Dr. Peggion presented a talk on the feasibility of a two-way nesting algorithm in a moderately parallel environment. This is a study that requires modeling the ocean (mainly the Gulf of Mexico). The main problem that this effort deals with is artificial boundaries. This model is used to study various conditions such as Loop Currents and El Nino effects.

Dr. Peggion first described and presented displays of the Gulf of Mexico to demonstrate the intrusion of the Loop Current over the Gulf of Mexico and the Mississippi Bight. Since this presentation was targeted to the numerical aspects of the study as opposed to the oceanography aspects, she presented several ocean modeling numeral issues such as:

• Explicit schemes are more accurate than implicit in representing wave propagation.
• The use of the N-S equation
• Time splitting

With the assistance of the displays, she was able to explain the problem with artificial boundaries. She then described in detail the two models used in this study:

• The course-grid resolution model (used for the Gulf of Mexico)
• The fine-grid resolution model (used for the Mississippi Bight)

Dr. Peggion next describe the parallel code used in this study:

• Uses C-preprocess
• Runs as serial or parallel
• Uses MPI or PVM
• Produces separate plot files
• Runs on: SGI Power Challenger, Origin 200, and Cray Y-MP

In conclusion, Dr. Peggion again pointed out that the major challenge of this study is the artificial boundary conditions. This effort is low cost, but a big model.