Ocean Data Portal: a standards approach to data access and dissemination

Ocean Data Portal: A Standards Approach to Data Access and Dissemination

View paper here

Return to overview »

5 open review comments to “Ocean Data Portal: a standards approach to data access and dissemination”

  1. Dear Authors,
    The Ocean Data Portal (ODP) is an excellent overall entry point to oceanographic data in general. A mention of the Global Observing Systems Information Center (GOSIC) at http://gosic.org in the whitepaper would be something to possibly consider in this whitepaper as the GOSIC portal complements the ODP by providing a specific and unique focus on access to observing datasets, metadata, and information (including GOOS data), with a particular emphasis on ocean climate data. A mention of the possible cooperation and coordination between the ODP and the GOSIC would go a long way towards enhancing the ability of the global ocean observing user community to get access to critical datasets and related information. Furthermore, the GOSIC Portal is a registered component of the Global Earth Observing System of Systems (GEOSS) and opens access to a wider audience of people who are interested in ocean data. Finally, the GOSIC is endorsed by GOOS as a data access point for GOOS data.

    Thanks for your consideration here.

    Regards.

    Howard Diamond
    U.S. Global Climate Observing System (GCOS) Program Manager
    NOAA/National Climatic Data Center

  2. Steve Hankin says:

    Dear Authors,

    As I’m sure you know the IT white papers submitted for OceanObs 09 all share the difficult challenge of describing highly technical material in non-technical language. Too much detail and the paper becomes unintelligible to the science audience that the paper wants to reach. Too much simplification and there is a risk of giving a false sense of simplicity about complex problems. In my opinion this paper, “Ocean Data Portal: a standards approach to data access and dissemination”, strays too far towards simplification. I would like to see it directly address some complex realities of data management that it currently side-steps. Examples of this are:

    * The notion of a single portal for access to widely distributed data has proven to be problematic for what are essentially “social” reasons. Namely that the visibility of participating organizations is often perceived to be reduced when its data are accessed through a portal that is designed elsewhere. The organization running the portal is seen as “taking credit”. For this reason much of the focus of discussion in recent years has been on service-level interfaces, rather than portals on the theory that defining enabling technologies will permit each participating organization can create its own “portal”. Can the paper more directly address this question? Is the intention that copies of the Ocean Data Portal will be hosted at each NODC?

    * The standards literature is full of analyses to explain what are the elements of Information Technology that make achieving standardization so difficult. The best examples of success are found to be in the IETF and the W3C, where great emphasis is placed on mature , competing implementations of technologies, before they can become “standards”. The design of the portal described in this paper seems to depend upon a standards process that is described in just a single sentence, “IODE, together with JCOMM, has initiated a standards process that will support the accreditation and adoption of core standards by the marine meteorological and oceanographic communities”. Does the OceanObs conference include a paper on this process that can be cited? If not, then can more detail on the IODE/JCOMM standards process be included in this paper.

    * The paper refers to “a number of internationally accepted standards which are applicable to oceanographic data [including] those developed by the International Organization for Standardization (ISO), the World Wide Web Consortium (W3C) and the Open Geospatial Consortium (OGC)”. It is implied that the OGC and ISO processes have produced mature standards that are ready to be applied to oceanography. The simple statement hides the complex reality that the the standard tend to be at such high levels of abstraction that another layer of standards is needed in order to apply them to oceanography. The case of OGC is particularly acute, where 1) the standards cannot provide interoperability until complex and subtle XML application schemas are developed and agreed upon across the ocean community and 2) multiple OGC standards such as WFS, WCS and SOS compete to perform the same task. In this predicament we are seeing some pockets of convergence and other major areas of divergence. The portal is apparently based upon the use of the WFS standard. Can you include in the paper a discussion of the particular application schema that is used, and the breadth of acceptance that the application schema enjoys?

    * Along the lines of the previous remark, the paper mentions that the “metadata record is based on ISO 19115″. Can you cite the particular profiles of ISO 19115 that are being used?

    * What are the “”E2E” conventions for netCDF” and where can they be found? Are they connected to the Climate and Forecast (CF) conventions? Would it make sense to list them at http://www.unidata.ucar.edu/software/netcdf/conventions.html as a part of this portal development effort?

    * Since the paper shows attractive and tantalizing screen snap shots of user interfaces, providing a URL at which the reader can experience the portal would add to the paper. If a URL is unavailable for reasons of security or BETA status, then can the paper suggest when it might become available and where the reader should look for it?

    Sergey Belov Reply:

    Dear Steve,
    Thank you for your comments. I’ll try to evaluate all your comments and forward them to appropriate persons (including me as one of the authors). Following your question related to the “”E2E” conventions for netCDF”, here is the reply from ODP team member who is dealing with this issue (with quite a lot of details):
    “NetCDF - E2E convention based on the NetCDF Classic Model (NetCDF-3) and CF conventions. E2E NetCDF transport data file uses simple structures as dimensions, variables and attributes for representing data arrays. NetCDF – E2E convention is concerning with naming rules inside NetCDF transport data file:
    • Each dimension and variable has system parameter name with thematic federation prefix. E.g.: odp:SDN_ALONZZ01, where “odp” is thematic prefix of ODP federation and SDN_ALONZZ01 is system parameter name for longitude. Dimension name has the same name as variable name to which it corresponds.
    • Attribute with name “resourceId” contains information resource identifier to make correspondence between current transport data file and information resource.
    • 2-dimensional variable is using for representing data values of String type. First dimension has byte data type, second dimension has int data type and describes maximum length of strings in the array.
    • Profile data measurements are represented the same way as plain surface measurements. E2E convention uses special variable with name “series” for defining array of start and stop positions which are used for retrieving the data for each profile.
    • E2E NetCDF transport data file can use attribute names for variables from CF convention, describing parameter name, units of measurement and fill value for missing value. This functionality can be used in case E2E NetCDF transport data file contains parameters which defines in parameter list of CF convention.”

    Sergey Belov Reply:

    The URL of the Portal interface, which screen-shots are provided in the article, is http://www.oceandataportal.net and it’s available for any user without registration (for a while).
    Answering your question on what particular ISO 19915 standard metadata re “metadata record is based on ISO 19115″

    Sergey Belov Reply:

    The URL of the Portal interface, which screen-shots are provided in the article, is http://www.oceandataportal.net and it’s available for any user without registration (for a while).
    Answering your question on what particular ISO 19915 standard metadata record is based, I can say that it’s based on “ISO 19115:2003 Geographic information — Metadata”.