One of the software tools being used by the Smart Telescope Working Group looks at photometric accuracy, and that requires us to correlate stars across multiple images, catalogs, the VSX database, and AAVSO star charts. I’m trying to understand how these coordinate system frames either match or require position adjustments to support that correlation. (And I apologize if this is a naive question, but I’m confused.)
For example, positions in VSX are labeled “J2000” on the download screen. Are these in the ICRS frame with proper motion adjusted to each star’s J2000 position? Or does this imply an earth-centered frame such as fk4 or fk5 using the earth’s axis alignment for the time 2000.0?
The VSP charts give a RA/Dec for the chart center position (usually, the position of a variable star) as “J2000”. Same question for what this means. In addition, the VSP charts now say that “Stars in the chart are drawn at J2015.5.” Is this referring only to proper motion? Or is this saying that positions on the chart are fk4 or fk5 based on earth’s axis alignment for the time 2015.5?
Photometry provided by VSP/VSD provide a RA/Dec that is unlabeled. What frame is used as the reference for this?
I know that I have to be careful with APASS positions, because they are “Coordinates are equinox J2000, but are measured on epoch of observation without any precession.”
I ask all this because I know what frame is used by astrometry.net for plate solutions (ICRS) and I’m trying to correlate these ICRS positions against AAVSO positions. (ICRS coordinates are not referenced to the Earth’s axis, so aren’t subject to Earth-based precession effects.)
Hi Mark,
I spoke with Sebastian, Bert, an George about your question. Unfortunately, the answer for questions regarding both AAVSO’s photometric reference frame and astrometric reference frame are the same: our catalogs are assembled from multiple sources so AAVSO has neither a single astrometric or photometric reference frame. While this may not sound great, it often doesn’t matter in practice.
Before I get into the details, I highly suggest reading a 1999 Annual Review paper entitled Reference Frames in Astronomy by Johnston and Vegt. They do an excellent job explaining the idea of reference frames and discuss major developments that took us from Hipparchus (the person) to HIPPARCOS (the satellite) with a preview of what GAIA could provide.
The vast majority of AAVSO’s astrometric data comes from one of the following sources:
- GAIA DR3: The data from this catalog are in ICRF J2016; however, when we imported the data into VSX, about 90% had proper motions so we propagated them back to ICRF J2000. The remaining 10% of stars without proper motions are in ICRF J2016. Generally speaking, these objects are either very faint or binaries.
- HIPPARCOS: The data from this catalog are in ICRF J1991.25; however, when we imported them we propagated them to IRCS J2000.
- UCAC4: This catalog consists of historic sources (see Section 2c of the documentation) that were re-reduced using the Hipparcos (ICRF) reference frame. The raw data have different epochs depending on the available data. When they were imported into AAVSO’s systems, they were propagated to ICRF J2000.
- Catalog specific source. There are a multitude of papers and catalogs with variable star discoveries that only list the positions of objects with no other cross-references. In these situations, the data were imported as-is. That means they are in an unknown reference frame (likely FK5 or ICRF) and at an unknown epoch (likely either J2000 or J-whenever-it-was-observed). For example, data from the Zwicky Transient Facility (ZTF) are ICRF J-whenever.
- FK5 (or earlier) - There are likely several instance of stars that have FK5 (or earlier) coordinates in the database; however, these are relatively rare. This system predates ICRS/ICRF and is both dynamical and kinematic in nature.
Generally speaking, this mixture of reference frames has not been a considerable source of problems. However, a few of our observing programs have special charts that are generated in J-this-year format. Programs involving nearby stars (e.g. red dwarf programs) or high proper motion stars (i.e. Bernard’s Star) are moving sufficiently quickly that the errors between J2000 and J-now are evident.
As I write this, we are importing GAIA’s variable star catalog (some 9 million objects) into VSX. Later this year, we intend to adopt GAIA’s positions and proper motions for all objects in VSX. For some of the brighter objects, we may need to use HIPPARCOS positions and proper motions. When GAIA DR4 and DR5 come out, we’ll do something similar. There are likely many objects in VSX where the discovery coordinates are too imprecise to make a match to either of these, but most objects should be good to go.
Brian
1 Like