[Paper Review] The KMTNet 2016 Data Release

H. W. Kim, Kyu‐Ha Hwang+19 more|arXiv (Cornell University)|Apr 10, 2018

Geophysics and Gravity Measurements4 references5 citations

TL;DR

This paper presents the 2016 data release from the Korea Microlensing Telescope Network (KMTNet), delivering 2,163 light curves from 97 deg² of sky observed across three southern hemisphere sites (Chile, South Africa, Australia) at cadences from 0.2 to 8 hr⁻¹. The release includes 1,856 'clear' and 307 'possible' microlensing events, with improved event-finding algorithms and data processing, and is publicly available with a 6-month embargo policy for publications.

ABSTRACT

We present Korea Microlensing Telescope Network (KMTNet) light curves for microlensing-event candidates for the 2016 season, which covers an area of 97 sq.deg observed at cadences ranging from Gamma=0.2/hr to Gamma=8/hr from three southern sites in Chile, South Africa, and Australia. These 2163 light curves are comprised of 1856 "clear microlensing" and 307 "possible microlensing" events (including 265 previously released from the K2 C9 field). The data policy is very similar to the one governing the 2015 release. The changes relative to 2015 in the algorithms to find and vet microlensing events are comprehensively described.

Motivation & Objective

• To provide a comprehensive, publicly accessible release of microlensing light curves from the 2016 KMTNet observing season.
• To improve the event-finding algorithm by integrating multi-site, multi-field light curves and removing known variable stars from catalogs.
• To reduce false positives through enhanced vetting procedures and improved photometric calibration.
• To maintain a consistent data policy with prior releases while enabling early access to K2 C9 campaign overlap events.
• To support future microlensing research by upgrading the star catalog and preparing for faster data release cycles.

Proposed method

• Collected and reduced photometric data from three 1.6m telescopes at KMTC (Chile), KMTS (South Africa), and KMTA (Australia), all observing in I-band for ~95% of observations.
• Applied the DIA and pySIS photometry pipelines to generate light curves from ~3×10⁸ individual light curves derived from stacked images.
• Implemented a multi-site, multi-field event-finding algorithm that jointly fits microlensing light-curve profiles across all three observatories.
• Incorporated vetting checks to exclude known variable stars from published catalogs and prior KMTNet seasons.
• Upgraded the star catalog by replacing the DoPhot-based catalog with the CFHT-based catalog of Schlafly et al. (2018), increasing the catalog size from 3×10⁸ to 5×10⁸ stars.
• Released data via the KMTNet web portal with two access tiers: full embargoed release and expedited access for K2 C9 overlap events.

Experimental results

Research questions

• RQ1 How can multi-site, multi-field microlensing event detection be improved through joint light-curve fitting across distributed observatories?
• RQ2 What is the impact of removing known variable stars from catalogs on reducing false positive microlensing event candidates?
• RQ3 How does the upgraded star catalog affect photometric precision and event detection efficiency in dense fields?
• RQ4 What are the performance differences between DIA and pySIS photometry pipelines in the context of microlensing light curve extraction?
• RQ5 Can a streamlined data release policy balance rapid public access with scientific priority for early-arriving publications?

Key findings

• The 2016 KMTNet data release includes 2,163 light curves covering 97 deg² of sky, with 1,856 classified as 'clear microlensing' events and 307 as 'possible microlensing' events.
• The event-finding algorithm was enhanced to process light curves from all three KMTNet sites simultaneously, improving detection efficiency and reducing false positives.
• A total of 265 events from the K2 C9 field were previously released, and the remaining 307 events are newly released in this dataset.
• The data were observed at cadences ranging from 0.2 hr⁻¹ to 8 hr⁻¹, with prime fields achieving up to 8 hr⁻¹ cadence in overlapping regions.
• The release policy allows public access to all data after publication of this paper and all papers using the data, with a 6-month embargo on submissions.
• A subset of 265 events overlapping the Kepler K2 C9 campaign was released earlier and remains publicly accessible without embargo at http://kmtnet.kasi.re.kr/ulens/event/2016k2/.