🔭 astrophysics

PlatoSim: An end-to-end PLATO camera simulator for modelling high-precision space-based photometry

This paper introduces PlatoSim, a state-of-the-art end-to-end simulator designed to accurately model the complex multi-telescope architecture of the ESA PLATO mission, thereby supporting critical preparatory activities such as mechanical integration, performance studies, and pipeline development through high-precision photometric predictions.

Original authors: N. Jannsen, J. De Ridder, D. Seynaeve, S. Regibo, R. Huygen, P. Royer, C. Paproth, D. Grießbach, R. Samadi, D. R. Reese, M. Pertenais, E. Grolleau, R. Heller, S. M. Niemi, J. Cabrera, A. Börner, S. Ai

Published 2026-04-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: N. Jannsen, J. De Ridder, D. Seynaeve, S. Regibo, R. Huygen, P. Royer, C. Paproth, D. Grießbach, R. Samadi, D. R. Reese, M. Pertenais, E. Grolleau, R. Heller, S. M. Niemi, J. Cabrera, A. Börner, S. Aigrain, J. McCormac, P. Verhoeve, P. Astier, N. Kutrowski, B. Vandenbussche, A. Tkachenko, C. Aerts

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 you are about to launch a massive, high-tech camera into deep space to take the most perfect, uninterrupted photos of the universe ever taken. This camera, part of the PLATO mission, is designed to find Earth-like planets orbiting other stars. But before you can fire up the real engine, you need to make sure the camera won't glitch, the software won't crash, and the photos won't be ruined by space dust or weird vibrations.

Enter PlatoSim.

Think of PlatoSim as a hyper-realistic "Flight Simulator" for a space camera. Just as a pilot uses a simulator to practice landing a plane in a storm without ever leaving the ground, the scientists building the PLATO mission use PlatoSim to "fly" the camera through virtual space, testing every possible scenario before the real hardware ever leaves Earth.

Here is a breakdown of how this digital twin works, using some everyday analogies:

1. The Camera: A Swarm of 26 Eyes

The PLATO mission isn't just one camera; it's a swarm of 26 small telescopes working together like a hive mind.

The N-CAMs (Normal Cameras): These are the main workers, taking steady, high-quality photos of stars for years.
The F-CAMs (Fast Cameras): These are the "eyes on the prize," moving super fast to keep the whole spacecraft steady, like a photographer using a gimbal to stop their hand from shaking.

PlatoSim simulates all 26 of these eyes at once, figuring out how they overlap and how they see the same star from slightly different angles.

2. The Simulation: Building a Virtual Universe

When you run a simulation in PlatoSim, it doesn't just draw a pretty picture. It builds a virtual physics engine that mimics reality step-by-step:

The Light Show: It starts by calculating how many photons (light particles) hit the camera. It accounts for the star's brightness, the "fog" of the Milky Way, and even the stray light bouncing off the Moon or Earth.
The Shaky Hand (Jitter): Even in space, the spacecraft wobbles slightly due to its internal motors. PlatoSim simulates this "shaky hand" effect. If the camera moves even a tiny bit while taking a photo, the star's image blurs. The simulator calculates exactly how much noise this adds to the data.
The "Ghost" Images: Just like when you take a photo of a bright light and see a reflection in your lens, space cameras get "ghosts." PlatoSim predicts where these ghost images will appear so scientists can teach the computer to ignore them.
The Cosmic Ray Rain: Space is filled with invisible, high-energy particles (cosmic rays) that hit the camera like tiny bullets, leaving bright white streaks in the photos. PlatoSim simulates this "rain" so the data scientists can write software to wipe those streaks out later.

3. The Detector: The Digital Canvas

The camera uses a special sensor called a CCD (the same kind of technology in your digital camera, but much more sensitive). PlatoSim simulates how this sensor behaves:

The "Fat" Pixel Effect: When a pixel gets too full of light (like a bucket overflowing), the extra water spills into the neighbor's bucket. This is called the "Brighter-Fatter Effect." PlatoSim models this spill so scientists know how to measure the star's true brightness even when it's blindingly bright.
The Aging Process: Space is harsh. Over 6 years, the camera gets "tired" from radiation. PlatoSim simulates this aging, predicting how the camera will perform on Day 1 versus Day 2,000.

4. Why Do We Need This?

You might ask, "Why not just build it and see what happens?"
Because space is expensive and you only get one shot. If the camera has a flaw, you can't fly up there to fix it.

PlatoSim allows the team to:

Stress Test: They can simulate a "worst-case scenario" (e.g., a massive solar storm or a total loss of pointing stability) to see if the mission survives.
Train the AI: The software that will analyze the data on the spacecraft needs to be trained. PlatoSim generates millions of fake photos with known answers (e.g., "Here is a fake planet transit") to teach the computer how to spot real planets.
Design the Strategy: They use it to decide the best way to point the cameras and how long to take photos to get the clearest data.

The Bottom Line

PlatoSim is the ultimate rehearsal.

It is a piece of software that says, "Let's pretend we are in space, let's pretend the camera is old, let's pretend the star is flickering, and let's see if our plan still works."

By running these simulations, the scientists are ensuring that when the real PLATO mission launches in 2026, it will be ready to capture the first clear, high-definition photos of Earth-like worlds around other stars, potentially answering the biggest question of all: Are we alone?

1. Problem Statement

The PLAnetary Transits and Oscillations of stars (PLATO) mission is an ESA M3 mission designed to detect and characterize exoplanets, particularly Earth-analogues, and determine stellar ages via asteroseismology. To achieve its stringent scientific goals (photometric precision $\le$ 50 ppm $h^{-1/2}$ for solar-like stars), the mission relies on a complex, innovative multi-telescope payload consisting of 26 cameras (24 "Normal" N-CAMs and 2 "Fast" F-CAMs).

Prior to launch, rigorous testing and validation are required to ensure the instrument meets performance requirements. However, existing simulation tools for PLATO (such as PSLS and PINE) were either simplified, pragmatic, or lacked the ability to simulate the full end-to-end data flow (from photon arrival to digital light curves). There was a critical need for a comprehensive, end-to-end simulator capable of:

Modeling the complex multi-telescope geometry and overlapping fields of view.
Simulating pixel-level physics, including non-linearities, charge transfer inefficiencies, and various noise sources.
Generating realistic time-series data (images, meta-data, housekeeping, and light curves) to test data processing pipelines and assess mission performance under realistic conditions.

2. Methodology

The authors present PlatoSim, a state-of-the-art, end-to-end simulator developed in C++ with a Python wrapper. It models the entire signal chain from incoming photons to the final digital measurement. The methodology is structured into several key modules:

Instrument Architecture: The simulator models the specific hardware of PLATO, including 26 refractive telescopes (Telescope Optical Units - TOUs), 104 CCDs (Teledyne-e2v CCD270), and Front-End Electronics (FEEs). It distinguishes between N-CAMs (25s cadence, full-frame) and F-CAMs (2.5s cadence, frame-transfer).
Image Acquisition Model:
- Photon Accumulation: Models the accumulation of photo-electrons from stars, sky background, and dark current, integrating over the exposure time.
- Gain and Non-linearity: Implements a polynomial model for CCD/FEE gain non-linearity (well-depth dependence) and temperature-dependent gain drifts.
- Readout Smearing: Accounts for the lack of a mechanical shutter, where charge accumulates during the readout phase, creating vertical trails.
Light and Electron Sources:
- Stellar Light: Uses a polychromatic Point Spread Function (PSF) derived from Zemax optical models and weighted by a G0V stellar spectrum. Includes charge diffusion effects.
- Ghost Images: Models internal reflections causing "point-like" and "extended" ghost images.
- Backgrounds: Simulates zodiacal light, galactic background, and cosmic ray (CR) hits using a simplified but validated Poisson-based model.
- Dark Current: Models thermal electrons with spatial non-uniformity (DSNU).
Focal Plane Dynamics:
- Kinematic Aberration: Calculates apparent stellar position shifts due to spacecraft velocity.
- Pointing Jitter: Simulates high-frequency attitude control system (AOCS) jitter using red noise models.
- Thermo-Elastic Drift (TED): Models slow drifts in pointing due to thermal flexure of the optical bench over mission quarters.
- Optical Distortion: Applies the Wang distortion model (radial, tangential, prism) to map sky coordinates to pixel coordinates.
Electron Redistribution:
- Brighter-Fatter Effect (BFE): Models electrostatic repulsion between pixels that alters effective pixel boundaries.
- Charge Transfer Inefficiency (CTI): Simulates charge trapping and release during readout, including time-dependent degradation (BOL to EOL).
- Blooming: Models overflow of charge from saturated pixels into neighboring pixels.
Photometry Module:
- Includes an optimal aperture algorithm that constructs binary masks to minimize the noise-to-signal ratio (NSR).
- Supports PSF inversion via microscanning to reconstruct high-resolution PSFs for accurate photometry.
- Handles mask updates to compensate for long-term drifts.

3. Key Contributions

End-to-End Fidelity: PlatoSim is the first tool to simulate the complete PLATO data chain, from sky photons to downlinked light curves, including housekeeping data and meta-data.
Modular and Configurable Architecture: The software is highly modular, allowing users to toggle specific physical effects (e.g., CTI, BFE, jitter) and configure different observation modes (N-CAM vs. F-CAM, microscanning).
Realistic Noise Modeling: It incorporates a comprehensive suite of noise sources, including instrumental (jitter, TED, CTI, BFE, PRNU) and astrophysical (stellar granulation, oscillations, contaminants) effects.
Open Source Release: The paper announces the public release of PlatoSim (v3) on GitHub, providing the scientific community with access to the simulator and the final frozen PLATO Input Catalogue (PIC).
Validation: The simulator's CR model was validated against the CosmiX code and flight model test data. The photometric performance was validated against theoretical noise estimators (PINE).

4. Results

Performance Validation: Simulations of 150,000 light curves for 10,000 stars demonstrated that the instrument meets the mission requirement of $\le$ 50 ppm $h^{-1/2}$ for $V \le 11$ stars when observed by 24 cameras ( $n_{CAM}=24$ ). The noise budget correctly identifies photon noise as dominant for bright stars and read/sky noise for faint stars.
Hare and Hound Exercise: The authors successfully simulated the detection of an Earth-analogue (transiting a $V=10$ Sun-like star) over 3.3 years. Using the TLS (Transit Least Squares) algorithm on the simulated data, they achieved a Signal Detection Efficiency (SDE) of ~47 for 4 transits, well above the detection threshold, proving the pipeline's ability to retrieve Earth-sized signals despite stellar variability and instrumental systematics.
Application to Mission Prep: PlatoSim has been instrumental in:
- Designing alignment tests for the cameras at ambient temperatures.
- Optimizing thermal vacuum test strategies.
- Validating data compression algorithms and Fine Guidance System feedback loops.
- Developing and testing the data reduction pipeline (stitching, detrending, and mask creation).

5. Significance

PlatoSim represents a critical milestone in the preparation for the PLATO mission. Its ability to generate realistic, pixel-level synthetic data allows the mission consortium to:

De-risk the Mission: Identify and mitigate performance bottlenecks (e.g., saturation, blooming, CTI) before launch.
Develop Pipelines: Train and validate the on-board and on-ground data processing algorithms, ensuring they can handle the complex, time-varying noise environment of space.
Optimize Strategy: Test different observation strategies (e.g., microscanning, mask updates) to maximize scientific yield.
Scientific Verification: Confirm that the mission can achieve its primary goal of detecting and characterizing Earth-like planets and determining stellar ages with high precision.

By bridging the gap between theoretical requirements and realistic instrumental behavior, PlatoSim serves as an indispensable tool for the PLATO consortium, ensuring the mission is ready to deliver high-precision space-based photometry upon its launch in late 2026.