The large surveys are looking for more than variable stars -- everything from extrasolar planets to weak lensing. Variable stars are a stated aim for most, but secondary. (High-z SN are the only variables mentioned on the LSST front page!) Small surveys (like ASAS) are geared towards variable stars. Theyre cheap! But we are bumping into the magnitude limits of these small telescopes.
Wish-list for Variable Star Surveys Photometry: multicolor, high-S/N, large dynamic range long time baseline dense temporal sampling Multi-site Wide area? Or not? It is very hard to design a survey that will satisfy everyones needs. Variable stars exhibit such a huge variety of behavior that one size cant fit all.
Surveys, in practice Telescope time / resources Dynamic range Data pipeline capacity Analysis capacity (& time, & interest...) Funding... (Sloan 2.5-m)
Astrophysics from cluster surveys: variability Intrinsic variability fraction Short-period pulsators, asteroseismology Pulsation-rotation connection Magnetic activity versus [age,rotation] Binaries, binary fraction, binary evolution Planet searching
So anyway, we have this data... How the NGC 2301 variables project started…
NGC 2301 In the plane, but low extinction (~820 pc) MSTO around A0 (~180 Myr) Near-solar abundances NGC 2301 provides a good test of young stellar evolution & isochrones, and searches for variability in a coeval group of upper main sequence stars.
NGC 2301 data (Tonry et al. 2005) UH 2.2-meter Two week span B-R calibration, R- band time-series ~2 dozen points per night
Analysis challenges Time-series analysis -- CPU time Aperiodic/quasiperiodic stars Difficult lightcurves (binaries, spots) Spurious signals, bad data detection? Classification Variable star data analysis is not always straightforward; large-scale, automated analysis design is non-trivial.
What kinds of variables? Close EBs (EB/EW/KW?) pulsators (dSct, gDor, aCyg?, bCep?) RR/delta Cep (background, probably) rotating? (RS CVn...) Because of the two-week data span, the data are naturally biased towards the shortest period stars. Thus, the close EBs dominate. Other types are hard to study and firmly classify with this amount of data.
Upper MS pulsators alpha Cygni beta Cephei SPBs delta Scuti gamma Doradus The MSTO of NGC 2301 lies around A0, so delta Scuti and gamma Doradus stars should be present. Perhaps some cluster Cepheids, too?
Gamma Doradus stars sp. type mid-late F, lum. class V, Pop I P: 8h - 3d (close to P(rot)!) sometimes multiperiodic g-mode pulsators (n < 0) The gamma Doradus class is a new designation (early 1990s). Before they were discovered, they were sometimes unknowingly used as comp stars!
The star is a few magnitudes fainter than the MSTO (B-R) ~ +1.5 But... The period is just right The field is in the Galactic plane The light curve is beta Cep (Sterken & Jaschek 1996)
Several other interlopers are suspected in this cluster, including background RR Lyrae and Cepheid variables. If this is a beta Cep star, it suggests that classifications based upon presumed HR diagram positions must be done with care. (That includes binaries.)
Variability study: in progress Common-envelope binaries are nearly sorted out, wider ones and interacting ones not Pulsators? Have done two multiperiodic tests (out of 4000...) Rotating & Magnetic stars? (probably some alpha CVn stars at the bright end -- spectra would be neat!) Background stars Bad data & alias rejections The photometry was a dry-run for upcoming surveys. The data analysis is becoming a dry-run for LONEOS...
Building a variable catalog quick and dirty time series sort by color, mag, period, amplitude visual inspection? computer methods? With 4000 stars, 3200 of which show some hint of variability, the analysis, classification, and study of this sample is proving a non- trivial task. This is ONLY 4000 stars -- what do you do with millions?
Learning from the past Large-scale time-series databases like Hipparcos, MACHO, OGLE, and ASAS provide reasonable examples for how to deal with massive databases: process them quickly, build a database, and either... Work on classes/clusters individually, and/or Release the data and let the community do the work MACHO kept their data proprietary for a long time, but still got lots of results. The others went public sooner -- great data, but less press?
A work in progress... At the end, a detailed variable star census of NGC 2301 will provide us with some interesting astrophysics (net variability fraction, variability versus [X]) Asteroseismology? (At least we can find new candidates for followup photometry) A good, small-scale project for building large-scale analysis pipelines -- find what works, what doesnt Archival data wants to be free! (And disks are only getting cheaper!)