Open AccessDD-OceanVar: A Domain Decomposition Fully Parallel Data Assimilation Software for the Mediterranean Forecasting System
Author(s) -
Luisa D’Amore,
Rossella Arcucci,
Luisa Carracciuolo,
A. Murli
Publication year - 2013
Publication title -
procedia computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.334
H-Index - 76
ISSN - 1877-0509
DOI - 10.1016/j.procs.2013.05.290
Subject(s) - computer science , data assimilation , scalability , software , domain decomposition methods , domain (mathematical analysis) , parallelism (grammar) , computational science , mediterranean sea , parallel computing , data mining , mediterranean climate , database , meteorology , programming language , mathematical analysis , mathematics , physics , finite element method , thermodynamics , ecology , biology
OceanVar is a Data Assimilation (DA) software which is being used in Italy within the Mediterranean Forecasting System\ud(MFS) to combine observational data (Sea level anomaly, sea-surface temperatures, etc.) with backgrounds produced by\udcomputational models of ocean currents for the Mediterranean Sea (namely, the NEMO framework). OceanVAR is based\udon a three-dimensional variational approach. We describe computational efforts aimed to design a fully parallel OceanVar\udsoftware, based on Domain Decomposition (DD), which involves modification of the variational scheme on each sub domain.\udOur approach aims to face to the ever greater multi-level parallelism and scalability of the current and of the next generation\udof leadership computing facility systems, while fulfilling the specific requirements of OceanVar within the MFS
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom