The NEMO code utilises the netCDF (network Common Data Form) library for inputting and outputting its data files which means before compilation can begin the netCDF library must be available or compiled from source. Further information on netCDF can be found at [3,4].
NEMO can potentially be run on any number of processors providing sufficient memory is available to fit the required subset of data onto a single processor. The number of processors must be specified at compile time as the dimensions of a number of statically allocated arrays are computed based on the processor count. This hard-wiring of the processor count into code means that the code must be recompiled if a different number of processors is used.
The NEMO Makefile is somewhat complex. The internal macro names (those pre- defined in make) such as FC, FFLAGS, LDFLAGS, etc are not used. Instead, the code authors use their own naming convention and as no comments have been included to describe these the user is left to make their own decision as to what each macro name stands for. Table 1 lists the NEMO Makefile macro names and what they are believed to correspond to using the standard Makefile macro naming conventions.