NETCDF - AN INTERFACE FOR SCIENTIFIC-DATA ACCESS

In scientific visualization, two obstacles prevent the full use of heterogeneous networks of powerful workstations: data access and data representation. Low-level sequentialdata access ties visualization applications to application- specific formats. Machine-dependent data representations make it difficult to distribute applications across networks or to display output from programs running on different workstation architectures simultaneously. The Network Common Data Form is a data abstraction for storing and retrieving multidimensional data. MetCDF is distributed as a software library that provides a concrete implementation of that abstraction. The implementation provides a machine-independent format for representing scientific data. Thus NetCDF is more than just another data format. Together, the abstraction, library, and data format support the creation, access, and sharing of scientific information. NetCDF has proven useful for supporting objects that contain dissimilar kinds of data in a heterogeneous network environment and for writing application software that does not depend on application-specific formats. NetCDF data achieve independence from applicationspecific formats because the data are self-describing. This means that a NetCDF data set includes information defining the data it contains. NetCDF data achieve independence from particular machine representations by using a nonproprietary standard for external data representation. Programs that use the NetCDF interface can thus read and write data in a form that is more widely usable than machinedependent binary data files. NetCDF files can provide a way to encapsulate structured scientific data for use among multiple application programs, and thus, these files can help to support high-level data access and shell-level application programming. © 1990 IEEE