KMTNet Synoptic Survey of Southern Sky II: Data Reduction and Real-Time Transient Detection Pipeline
Mankeun Jeong, Myungshin Im, Joonho Kim, Seo-Won Chang, Sungho Jung, Chung-Uk Lee, Dong-Jin Kim, Bomi Park, Jaewon Lee, Jiseop Shin, Changwan Kim, Gregory S. H. Paek · 2026
Read KMTNet Synoptic Survey of Southern Sky II: Data Reduction and Real-Time Transient Detection Pipeline by Mankeun Jeong, Myungshin Im, Joonho Kim, Seo-Won Chang, Sungho Jung, Chung-Uk Lee, Dong-Jin Kim, Bomi Park, Jaewon Lee, Jiseop Shin, Changwan Kim, Gregory S. H. Paek (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-17442.txt.
Introduction
Jeong and co-authors describe the data pipeline for the KMTNet Synoptic Survey of the Southern Sky. The pipeline checks image quality, solves astrometry, masks bad pixels, and stacks frames. Two-stage photometric calibration ties zero points to APASS, SkyMapper, and Gaia data. The release covers over 4,000 square degrees at 22 to 23 magnitude depth. Difference imaging on the stacks found transients in real time, they report, including gravitational-wave follow-ups.
Catalog record
| CREATOR | Mankeun Jeong, Myungshin Im, Joonho Kim, Seo-Won Chang, Sungho Jung, Chung-Uk Lee, Dong-Jin Kim, Bomi Park, Jaewon Lee, Jiseop Shin, Changwan Kim, Gregory S. H. Paek |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 89575 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.17442
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
KMTNet Synoptic Survey of Southern Sky II: Data Reduction and Real-Time Transient Detection Pipeline Mankeun Jeong, Myungshin Im, Joonho Kim, Seo-Won Chang, Sungho Jung, Chung-Uk Lee, Dong-Jin Kim, Bomi Park, Jaewon Lee, Jiseop Shin, Changwan Kim, Gregory S. H. Paek Abstract 0009-0003-1280-0099 0000-0002-8537-6714 0000-0002-1294-168X 0000-0002-3118-8275 0009-0009-5870-4266 0000-0003-0043-3925 0000-0002-4292-9649 0009-0008-3499-3043 0009-0005-3910-0337 0009-0005-3944-1457 0000-0002-5266-1658 0000-0002-6639-6533 1]SNU Astronomy Research Center, Astronomy Program, Department of Physics and Astronomy, Seoul National University, Gwanak-gu, Seoul 151-742, Republic of Korea 2]Daegu National Science Museum, 20, Techno-daero 6-gil, Yuga-myeon, Dalseong-gun, Daegu 43023, Republic of Korea 3]Korea Astronomy and Space Science Institute, Daejeon 34055, Republic of Korea 4]Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822, USA \jkashead 1 Introduction In the era of time-domain astronomy, the Korea Microlensing Telescope Network (KMTNet) offers significant advantages through its continuous, wide-field observing capability. KMTNet consists of three identical 1.6-meter telescopes, each equipped with a camera providing a 2 \times 2 square degree field of view (FOV). These telescopes are installed at the Cerro Tololo Inter-American Observatory (CTIO), the South African Astronomical Observatory (SAAO), and the Siding Spring Observatory (SSO) ( Kim et al. 2016a ) . This global distribution enables near-continuous sky coverage with minimal interruption. Such capabilities are ideally suited for the discovery and monitoring of transient phenomena, as demonstrated by several successful target-of-opportunity (ToO) and monitoring programs ( Moon et al. 2016a ; Kim et al. 2018 ; Shin et al. 2019 ; Yang et al. 2024 ) . Furthermore, in the context of gravitational-wave (GW) astronomy, KMTNet’s prompt response across the southern hemisphere is particularly valuable. The discovery of GW170817, which was accompanied by a rapidly fading kilonova, required deep and fast-response optical imaging for successful identification ( Abbott et al. 2017 ; Im et al. 2017 ; Troja et al. 2017 ) . Since that event, timely electromagnetic observations have been recognized as crucial for characterizing a wide range of GW sources and for constraining the nature of their progenitors and environments ( Coughlin et al. 2019 ; Hosseinzadeh et al. 2019 ; Gompertz et al. 2020 ; Kasliwal et al. 2020 ) . KMTNet can reach a 5 \sigma limiting depth of \gtrsim 22 mag with 480-second exposures, making its rapid and sensitive follow-up observations well suited for surveying significant cosmic volumes ( Kim et al. 2021 ; Paek et al. 2024b ; Paek et al. 2025a ; Pillas et al. 2025 ) . However, effective transient identification relies on the availability of pre-existing reference images. Building a large, deep, and uniform reference image database is therefore essential for real-time transient detection, especially given that GW localization area can span several hundred square degrees across unpredictable parts of the sky ( Abbott et al. 2020 ) . Despite this need, no public reference image database has previously been available for KMTNet. The KMTNet Synoptic Survey of the Southern Sky (KS4) was initiated to resolve this limitation (Im et al., in preparation). This survey targets the southern sky at declinations between -85^{\circ} and -30^{\circ} (excluding most of the Galactic plane) with the aim of building a comprehensive reference image database. This paper presents the data reduction pipeline and the procedures used for the data in Data Release 1 (DR1) of KS4, covering observations from November 2019 to December 2023. The detailed survey footprint and field distribution of the 4,000\text{deg}^{2} covered in DR1, as well as the anticipated future coverage, are illustrated in Figure 1 of the companion paper ( Chang et al. 2026 ) and are available on the KS4 project page at NOIRLab’s Astro Data Lab 1 1 1 https://datalab.noirlab.edu/data/ks4 . This dataset includes a catalog of over 200 million sources (with signal-to-noise ratio above 5) and reference images covering more than 4,000 deg 2 across 979 KS4 fields, observed in the Johnson–Cousins BVRI filters with at least four 120-second exposures per filter per field. This data reduction pipeline includes data quality assessment, astrometric and photometric calibration, and image stacking. Previous efforts, such as the Deep Ecliptic Patrol of the Southern Sky ( Moon et al. 2016b ; Chang et al. 2018 , DEEP-South;) and the KMTNet Supernova Program ( Moon et al. 2016a ; Park et al. 2017 ; Park et al. 2019 , KSP;) , have demonstrated the scientific utility of KMTNet for transient science, providing calibrated photometry and reduction procedures for specific targets. Building on that foundation, our pipeline extends this capability to full-sky calibration across thousands of square degrees, incorporating advanced photometric calibration techniques using Gaia XP synthetic photometry ( Gaia Collaboration et al. 2023 ) . The final stacked images reach 5 \sigma depths of 22–23 AB mag in the BVRI filters, with a photometric root-mean-square error (RMSE) of approximately 0.03 mag for sources brighter than 19 AB mag. In addition to reference image construction, the KS4 pipeline is designed to handle real-time processing of newly acquired images from ToO observations using the same modular framework. This includes immediate reduction of ToO imaging data obtained during time-critical follow-up campaigns such as GW events. By applying consistent procedures—data quality assessment, astrometric and photometric calibration, and image subtraction—the pipeline ensures compatibility with pre-built reference images and minimizes latency in transient identification. Recent large-scale surveys such as SkyMapper ( Scalzo et al. 2017 ) , the Dark Energy Survey ( Morganson et al. 2018 ) , and Pan-STARRS ( Magnier et al. 2020 ) have similarly demonstrated that combining deep reference images with automated pipelines is essential for maximizing time-domain science returns. Following this approach, we have made the KS4 DR1 reference image set and the reduction pipeline publicly available 2 2 2 https://github.com/jmk5040/KMTNet_ToO , thereby extending these capabilities to the broader KMTNet community. Lastly, we present selected transient follow-up cases associated with GW events detected by the Advanced Laser Interferometer Gravitational-Wave Observatory (aLIGO; LIGO Scientific Collaboration et al. 2015 ), the Advanced Virgo detector (aVirgo; Acernese et al. 2015 ), and the Kamioka Gravitational Wave Detector (KAGRA; Akutsu et al. 2021 ), which collectively form the LIGO–Virgo–KAGRA (LVK) collaboration. These cases include the events S230518h ( Ligo Scientific Collaboration et al. 2023 ) , S240422ed ( Ligo Scientific Collaboration et al. 2024a ) , S240915b ( Ligo Scientific Collaboration et al. 2024b ) , S250206dm ( Ligo Scientific Collaboration et al. 2025a ) , and S250830bp ( Ligo Scientific Collaboration et al. 2025b ) . For each of these follow-ups, the KS4 reference images and reduction pipeline enabled efficient image subtraction and successful transient identification. These results highlight the practical value of KS4 in real-time ToO campaigns, reinforcing its scientific utility in the context of multi-messenger astronomy. This paper is organized as follows. Section 2 outlines the KMTNet data structure and quality selection criteria. Section 3 describes the reduction pipeline, including astrometric and photometric calibration, bad pixel masking, and stacking procedures. Calibration performance and practical applications in ToO observations are presented in Section 4 . The current limitations of the pipeline and potential directions for future development are discussed in Section 5 . Section 6 concludes with a summary of key results. 2 Data 2.1 KMTNet Image Data Structure The KMTNet camera consists of a mosaic of four charge-coupled device (CCD) chips (each referred to as a “chip image”) with