🔭 astrophysics

Systematic KMTNet Planetary Anomaly Search. XIII. Complete Sample of 2021 Prime Field Planets

This paper presents the thirteenth installment of the Systematic KMTNet Planetary Anomaly Search, identifying seven hidden planetary systems and three candidates in 2021 Prime Field data to demonstrate the continued necessity of systematic searches for constructing a complete, unbiased sample of microlensing planets.

Original authors: In-Gu Shin, Jennifer C. Yee, Weicheng Zang, Cheongho Han, Andrew Gould, Shude Mao, Chung-Uk Lee, Yoon-Hyun Ryu, Ian A. Bond, Takahiro Sumi, Michael D. Albrow, Sun-Ju Chung, Kyu-Ha Hwang, Youn Kil Jung

Published 2026-05-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: In-Gu Shin, Jennifer C. Yee, Weicheng Zang, Cheongho Han, Andrew Gould, Shude Mao, Chung-Uk Lee, Yoon-Hyun Ryu, Ian A. Bond, Takahiro Sumi, Michael D. Albrow, Sun-Ju Chung, Kyu-Ha Hwang, Youn Kil Jung, Yossi Shvartzvald, Hongjing Yang, Sang-Mok Cha, Dong-Jin Kim, Seung-Lee Kim, Dong-Joo Lee, Yongseok Lee, Byeong-Gon Park, Richard W. Pogge, Fumio Abe, David P. Bennett, Aparna Bhattacharya, Ryusei Hamada, Yuki Hirao, Stela Ishitani Silva, Shota Miyazaki, Yasushi Muraki, Kansuke NUNOTA, Greg Olmschenk, Cl'ement Ranc, Nicholas J. Rattenbury, Yuki K. Satoh, Daisuke Suzuki, Takuto Tamaoki, Sean K. Terry, Paul. J. Tristram, Aikaterini Vandorou, Hibiki Yama

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the night sky as a giant, cosmic game of "whack-a-mole." In this game, stars occasionally line up perfectly, and the gravity of a foreground star acts like a magnifying glass, briefly brightening a background star. This is called a microlensing event. Sometimes, if that foreground star has a planet, the planet creates a tiny, extra "blip" or "bump" in the brightness curve.

For years, astronomers have been watching these events, mostly looking for these blips with their own eyes. But just like trying to find a needle in a haystack by staring at it, you might miss the tiny needles hidden in the hay.

This paper is the report from a team of astronomers who decided to stop just staring and start using a robotic detective (a semi-machine algorithm called AnomalyFinder) to scan through a massive archive of data from 2021. They were looking specifically at the "Prime Fields," which are the most crowded, star-rich areas of the sky, hoping to find planets that human eyes had missed.

Here is what they found, explained simply:

1. The Robot Found Hidden Treasures

The team analyzed data from the KMTNet telescope network (which has three telescopes in Chile, South Africa, and Australia, working together like a relay team to watch the sky 24/7).

The Result: They discovered 7 confirmed new planetary systems and 3 strong candidates (events that look like planets but have some confusing features that need more study).
The Impact: These new discoveries represent about one-third of all the planets found in that specific area of the sky in 2021. This proves that even though human astronomers are good at their job, a systematic, robot-assisted search is essential to find the "hidden" planets that would otherwise be missed. Without this robot, our picture of the galaxy's family of planets would be incomplete.

2. What Are These Planets Like?

Most of the planets they found are what you might call "standard issue" for this type of cosmic search:

The Hosts: They mostly orbit M-dwarf stars, which are small, cool, red stars (the most common type of star in our galaxy).
The Planets: They range from "Super-Jupiters" (huge gas giants) to "Sub-Neptunes" (smaller, rocky or icy worlds).
The Location: Most of these planets are orbiting far away from their stars, beyond the "snow line" (the distance where it's cold enough for ice to form). This is similar to how Jupiter and Saturn are far from our Sun.

3. The "Debate" Over Some Events

Just like a detective might have two theories about a crime, the astronomers found some events where the data could be explained in two different ways.

The "Parallax" vs. "Xallarap" Mystery: For one specific event (KMT-2021-BLG-0690), the data looked like it could be caused by the Earth moving around the Sun (a "parallax" effect) OR by the source star itself wobbling because it has a companion (a "xallarap" effect). The robot couldn't definitively pick a winner, so the team presented both possibilities. It's like seeing a shadow and not being sure if it's a person walking or a tree swaying in the wind.
The "Candidate" List: For three events, the data was too messy to be 100% sure. They could be planets, or they could be binary stars (two stars orbiting each other) mimicking a planet. The team labeled these "candidates" and put them on a "watch list" for future study.

4. Why Does This Matter?

The authors explain that to understand the "demographics" of our galaxy (how common different types of planets are), you need a complete sample. If you only count the easy-to-find planets, your statistics will be wrong.

The Analogy: Imagine trying to count all the fish in a lake. If you only use a net that catches big fish, you'll conclude there are no small fish. This paper is about using a finer mesh net to catch the small, hidden fish so we can get an accurate census of the lake.
The Conclusion: By finding these hidden planets, the team is helping to build a more accurate map of how common planets are in the Milky Way. They found that the distribution of planet sizes doesn't follow a simple, broken line (as some older theories suggested) but looks more like a "double hill" (two distinct groups of planet sizes).

5. The "False Alarms"

The paper also includes an appendix listing 12 events that looked like planets at first glance but turned out to be something else (like two stars orbiting each other or a binary star system). The team included this list to save other astronomers time, essentially saying, "Don't bother checking these; we already did, and they aren't planets."

In Summary:
This paper is a success story of teamwork between humans and machines. By letting a computer algorithm sift through the data, the team found a significant number of hidden planets that would have been missed otherwise. This helps us build a more complete and unbiased picture of the planetary families living in our galaxy.

Technical Summary: Systematic KMTNet Planetary Anomaly Search. XIII. Complete Sample of 2021 Prime Field Planets

Problem and Motivation
The Systematic KMTNet Planetary Anomaly Search series aims to construct a complete, unbiased sample of microlensing planets to facilitate robust statistical studies of planet demographics in the Galaxy. While the "by-eye" search method remains the primary channel for detecting microlensing planets (accounting for approximately two-thirds of discoveries), previous phases of this series (covering 2016–2019) demonstrated that systematic, semi-machine-based searches are crucial for uncovering "hidden" planetary systems that evade visual inspection. This work extends that series to the 2021 bulge season, specifically targeting the KMTNet high-cadence "Prime fields" (BLG01, BLG02, BLG03, BLG41, BLG42, and BLG43), to expand the existing microlensing planet sample and ensure its completeness.

Methodology
The study utilizes the AnomalyFinder algorithm, a semi-machine-based search tool previously developed and refined in works by Zang et al. (2021, 2022b) and Shin et al. (2023b, 2024). The methodology proceeds as follows:

Anomaly Detection: AnomalyFinder is applied to the KMTNet datasets (processed via the pySIS pipeline) to identify anomalous events. For the 2021 season, OGLE-IV data were unavailable due to the pandemic; thus, the analysis relied on KMTNet data and, where available, MOA datasets.
Light Curve Modeling: Identified anomalies undergo rigorous modeling. The search grid covers $\log_{10}(s) \in [-1.0, 1.0]$ and $\log_{10}(q) \in [-5.0, 1.0]$ with $200 \times 200$ grids. Local minima are refined using $\chi^2$ minimization and heuristic analysis.
Degeneracy Resolution: The authors apply strict criteria to claim a planet detection:
- The best-fit model must have a mass ratio $q < 0.03$ .
- Planet/binary degeneracy must be resolved with $\Delta\chi^2 > 10$ .
- 2L1S/1L2S (binary source) degeneracy must be resolved with $\Delta\chi^2 > 15$ .
- Events failing these criteria but showing $q < 0.03$ in at least one model are classified as "planet candidates."
Higher-Order Effects: For events with timescales $t_E > 15$ days, annual microlensing parallax (APRX) and orbital motion (OBT) effects are investigated. For events where standard models yield physically improbable parameters (e.g., extremely low proper motion), alternative interpretations such as the xallarap (XRP) effect are explored.
Source Characterization: Angular source radii ( $\theta_*$ ) are determined via Color-Magnitude Diagram (CMD) analysis using KMTNet and VVV data.
Lens Property Estimation: Lens masses and distances are derived either directly (if both $\theta_E$ and $\pi_E$ are measured) or via Bayesian analysis using Galactic priors, incorporating constraints from $t_E$ , $\theta_E$ , $\rho_*$ , and $\pi_E$ distributions.

Key Results
The systematic search of the 2021 Prime fields identified seven confirmed planetary events and three planet candidates. These new discoveries represent approximately 33% of the total microlensing planets found in the 2021 Prime fields, a fraction consistent with the ~35% contribution of systematic searches in the 2016–2019 phase.

Confirmed Planets:
- KMT-2021-BLG-0424: A sub-Neptune ( $0.78 M_N$ ) orbiting an M-dwarf ( $0.23 M_\odot$ ) at $\sim 2.0$ au, located in the Galactic disk ( $\sim 2.5$ kpc). The solution involves an ecliptic degeneracy in the APRX model.
- KMT-2021-BLG-0457: A Jupiter-mass planet ( $\sim 1.1 M_J$ ) orbiting an M-dwarf ( $\sim 0.23 M_\odot$ ) at $\sim 1.9$ au, located in the bulge ( $\sim 6.9$ kpc).
- KMT-2021-BLG-0690: A complex case with competing APRX and XRP solutions. Both suggest a super-Jupiter ( $\sim 10-12 M_J$ ) orbiting an M or K dwarf. The APRX solution implies an extremely low proper motion and a very small lens-source separation, while the XRP solution suggests a more typical disk location.
- KMT-2021-BLG-1063: A sub-Jupiter planet ( $\sim 0.3-0.7 M_J$ ) orbiting an M-dwarf ( $\sim 0.4 M_\odot$ ) at $\sim 1.7-2.9$ au.
- KMT-2021-BLG-1691: A multi-planet system interpretation involving either a 2L2S or 3L1S model. The solutions suggest super-Jupiter or sub-Jupiter planets orbiting an M-dwarf.
- KMT-2021-BLG-2213: A Jupiter-mass planet ( $\sim 1.1 M_J$ ) orbiting an M-dwarf ( $\sim 0.23 M_\odot$ ) at $\sim 1.2$ au.
- KMT-2021-BLG-3290: A super-Jupiter ( $\sim 1.9 M_J$ ) orbiting an M-dwarf ( $\sim 0.55 M_\odot$ ) at $\sim 2.4-3.2$ au.
Planet Candidates:
- KMT-2021-BLG-1385, 1751, 1907: These events show $q < 0.03$ in some models but suffer from unresolvable degeneracies (e.g., 2L1S/1L2S or planet/binary) that prevent definitive classification as planets.
Non-Planetary Events: The authors identified 12 additional events that initially appeared planet-like but were conclusively modeled as binary-lens or binary-source systems (e.g., KMT-2021-BLG-0148, 0338, 0341, etc.) upon detailed analysis.

Significance and Claims
The paper asserts that this work reinforces the necessity of systematic anomaly searches to build a complete microlensing planet sample. The authors highlight that relying solely on visual inspection leaves a significant fraction of planetary systems undetected.

The primary significance lies in the contribution to statistical studies of planet demographics. The authors note that previous mass-ratio distributions based on incomplete samples showed conflicting features, such as "breaks" in the power law or dips at specific mass ratios. By adding the 2021 sample to the existing dataset, the authors observe that the "dip" feature in the mass-ratio distribution ( $-3.6 < \log_{10}(q) < -3.0$ ) becomes clearer, and the distribution is better described by a double Gaussian rather than a broken power law. They caution that while the 2021 sample supports the double Gaussian model, the inclusion of older, uncontrolled samples (pre-KMTNet) obscures these features, likely due to publication bias and sample contamination. Consequently, the authors conclude that continuing systematic searches is essential to resolve the true morphology of the planet mass-ratio distribution and to conduct unbiased statistical studies of the Galactic planet population.

1. The Robot Found Hidden Treasures

2. What Are These Planets Like?

3. The "Debate" Over Some Events

4. Why Does This Matter?

5. The "False Alarms"

More like this