UN Science Workshop in Tokyo, June, 2007

[Rob Seaman]

Hi Silvia,

> as you mention there is plenty of scope for interaction of
> VOEvent with the Heliospheric and Ionospheric community.
>
> However, I am not entirely clear about what you mean with
> 'VOEvent is a reasonable choice for reporting almost any time
> sequence of measurements'. Please could you elaborate?

We set out to build a data and transport protocol that could address  
the use case of short latency, high reliability delivery of alert  
messaging in a heterogeneous environment.  The key word there is  
"heterogeneous", since our format needed to be able to support an  
arbitrary collection of dependent variables (<What>) as well as a  
self-consistent and interoperable description of the independent  
variables of an observation (<WhereWhen>).  As Rick points out, the  
various solar and solar system communities have an even greater need  
for coherent world coordinate systems than the classic "fixed star"  
astronomers.

By addressing the specific problem of transient celestial alert  
reporting, we also addressed many of the requirement of a generalized  
format for reporting time domain observations.  In fact, VOEvent  
could be used to report almost any data from remote instrument  
loggers, astronomical or not.  Instrumentation embedded in the canopy  
of the Amazon rain forest faces many of the requirements of  
mountaintop instrumentation.  I wouldn't likely recommend VOEvent be  
used for such purposes, but I also wouldn't automatically reject this  
choice.  It depends on questions regarding other parts of the data  
transport infrastructure, data handling applications, and ultimately  
the experiment and user community in question.

For instance, FITS and IRAF have been used by several non- 
astronomical projects over the years, e.g., STScI had a collaboration  
with a medical project researching mammogram technology.  We had a  
guy sending us pictures of brain slices that he was processing using  
astronomical image processing algorithms.

> There have been efforts to standardise data formats for time
> series data - one such standard, becoming more and more popular,
> is CDF:
> http://cdf.gsfc.nasa.gov/
>
> Another data format that is in use is the CEDAR one, used
> especially for ionospheric data, eg:
> http://www.openmadrigal.org/

Do you happen to know if any of the experiments listed in the IHY  
site are using these formats, or are otherwise constructed to  
interoperate one with the other?

> I don't think there is currently a standard way of reporting
> events, nor of querying for data within a specified time
> range - this is why we drafted the Simple Time Access Protocol
> proposal:
> http://wiki.astrogrid.org/bin/view/Astrogrid/SimpleTimeAccessProtocol

Yes, I was aware of this, but Rick points out some other requirements  
for STAP.  Data transport is also a different discipline than data  
storage and retrieval.  Identifying and benefiting from the areas of  
overlap is precisely the issue I was trying to generate a discussion  
about.

Rob