Detecting false positives with PLATO using double-aperture photometry and centroid shifts

This paper proposes and validates a double-aperture photometry strategy as a computationally efficient alternative to on-board centroid shifts for detecting false positives in the PLATO mission, demonstrating that secondary flux measurements achieve the highest detection efficiency (92%) for rejecting eclipsing binary contaminants among Sun-like stars.

The Gist

Imagine you are a detective trying to find a tiny, invisible mouse (an exoplanet) scurrying across a giant, bright stage light (a star). The mouse is so small that when it crosses the light, it only dims the beam by a tiny fraction. Your job is to spot that tiny dip in brightness.

However, there's a problem: the stage is crowded. Other actors (background stars) are standing right next to the main light. Sometimes, one of these background actors is actually a pair of stars dancing in a circle, eclipsing each other. When they do, they create a dip in brightness that looks exactly like the mouse crossing the main stage. These are "False Positives" (FPs)—fake mice that trick your detector.

This paper is about a new, clever strategy for the PLATO space mission (a giant space telescope launching soon) to tell the difference between a real mouse and a fake one, without running out of battery or data space.

The Problem: Too Much Data, Not Enough Brainpower

PLATO will stare at hundreds of thousands of stars. Every day, it generates a mountain of data (over 100 Terabits!). But the radio link to Earth is narrow (like a straw trying to drink from a firehose). PLATO can't send all the raw images back to Earth to be analyzed. It has to do the hard work on the spaceship itself.

The spaceship has limited computer power (CPU) and memory. It can't run complex calculations on every single star to check for fakes. It needs a quick, cheap, and efficient way to say, "This looks like a real planet," or "This is probably a fake, throw it away."

The Old Trick: The "Center of Gravity" Test

One standard way to catch a fake is to look at the Centroid Shift.

• The Analogy: Imagine a seesaw. If a real mouse (planet) crosses the main star, the center of the light stays almost exactly in the middle. But if a background star (the fake) dims, the "center of gravity" of the light in your camera window will wobble toward that background star.
• The Catch: Calculating this wobble is computationally expensive. It requires the spaceship to do heavy math for every star. PLATO can only afford to do this for a tiny fraction (5% to 20%) of its targets. For the rest, they need a cheaper trick.

The New Trick: The "Double-Aperture" Strategy

The authors propose a new method called Double-Aperture Photometry. Instead of just looking at the star through one "window" (aperture), they look through two different windows simultaneously.

Think of it like wearing two pairs of glasses with different frames:

1. The Nominal Mask (The Standard Frame): This is the perfect, tight frame around the main star. It's designed to get the clearest picture of the target star.
2. The Extra Frame: They add a second frame. They test two types of extra frames:
   • The "Extended" Frame: This is a slightly larger frame that includes the main star plus a ring of neighbors. It's like widening your view to see if a neighbor is sneaking in.
   • The "Secondary" Frame: This is a small, focused frame placed directly on the most suspicious neighbor star. It's like putting a magnifying glass directly on the guy standing next to the star who looks most guilty.

How It Works: The "Deeper Dip" Test

Here is the magic logic:

• Scenario A (Real Planet): If a real planet crosses the main star, it blocks light from the main star. In the Standard Frame, you see a dip. In the Extended Frame (which includes more light from neighbors), the dip looks smaller or the same because the extra light from the neighbors "dilutes" the darkness.
• Scenario B (Fake Planet/Background Star): If a background star (the fake) is the one dimming, the Standard Frame sees a small dip (because the main star's light is overwhelming it). But the Secondary Frame (focused on the fake star) or the Extended Frame (which catches more of the fake star's light) will see a much deeper, more dramatic dip.

The Rule: If the dip looks deeper in the extra frame than in the standard frame, it's a fake! The spaceship can instantly flag it and discard it without needing to calculate the expensive "wobble" (centroid shift).

The Results: What Worked Best?

The authors ran millions of simulations to see which method caught the most fakes.

1. The "Secondary Flux" (The Magnifying Glass): This was the champion. By focusing specifically on the most suspicious neighbor, it caught 92% of the fakes. It's the most efficient tool.
2. The "Extended Centroid" (The Wobble with a Wide View): This came in second, catching 87% of fakes.
3. The "Secondary Centroid": Caught 75%.
4. The "Extended Flux" (The Wide View): Caught 73%.

Why does this matter?
The "Flux" methods (looking at brightness dips) are 50% cheaper for the spaceship's computer than the "Centroid" methods (calculating wobbles). They require less data to be sent back to Earth.

The Final Strategy: A Smart Mix

The paper concludes with a smart plan for PLATO to use its limited resources:

• If a star has one suspicious neighbor: Use the Secondary Flux (the magnifying glass). It's cheap and catches almost everything.
• If a star has multiple suspicious neighbors: Use the Extended Flux or Centroid shifts to keep an eye on the whole group.
• If there are no obvious neighbors: Just use a default wide view to be safe.

The Takeaway

This paper shows that PLATO doesn't need to be a supercomputer to find Earth-like planets. By using a clever "two-window" trick to compare brightness, it can filter out the vast majority of fake signals right on the spaceship. This saves massive amounts of data and computing power, allowing the mission to focus its resources on the stars that actually have real planets.

It's like having a bouncer at a club who doesn't need to check everyone's ID (expensive centroid shifts); instead, they just ask, "Did you come in alone?" (checking the secondary mask). If you came in with a suspicious friend, you're out!

Technical Summary

Here is a detailed technical summary of the paper "Detecting false positives with PLATO using double-aperture photometry and centroid shifts."

1. Problem Statement

The PLATO (PLAnetary Transits and Oscillations of stars) space mission aims to discover exoplanets around Sun-like stars. A critical challenge for PLATO is the massive volume of data generated (over 100 Terabits/day), which far exceeds the downlink bandwidth (435 Gbits/day). Consequently, photometry for the vast majority of targets (specifically the P5 statistical sample, containing ~245,000 stars) must be processed on-board.

• The False Positive (FP) Issue: Many transit-like signals are not caused by planets but by Eclipsing Binaries (EBs) located in the background or nearby (contaminants). Distinguishing these FPs from true planetary transits is essential for mission validation.
• The Constraint: The standard method for FP detection is measuring centroid shifts (the movement of the light source's center of brightness during a transit). However, due to computational and telemetry limits, on-board centroid calculations can only be performed for 5% to 20% of the P5 targets.
• The Gap: An efficient, low-cost on-board strategy is required to detect FPs for the remaining ~80% of targets where centroids cannot be computed.

2. Methodology

The authors propose and evaluate a double-aperture photometry strategy as an alternative or complement to centroid shifts. This involves extracting flux using two different masks per target window:

1. Nominal Mask: The standard aperture optimized for the target star to maximize Signal-to-Noise Ratio (SNR).
2. Alternative Mask: An additional aperture designed to capture contaminant flux. The authors tested two types:
   • Extended Mask: The nominal mask expanded by a ring of pixels (typically 1 pixel) to encompass surrounding contaminants.
   • Secondary Mask: A smaller mask centered specifically on the most prominent contaminant (the one with the highest Stellar-Pollution Ratio, SPR) in the window.

Simulation Framework:

• Data Source: Gaia DR3 catalog (13.6 million stars) to define P5 targets and their background contaminants.
• Assumptions: Contaminants are modeled as Eclipsing Binaries (EBs) with transit depths and durations drawn from observed distributions (Kepler catalog).
• Metrics Evaluated:
  • Flux Metrics: Comparing transit depths in the nominal vs. extended/secondary masks. A deeper transit in the alternative mask indicates the signal originates from a contaminant.
  • Centroid Metrics: Computing centroid shifts using nominal, extended, and secondary masks.
• Efficiency Definition: The ratio of detectable FPs by a specific method to the total number of FPs detectable by the nominal mask (the ground truth baseline).

3. Key Contributions

• First Efficiency Assessment: This is the first study to quantitatively assess the efficiency of double-aperture photometry for PLATO's on-board FP detection.
• Resource Optimization: The study demonstrates that double-aperture photometry requires 50% less CPU and telemetry than centroid shift calculations (requiring only one scalar flux value vs. x/y coordinates).
• Strategic Framework: The authors develop a decision matrix for assigning detection metrics to targets based on the number of potential contaminants ($N_{FP}$) and available on-board resources.

4. Key Results

The study simulated the P5 sample and compared the detection efficiency of various metrics:

A. Contamination Landscape:

• ~40% of P5 targets have no contaminants capable of creating an FP.
• ~35% have exactly one FP-creating contaminant.
• ~22% have two or more FP-creating contaminants.

B. Detection Efficiency (Averaged across P5 magnitude range):

1. Secondary Flux: 92.1% (Most efficient).
2. Extended Centroid: 87.1%.
3. Nominal Centroid: 83.7%.
4. Secondary Centroid: 75.4%.
5. Extended Flux: 73.5% (Standard condition); rises to 87.2% if a differential analysis (subtracting nominal from extended flux) is applied.

C. Resource vs. Performance Trade-off:

• Secondary Flux is the most efficient metric overall, particularly for targets with a single dominant contaminant.
• Centroid shifts (Nominal/Extended) are highly effective for targets with multiple contaminants, as they can detect shifts from any of them, whereas secondary flux only monitors the single brightest contaminant.
• Cost: Flux measurements are significantly cheaper in terms of telemetry and CPU than centroid calculations.

5. Proposed Strategy for On-Board Implementation

Based on the efficiency and resource constraints, the authors propose a tiered assignment strategy for the 80,000 on-board targets per camera:

Number of Contaminants ($N_{FP}$)	Recommended Metric	Rationale
$N_{FP} = 0$	Extended Flux (Default)	Low cost; acts as a safety net for missed contaminants.
$N_{FP} = 1$	Secondary Flux	Highest efficiency (92%) for single contaminants; low resource cost.
$N_{FP} \ge 2$	Extended/Nominal Centroids	Needed to monitor multiple contaminants simultaneously; flux methods may miss signals from non-dominant contaminants.

Resource Allocation Outcome:
The proposed strategy fits within the strict on-board limits (max 14,800 centroids and 45,400 flux measurements per camera):

• Secondary Flux: ~25,760 targets.
• Extended Flux: ~19,830 targets.
• Centroids (Nominal + Extended): ~6,400 targets.
• Total: ~52,000 targets (well within the 80,000 limit), ensuring optimal FP rejection while respecting telemetry budgets.

6. Significance

This work provides a critical roadmap for the PLATO mission operations. By validating that double-aperture photometry is a highly efficient, low-cost alternative to centroid shifts, the paper enables PLATO to:

1. Discard a large fraction of False Positives (specifically those caused by EBs) before data is even downlinked to Earth.
2. Optimize Telemetry: Prioritize the download of light curves that are most likely to contain true planetary signals, maximizing the scientific return of the mission.
3. Future Missions: The methodology serves as a blueprint for future transit-based exoplanet missions facing similar data volume constraints.

In conclusion, the paper demonstrates that a hybrid approach—using secondary flux for single-contaminant targets and centroid shifts for multi-contaminant targets—allows PLATO to effectively validate its massive target list within strict hardware limitations.