Time Domain presentation at CSP

[Douglas Tody]

On Wed, 17 Jun 2015, Roy Williams wrote:

> LIGO has recently published several terabytes of time-series data, and it is 
> not about photometry or light curves, but rather gravitational-wave strain at 
> 4 kHz (https://losc.ligo.org/S6/). There are also megabytes of time-series 
> data that are 1 Hz boolean flags about data quality.  So its a bit strange to 
> read Tody's document that says: "a time series deals with flux vs time".

A time series of course, expresses some quantity as a function of time.
For time series, this is often analysis as much as obsevation so the
measured quantity could be anything.

Here is what I actually said in the document you guys reference:

     The class of time series data dealt with here consists of a sequence of
     discrete values represented as a function of time. A typical example
     would be a light curve measuring flux as a function of time for a single
     astronomical object, possibly in several colors or photometric bands,
     but in general any quantity could be the dependent variable.

I haven't been keeping up to date with all the latest developments with
SDM (the IVOA spectral data model), but there is little difference
between a spectrum and a time series, so the basic model should be fine
for both.  Time series do have some unique attributes of course, but
much of the basic metadata should be common.

> Perhaps the IVOA could more carefully distinguish between "time series" and 
> its sub-class "light curve"?

Yes, a light curve is a subclass of a time series.  Probably the most
important subclass, but time series is much more than that.

 	- Doug

> Roy
>
> ---
> Caltech LIGO
> roy at caltech.edu
> 626 395 3670
>