Stony Brook Parallel Ocean Model

System requirements

The sbPOM code is intended to be highly portable among unix/linux operating systems. The sbPOM runs on workstations, compute-servers, clusters and massively parallel supercomputers. To run sbPOM, you will need:

  • A fortran compiler: The sbPOM is predominantly compiled on several linux/x86 platforms with several versions of the Intel Fortran Compiler for Linux, which is free for individual scientific use. It has also been successfully compiles with other fortran compilers (IBM XL Fortran Compiler, Portland Group Fortran Compiler, GNU Fortran Compiler...).
  • A distributed-memory parallelism: The communications between processors to exchange information are based on the Message Passing Interface (MPI) standard interface. The MPI is based on MPICH2 version, which is free of charge. If your system uses an older version of MPI, it won't be compatible.
  • Unidata's NetCDF: The original POM used NetCDF for I/O operations, which makes the files space-efficient, self-describing and machine independent. NetCDF is also recognized by many graphics and post-processing utilities. To maintain those advantages, the sbPOM uses Parallel-NetCDF which provides high-performance parallel I/O while still maintains file-format compatibility with Unidata's NetCDF. Both NetCDF library and Parallel-NetCDF are free and must exist on your system.
  • The GNU make utility: sbPOM uses make to build and install the program. Make controls the generation of the sbPOM executable from the code's source files. Make gets its knowledge of how to build the sbPOM from a file called makefile.
  • Matlab©: The sbPOM uses scripts written for Matlab© to generate the input files and to visualize/process the model output. Although Matlab© R2008B has native NetCDF support, you still need the mexnc and snctools toolboxes (appropriate for your version of Matlab©). Other matlab toolboxes such as the M-map package for map projection are also required.

    Antoni JordiIMEDEA (UIB-CSIC)