[Paper Review] The u'g'r'i'z' Standard Star Network

J. A. Smith, D. L. Tucker+20 more|ArXiv.org|Jan 10, 2002

Stellar, planetary, and galactic studiesPhysics and Astronomy995 citations

TL;DR

This paper establishes a 158-star u'g'r'i'z' standard star network at the U.S. Naval Observatory, Flagstaff, to enable photometric calibration of the Sloan Digital Sky Survey (SDSS). Using precise, multi-night observations with a 2.5-m telescope and specialized reduction techniques, the authors achieve sub-1.5% photometric accuracy across all bands, meeting or exceeding SDSS's stringent calibration goals for uniformity and precision in the u', g', r', i', z' filters.

ABSTRACT

We present the 158 standard stars that define the u'g'r'i'z' photometric system. These stars form the basis for the photometric calibration of the Sloan Digital Sky Survey (SDSS). The defining instrument system and filters, the observing process, the reduction techniques, and the software used to create the stellar network are all described. We briefly discuss the history of the star selection process, the derivation of a set of transformation equations for the UBVRcIc system, and plans for future work.

Motivation & Objective

• To create a robust, multi-band photometric standard star system for the Sloan Digital Sky Survey (SDSS) using u'g'r'i'z' filters.
• To achieve photometric calibration accuracy of ≤1.5% in u' and z', and ≤1% in g', r', and i' for primary standard stars.
• To develop a self-contained, reproducible standard star network that supports uniform, high-precision photometry across the Northern Galactic Cap.
• To enable transformation between the u'g'r'i'z' system and the traditional UBVRcIc system via calibrated transformation equations.
• To ensure long-term stability and accuracy through rigorous monitoring of instrumental response, extinction, and mirror reflectivity changes.

Proposed method

• Observations were conducted at the U.S. Naval Observatory, Flagstaff Station, using a 2.5-m telescope equipped with u'g'r'i'z' filters to collect multi-night photometric data on 158 primary standard stars.
• Instrumental magnitudes were derived using aperture photometry and calibrated against known standard stars, with corrections applied for atmospheric extinction and instrumental response.
• Zeropoint and extinction coefficients were computed nightly and monitored over time, with breaks in the data stream indicating mirror re-aluminization events.
• Red leak correction was applied to u'-band magnitudes to account for residual red leakage in the u' filter, improving u'-g' color accuracy.
• Transformation equations between the u'g'r'i'z' system and the UBVRcIc system were derived using overlapping measurements from Landolt standards.
• Statistical analysis of mean errors across magnitude and color ranges confirmed the system's compliance with SDSS's photometric accuracy requirements.

Experimental results

Research questions

• RQ1 What is the optimal set of 158 stars that can serve as a stable, uniformly distributed u'g'r'i'z' photometric standard network across the Northern Galactic Cap?
• RQ2 How can photometric calibration accuracy be maintained at the 1%–1.5% level across all five u'g'r'i'z' bands despite instrumental and atmospheric variations?
• RQ3 To what extent do filter characteristics and mirror degradation affect the photometric zeropoints and extinction coefficients over time?
• RQ4 How accurately can the u'g'r'i'z' system be transformed into the traditional UBVRcIc system using overlapping standard stars?
• RQ5 What is the impact of red leakage on u'-band photometry, and how can it be corrected to maintain color accuracy?

Key findings

• The u'g'r'i'z' standard star network achieves a mean calibrated magnitude uncertainty of ≤1.5% in u' and z', and ≤1% in g', r', and i', meeting or exceeding the SDSS calibration targets.
• Red leak correction reduced systematic errors in u'-g' colors, with corrections below 0.01 mag for all but a few stars, ensuring high color accuracy.
• Zeropoint stability was maintained over time, with measurable shifts only at mirror re-aluminization events, confirming the robustness of the calibration process.
• The network spans a wide range of magnitudes and colors, with stars distributed across the celestial sphere, including dense coverage near the celestial equator and sparse coverage in the southern hemisphere.
• Transformation equations between the u'g'r'i'z' and UBVRcIc systems were derived with high confidence, enabling cross-calibration with existing photometric catalogs.
• The mean error in photometry was consistently below the SDSS requirement of 0.02 mag in r', (r'-i') and (g'-r'), and 0.03 mag in (u'-g') and (i'-z') for blue objects.