Time series in IVOA?

[Matthew Graham]

Hi Rob,

On Jan 30, 2012, at 6:25 AM, Rob Seaman wrote:

> Howdy,
> 
> The IVOA Roadmap for 2012 is now available:
> 
>> From: Christophe Arviset <Christophe.Arviset at esa.int>
>> Subject: The IVOA in 2011: Technical Assessment and Roadmap for 2012
>> Date: January 30, 2012 3:07:44 AM MST
>> To: interop at ivoa.net
>> 
>> Dear all
>> 
>> Following many inputs from Working Groups and Interest Groups chairs and vice-chairs, "The IVOA in 2011: Technical Assessment and Roadmap for 2012" has been endorsed by the IVOA Exec and is now available at:
>> 
>> http://www.ivoa.net/Documents/Notes/IVOATechRoadmap2011/index.html
>> 
>> Many thanks to all the people who have participated to this document !
> 
> VOEvent is mentioned a couple of dozen times.  We should be eager to see the registry work move forward, for instance.  I'm skeptical that cone search in a time window really addresses the science use cases for SEAP, however.

Our community solicitation exercise last summer revealed that most are just happy with cone search + time window and don't want a full query against the VOEvent data model. 

> On the other hand, time series are mentioned just three times.  Massively backpedaling on the schedule (2012 is for use cases when 2010/2011 was previously a target for the final product):
> 
>> 3.3.7 Other DAL standards
>> ...
>> Time Series and Event access protocols require that data model work be closer to completion before the DAL WG can assess whether existing standards can be used or new standards are required. Specifically, we expect to use TAP for discovery (along the same lines as the ObsCoreDM is used) and will try to use other DAL standards (e.g. DataLink and PQL) to satisfy other use cases. We will be collecting use cases in 2012.
> 
> and similarly underwhelming:
> 
>> 3.4.1 Relationship between the various Data Models
>> ...
>> 	o A potential TimeSerieDM still to be defined
> 
> and very subjunctively:
> 
>> 3.8.2 Other VOEvent standards
>> Following the early discussions which have taken place within the VOEvent WG about a possible Simple Light Curve Access Protocol (SLiCAP), it now appears that once the time series data model is defined, light curves could be retrieved through a suitable DAL interface, e.g. TAP, hence there would be no need to defined a specific access protocol for light curves.
> 
> In the absence of an undefined potential TimeSeries DM, thoughts on conveying such data sets pragmatically?  References to external SimpleTimeSeries?  Embedded table with interim convention for representing some useful subset of same?

As you aware, IVOA schedules are not set in stone (nor are they contractually binding) and sometimes things need to be changed to reflect strategy changes. This is what happened with TimeSeries last year (in Naples). TimeSeries is part of the suite of DMs that inherit from Spectral 2.0 DM, which is actively worked on by the DM group. Once this is stable, a TimeSeries spec will appear. We do have a need for it beyond including time series in VOEvents - the VAO time series work would really benefit from it. VOEvent 2.0 allows the inclusion of tabular data and this is the mechanism we should be promoting if you want to bloat your VOEvent packet with light curves.

	Cheers,

	Matthew




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/voevent/attachments/20120130/1096d7cc/attachment.html>