Unidata - To provide the data services, tools, and cyberinfrastructure leadership that advance Earth system science, enhance educational opportunities, and broaden participation. Unidata
         
  advanced  
 

Community Standards for Unstructured Grids Workshop Abstracts

Community Standards for Unstructured Grids home page


Tom Gross - Overview of Workshop Goals and Unstructured Metadata Standards in Use Today:  Straw Man Example

This workshop is convened to discuss the development of standards for output products of irregular grid hydrodynamic models.  These models are clearly a very important development in ocean science and are already in use by environmental managers and emergency response and planning management.  The acceptance and support of these techniques will depend upon a usable interface between the modeling community and the 3rd party users of the model results.  Many systems exist to work with simple, logically rectangular gridded models.  The development of systems to work with triangulated meshes will require a community effort to provide a workable, reproducible, community supported irregular grid description.  We believe that this is technically possible, and probably not very difficult.  But it does require community buy-in for standards to be adopted and actually used.   Community standards can provide non-altruistic benefits to model developers.  If well designed then these descriptions can be the foundation of modeling tools which can be shared and will help the pace of development of irregular grid modeling methods.

We have a small collection of grid specifications for a few of the most popular irregular grid hydrodynamic models.  A straw man “CDL” of these files will be posted and discussed.  The possible uses of the data model will be described.  The straw man is meant to be a starting point for the breakout session discussions of the details, exceptions and solutions for describing the greatest super-set of irregular griding methods.

This workshop will discuss the needs of the modelers and the 3rd party users of the models.  Based on these needs we will build a description of an irregular grid data structure.  Tangible examples will be devised for several existing modeling systems.  Finally, we expect to propose to the CF Conventions for NetCDF a description of irregular meshes to be used by NetCDF output files.  Other technologies will also be addressed, such as HDF5, Java API’s etc.  

Steve Hankin - Metadata Standards in the IOOS Context

This workshop is sponsored through funding to develop a US Integrated Ocean Observing System (IOOS).  This fact illustrates the degree to which coastal ocean modeling is a maturing field.  Coastal ocean models are increasingly viewed as a primary tool for addressing day-to-day societal needs in areas such as mitigation of natural disasters; ecosystem and coastal resource management; public health; search and rescue; climate change impacts; navigation and homeland security.  Addressing these “operational” needs today and in the near future requires community-wide adoption of standards for model output (among others) and an assurance that those standards will remain adequately stable.

The environment, in which modelers work, however, continues to evolve rapidly in many areas: numerical techniques, understanding of scientific processes, and the state of the art in information technology.  The development of successful standards in areas of rapid technological advancement inevitably requires compromises that balance practical needs today, against anticipated needs in the future.  This session will explore these issues and suggest some guidelines for the development of a successful standard.


John Caron and Russ Rew - Important Aspects of the CF Conventions and Unidata Common Data Model    

The CF Conventions for netCDF metadata have proven extremely useful for describing climate and forecast model datasets stored in netCDF files. The Unidata Common Data Model (CDM) is an abstract data model for scientific datasets, combining the netCDF, HDF5, and OPeNDAP data models, and adds a coordinate system layer corresponding to that of CF. This talk will briefly summarize the coordinate system models of CF and the CDM, as background to possible extensions of CF for unstructured grids. We will also summarize the status and future developments of netCDF.


V. Balaji - Grids, Meshes and Mosaics

A grid-spec for irregular grids will need to fit into the larger context of other grid specifications in use by atmospheric climate models and the other world of regular gridded hydrodynamic models.  A hierarchical grid spec is presented which may be used to contain complex grids which include staggered grids, sub-meshes, mosaics and other formulations necessary to efficient computing on cluster super computers. http://www.unidata.ucar.edu/events/2006GridWorkshop/balaji.pdf


Robert Wallace - XMDF Data Specification and API in Support of Structured and Unstructured Data

The XMDF API is probably the most complete API for Geophysical Fluid Dynamics. XMDF has a good users manual, both C and Fortran API, sits on top of HDF5. Uses include watershed, river, estuary and ocean models, and is evolving with HPC and GRID computing, OpenMI, ESMF and more. XMDF is built on HDF5 and includes both a data specification and an API to support structured and unstructured data.  http://emrl.byu.edu/xmdf1/index.html


CJ Beegle-Kraus and Chris Barker - User’s Often Need More Information About the Model Than You Think They Do

People that use model results, rather than run or develop models, often have different metadata needs than the modelers themselves.  Potential uses are: emergency response, environmental effects, ecosystem modeling, and boundary conditions for other models.  Key elements: The users are likely to need (1) model output values at arbitrary (x,y,z,t) locations in time and space, and (2) a great deal of knowledge about the grid, boundaries, and model domain.


Rich Signell - The importance of the API: some lessons from CF conventions for structured grid

The CF conventions seek to represent increasingly complex data relationships and grid structures.  An Application Programming Interface, API, is a set of programs, subroutines and data structures which interface between the conventions and their actual use.  An API is necessary to take advantage of the metadata conventions and allows generalization of application tools well beyond the one or two models originally targeted by the programmers.

A well conceived and supported API will avoid recoding the same logic and formula into each client application. NetCDF-Java is an API that gives access to many higher level CF operations for structured grids, such as returning coordinates for requested variables and computing vertical positions on-the-fly  from stretched coordinate models.   Examples will be shown using the NetCDF Java API in a stand-alone java client (IDV) and in a java-enabled analysis and visualization environment (Matlab).

How the API is designed and implemented can play a large role in how easy it is to develop clients and how efficiently and effectively those clients work.  We need to discuss the relationship between the API, clients, and conventions as we formulate a strategy for interoperability with unstructured mesh data.
 
 
  Contact Us     Site Map     Search     Terms and Conditions     Privacy Policy     Participation Policy
 
National Science Foundation (NSF) UCAR Office of Programs University Corporation for Atmospheric Research (UCAR)   Unidata is a member of the UCAR Office of Programs, is managed by the University Corporation for Atmospheric Research, and is sponsored by the National Science Foundation.
P.O. Box 3000     Boulder, CO 80307-3000 USA     Tel: 303-497-8643     Fax: 303-497-8690