Re: CUAHSI Observations Data Model

• To: Ben@xxxxxxxxxxxxxxxx
• Subject: Re: CUAHSI Observations Data Model
• From: "Rudolf Husar" <rhusar@xxxxxxxxxxxx>
• Date: Wed, 3 Jan 2007 12:26:41 -0600

Hi Ben, David, All,

Thank you for bringing the hydrology data modeling effort to our attention.

I can offer a few thoughts on this data model hamonization issue.

- In dealing with air quality point monitoring data, both practice and
theory has shown that the relational data model proposed by CUAHSI is a
robust way of encoding monitoring data. In fact virtually all AQ monintoring
data are being managed through Relational DMS, in SQL servers. A typical
(actual) AQ data schema is in slide
2<http://capita.wustl.edu/capita/capitareports/070103OGC_WCS_SQL/AQ_Pt_OGC.ppt>.


- The relational model is suitable to encode and to respond to typical
OGC queries for data (WCS), station description (WFS), sensor description
(SOS) etc. Some ideas on this are noted in slide
3<http://capita.wustl.edu/capita/capitareports/070103OGC_WCS_SQL/AQ_Pt_OGC.ppt>
.

- For web service-based data access, it should not matter how the data are
encoded and managed, as long as they repond to the servvice query.

- In this view, the key in harmonizing between the two (two or five) data
models is in the service return payload format. netCDF?, XLS, CSV, whatever
is appropriate for the particualar client (community).

-  So, connecting the pipes is a matter of using different data format
adopters. For the GALEON experiments, for example, we have formatted the
monitoring data as a 1 dim netCDF-CF data structure.

We are very much looking forward comparing notes on this topic and also
experimenting with implementation in GALEON 2.

On 1/3/07, Ben Domenico <bendomenico@xxxxxxxxx> wrote:

Hi all,

The US hydrology community represented by CUAHSI  (Consortium of
Universities for the Advancement of Hydrologic Sciences, Inc.) HIS
(Hydrologic Information System) has drafted a document called "CUAHSI
Community Observations Data Model Working Design Specifications."
These design specifications are targetted at a relational database for
hydrologic point datasets that are collected at distributed hydrologic
observatories.  David Tarboton welcomes input from the THREDDS and OGC
communities, so I'm forwarding his message which has pointers to the
draft specifications and a questionnaire.

I am struck by the fact that the THREDDS group is developing storage
and access systems for the same sort of point/station observations via
our traditional netCDF approach


http://www.unidata.ucar.edu/software/netcdf-java/formats/UnidataObsConvention.html

and the OGC Sensor Web Enablement (SWE) initiative addresses many of
the same data access issues from the perspective of standards-based
interfaces.

http://www.opengeospatial.org/pt/06-046r2

So we have two discipline-oriented communities of practice developing
very different approaches to data storage and access issues for
datasets that have many common characteristics and a group working
with evolving international standards in the same general area.  If
you look at the diagram in "The Middle Ground" section of


http://www.unidata.ucar.edu/projects/THREDDS/GALEON/Phase2Connections/FeaturesFilesScientificDataModels.htm

these projects are classic examples of the stovepipe foundations and
-- hopefully -- the conduit pipe between them.

I thought each group should at least be aware of the work of the others.

In addition, if any of you have input for David Tarboton, he welcomes
it before January 31st and has specific requests for feedback in the
questionnaire that he points to.

David Tarboton <dtarb@xxxxxxxxxx>

Thanks.
-- Ben

• References:
  • Fwd: CUAHSI Observations Data Model
    • From: Ben Domenico