Read more
High Performance Computing in the Geosciences surveys the state of the art of programs presently being developed which require high performance computing for their implementation, provides a guide for decision making in regard to computing directions in future numerical models, and provides an overview of future developments in massively parallel processing and their implications for numerical modelling in the geosciences.
List of contents
Performance and Portability in Parallel Computing: a Weather Forecast View.- The New Canadian Variable-Resolution Model: a Universal Tool to Study Atmospheric Phenomena on a Broad Temporal and Spatial Scale.- Experiences with Parallel Programming in Meteorology.- Porting the Quasi-Nonhydrostatic Meteorological Model to the Kendall Square Research KSR1.- Automatic Differentiation, Tangent Linear Models, and (Pseudo-) Adjoints.- Iterative Domain Decomposition Algorithms: Theory and Applications.- Parallel Computing and Operational Sea-Ice Forecasting.- Coupled Ocean-Atmosphere Simulations.- First Considerations about Modelling the Ocean General Circulation on MIMD Machines by Domain Decomposition Method.- Predicibility of the Oceanic Circulations.- Application of Fifth-Order Compact Upwind Differencing to Moisture Transport Equation in AGCM.- A Parallel Algorithm for Simulation of Long Range Transport of Air Pollution.- Multidisciplinary Mathematical Model for Earthquake Prediction Studies.- Basic Functional Representation of Programs for Automatic Differentiation in the Odyssée System.- Translation Tools for High Performance Computing.- Global Atmospheric and Ocean Modeling on the Connection Machine.- Efficiency of a Number of Processors in Integrating a Global Weather Model.
Summary
Proceedings of the NATO Advanced Research Workshop, Centre de Physique, Les Houches, France, 21--25 June 1993
