[Paper Review] Exploring the Variable Sky with LINEAR. I. Photometric Recalibration with SDSS

Branimir Sesar, J. Stuart+4 more|arXiv (Cornell University)|Sep 23, 2011

Stellar, planetary, and galactic studies34 references66 citations

TL;DR

This paper presents a photometric recalibration of the LINEAR asteroid survey data using SDSS as a reference, achieving 0.03 mag precision for bright sources and 0.2 mag at r ∼18. The recalibrated dataset provides ~200 epochs per object for over 25 million stars, enabling deep time-domain studies of variable objects, including 7,000 confirmed periodic variables such as RR Lyrae and eclipsing binaries, significantly extending the dynamic range of prior surveys like NSVS by 3 magnitudes.

ABSTRACT

We describe photometric recalibration of data obtained by the asteroid survey LINEAR. Although LINEAR was designed for astrometric discovery of moving objects, the dataset described here contains over 5 billion photometric measurements for about 25 million objects, mostly stars. We use SDSS data from the overlapping ~10,000 deg^2 of sky to recalibrate LINEAR photometry, and achieve errors of 0.03 mag for sources not limited by photon statistics, with errors of 0.2 mag at r~18. With its 200 observations per object on average, LINEAR data provide time domain information for the brightest 4 magnitudes of SDSS survey. At the same time, LINEAR extends the deepest similar wide-area variability survey, the Northern Sky Variability Survey, by 3 mag. We briefly discuss the properties of about 7,000 visually confirmed periodic variables, dominated by roughly equal fractions of RR Lyrae stars and eclipsing binary stars, and analyze their distribution in optical and infra-red color-color diagrams. The LINEAR dataset is publicly available from the SkyDOT website (http://skydot.lanl.gov).

Motivation & Objective

• To enable precise photometric calibration of the LINEAR survey, originally designed for astrometry, using SDSS as a reference standard.
• To extend the depth and cadence of wide-area variability surveys beyond existing datasets like NSVS and ASAS.
• To provide a publicly available, high-cadence, multi-epoch photometric dataset for studying variable stars and transient phenomena.
• To characterize the distribution of periodic variables in optical and infrared color-color diagrams using the recalibrated data.
• To support future automated variable star detection by providing a visually classified sample of 7,000 confirmed periodic variables.

Proposed method

• Utilized overlapping SDSS data across ~10,000 deg² to recalibrate LINEAR photometry using common stars as photometric standards.
• Applied a statistical framework to correct for non-photometric conditions and instrumental systematics in LINEAR data.
• Calibrated magnitudes using a zero-point offset and color-dependent corrections derived from matched star pairs between LINEAR and SDSS.
• Computed time-series statistics (median magnitude, RMS, chi-squared, skewness, kurtosis) to characterize variability and identify spurious signals.
• Cross-matched LINEAR objects with SDSS and 2MASS catalogs to enable multi-band photometric analysis.
• Conducted visual classification of ~7,000 periodic variable light curves to validate and calibrate automated detection methods.

Experimental results

Research questions

• RQ1 What is the photometric precision achievable after recalibrating LINEAR data using SDSS as a reference?
• RQ2 How does the LINEAR dataset compare in depth and cadence to existing wide-area variability surveys such as NSVS and ASAS?
• RQ3 What is the distribution of periodic variable stars—particularly RR Lyrae and eclipsing binaries—in optical and infrared color-color diagrams?
• RQ4 To what extent do non-Gaussian errors and instrumental artifacts contribute to false positive variability detections in high-cadence surveys?
• RQ5 How can visual classification of light curves improve the reliability of automated variable star detection pipelines?

Key findings

• The photometric recalibration achieved a precision of 0.03 mag for sources not limited by photon statistics, and 0.2 mag at r ∼18.
• LINEAR provides an average of 200 observations per object, enabling time-domain studies of the brightest 4 magnitudes of the SDSS survey.
• The dataset extends the Northern Sky Variability Survey (NSVS) depth by 3 magnitudes, reaching r ∼18.
• The distribution of periodic variables in color-color diagrams shows a bimodal distribution dominated by roughly equal fractions of RR Lyrae and eclipsing binary stars.
• The χ² per degree of freedom distribution for bright LINEAR objects is centered at ∼1, but exhibits a long tail (χ² > 3) indicating a significant fraction of spurious variable detections due to non-Gaussian errors.
• The visual classification of ~7,000 periodic variables provides a high-fidelity training set for future automated variable star detection algorithms.