Enhancing the Grid Analysis & Display System (GrADS) to read HDF-EOS data

NASA's Earth Observing System Data and Information System (EOSDIS) project is producing many data of interest to earth scientists. EOSDIS, begun in 1997, manages data from NASA's past and current Earth science research satellites and field measurement programs, providing data archiving, distribution, and information management services. There is a need for easy-to-use tools for viewing and analyzing these data from the EOS project. Many of these data are in the Hierarchical Data Format (HDF). HDF is a data format and related access and creation Application Programmer Interface (API) library from the National Center for Supercomputing Applications (NCSA) (http://hdf.ncsa.uiuc.edu/). The EOSDIS HDF files use a metadata convention specific to the project (http://spsosun.gsfc.nasa.gov/New_EOSDIS.html).

Previously, the Grid Analysis & Display System (GrADS) (from the Center for Ocean, Atmosphere, and Land studies (COLA) (http://grads.iges.org/grads/) has been extended to read netCDF (http://www.unidata.ucar.edu/packages/netcdf) data that conforms to the COARDS (http://ferret.wrc.noaa.gov/noaa_coop/coop_cdf_profile.html) metadata convention (Hooper, 12th IIPS, 1996). NetCDF is a data format and associated API from the Unidata project at the National Center for Atmospheric Research (NCAR). Because of work by the developers of netCDF and HDF, the enhancements to read netCDF data can also read HDF's Scientific Data Sets (HDF-SDS)). This was later extended to apply to almost any netCDF/HDF-SDS data by reading an external description file. This is based on existing mechanism in GrADS called a Data Descriptor File (DDF) (Hooper, 14th IIPS, 1998).

This paper reports on enhancing GrADS to read HDF-EOS data. HDF-EOS data can include gridded data, as well as "image" or "swath" data. Since GrADS already works well with gridded data, improving it to read HDF-EOS gridded data is a relatively straight-forward enhancement. However, since the EOSDIS metadata convention is radically different from the COARDS one, both in form and intent, new techniques must be developed to read the data. Additionally, many of the EOS data are in "image" format, having their origin in remote sensing (satellite) instruments. This is a new data type for GrADS to deal with. The implication here is that farther-reaching changes to the GrADS software are required. This paper discusses the portions of the work dealing with reading the HDF-EOS data format (including reading the image/swath data). Adding the "image" data type to the GrADS data model is the work of fellow GrADS developers (notably Brian Doty, the primary GrADS developer).