Re: netCDF vs HDF

• To: "Terry Westley" <twestley@xxxxxxxxxxxxxxxxxxxx>
• Subject: Re: netCDF vs HDF
• From: jps@xxxxxxxx (John Sheldon)
• Date: Fri, 3 Sep 1999 23:02:20 -0400

Hi Terry-

We (GFDL) chose netCDF over HDF about 5 years ago for several reasons:

 1. after trying both netCDF and HDF, netCDF seemed easier to use,
     especially for more casual users, and was better suited to our
     datasets, almost all of which are multidimensional gridded
     blocks;

 2. there was much more software available for manipulating and
     visualizing netCDF data, including NCAR Graphics which was
     our primary graphics application at the time;

 3. the development of applications for handling/viewing HDF data 
     seemed to stall (at least at NCSA), so we worried about future
     support for those tools;

Scott Pendergrass' observation that "HDF seems directed toward image
type data" is our impression also, and I think this is reflected in the
types of organizations that have adopted HDF (ie, primarily satellite
imagery related missions).

If you are interested in the rationale used by the NCAR Climate System
Model (CSM) group for adopting netCDF, check out:
   http://www.cgd.ucar.edu/ccr/bettge/CSM-netCDF/csm_why_netcdf.html

There is also a pretty good discussion of the characteristics of
netCDF and HDF in a FAQ about CDF put together by David Han at NASA
Goddard Space Flight Center (see his "Question #8"):
   http://nssdc.gsfc.nasa.gov/cdf/html/FAQ.html

You may want to take a look at the netCDF section (or maybe all) of 
our Scientific Visualization Guide:
   http://www.gfdl.gov/~jps/GFDL_Vis_Guide.html
This guide assumes that your data is in netCDF format; GFDL's adoption
of netCDF was a precondition I set when I took on the job of organizing
and supporting our data management and visualization capabilities, for 
the obvious reason that I couldn't possibly accommodate everyone's
individual data file format.

Last, but by no means least, the folks at Unidata have always been
extremely responsive and helpful whenever I've had questions or problems.
They helped us get a version of netCDF ported to what was at that time
the very unique Cray T3E architecture, and they even wrote us a utility
on-the-spot to help recover data in an incompletely written file.
(See http://www.gfdl.gov/~jps/GFDL_VG_NetCDF_Utils.html#nrecreset)

Anyhow, that's *my* $0.02...
Hope it helps!

John Sheldon 
Technical Services Manager
----------------------------------------------+---------------------------
 NOAA/Geophysical Fluid Dynamics Laboratory   | Email: jps@xxxxxxxx
 Princeton University/Forrestal Campus/Rte. 1 | Tel:   (609) 987-5053
 P.O. Box 308                                 | Fax:   (609) 987-5063
 Princeton, NJ, USA  08542                    | WWW:   http://www.gfdl.gov
----------------------------------------------+---------------------------

• References:
  • netCDF vs HDF
    • From: Terry Westley