for U, V, H, T, S + tracers
• No slip, partial slip or free slip coastal boundaries
• Leapfrog scheme for momentum and optionally for advection
• Optional Takacs, Arakawa-Hsu or 3. order upstream scheme for advection
• Gravity wave retardation for barotropic mode
• Asselin filter, forward scheme or time step averaging for filtering

Code

• Modular: 104 subroutines and 48 include files
  selected by C-preprocessing
• Fully vectorized, but highly portable (still using f77)
• Runs in parallel on Shared Memory Processors:
  CRAY C90, CRAY SV1ex; SGI Origin, SGI Altix
• Has been run on workstations:
  IBM SP2, Sun Ultra 1,Sun Ultra30, SGI Octane, SGI Indigos
  HP 755, DEC Alpha, Intel (Linux)
• Fast:
  SGI Origin 3000/400Mhz, 24 CPUs: over 2.3 Gflops
  SGI Origin 2000/250Mhz, 24 CPUs: over 1.2 Gflops
  SGI Origin 2000/250Mhz, 16 CPUs: over 900 Mflops
  CRAY-SV1, 8 CPU: 617 Mflops
  CRAY-C90, 1 CPU: > 400 Mflops
  SGI Power Challenge, 4 CPUs: about 200 Mflops
  Workstations, 1 CPU: up to 70 Mflops

Home

Tommy G. Jensen      
tjensen@hawaii.edu

Last update: March 24, 2005