🔭 astrophysics

Nancy Grace Roman Space Telescope Wide Field Instrument: Bright Point Source Saturation Response and Persistence Properties from Thermal-Vacuum Testing

This paper presents pre-flight thermal-vacuum test results characterizing the Nancy Grace Roman Space Telescope's Wide Field Instrument, revealing that deep saturation from bright stars (~4 mag) creates large saturated regions (~150 pixels) with adjacent non-linear charge leakage, while persistence signals decay to background levels within approximately 20 minutes regardless of source magnitude.

Original authors: Dana R. Louie, Robert F. Wilson, Thomas Barclay, Joshua E. Schlieder, Nicholas Bond, Evan Bray, Mario Cabrera, Stephanie Cheung, Ami Choi, Analia Cillis, Nicholas Collins, Tyler D. Groff, Robert J. Hi

Published 2026-07-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Dana R. Louie, Robert F. Wilson, Thomas Barclay, Joshua E. Schlieder, Nicholas Bond, Evan Bray, Mario Cabrera, Stephanie Cheung, Ami Choi, Analia Cillis, Nicholas Collins, Tyler D. Groff, Robert J. Hill, Jeffrey Kruk, Gregory Mosby, Jennie Paine, Bernard J. Rauscher, Timothy A. Reichard, Maxime Rizzo, Scott Rohrbach, Nicole Schanche, Eric R. Switzer

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 trying to take a perfect photograph of a starry night sky. You want to capture the faint, twinkling lights of distant galaxies, but your camera has a problem: if a bright star shines too directly into the lens, the sensor gets overwhelmed. In the world of astronomy, this is called "saturation." Think of a pixel in a camera sensor like a tiny bucket waiting to catch rain (light). If the rain falls too hard, the bucket overflows. In a digital camera, that overflow doesn't just spill over the edge; it can leak into the neighboring buckets, filling them up with water they didn't earn. Even after the storm passes and the buckets are empty, some water might cling to the sides or leak back in later, creating a ghostly "afterimage" known as "persistence."

This is a big deal for the Nancy Grace Roman Space Telescope, a massive space observatory designed to take wide-angle pictures of the universe. It will be staring at incredibly crowded star fields, like the Galactic Bulge, where thousands of bright stars will hit its sensors at once. If the telescope's cameras get too full of light from these bright stars, the overflow could ruin the pictures of fainter objects nearby, or leave behind lingering ghosts that mess up future observations. Scientists needed to know exactly how much light the camera could handle before it started leaking and how long those ghosts would stick around, especially for stars much brighter than anything they had tested before.

The Big Test: Shining a Flashlight on a Camera

To solve this mystery, a team of scientists put the Roman telescope's camera, called the Wide Field Instrument (WFI), through a grueling test inside a giant vacuum chamber that simulates the cold, empty conditions of space. They didn't just use regular lights; they used a special projector to beam simulated stars onto the camera's sensors. These "stars" ranged from very dim to incredibly bright—some as bright as the planet Venus or even brighter, far exceeding what the camera would normally see in a single snapshot.

The researchers wanted to see two main things: first, how big the "overflow zone" gets when a super-bright star hits the camera, and second, how long the "ghosts" (persistence) linger after the light is turned off. They tested this on two specific camera chips, which they called SCA 4 and SCA 11, shining nine different "stars" on them, ranging from a magnitude of 4 (very bright) to 18 (very faint).

The Overflow: A Spreading Wave of Light

The results were fascinating. When they shone a moderately bright star (about magnitude 12) on the sensor for roughly 170 seconds, the area where the pixels were completely full and overflowing grew to about 15 pixels wide. But when they turned the dial up to a super-bright star (magnitude 4), the overflow zone exploded in size, growing to a massive circle about 150 pixels wide after the same amount of time.

Even more interesting was what happened to the pixels right next to the overflow. The scientists found that as the "flood" of light expanded, the pixels on the edge didn't just sit there; they started acting strangely. They began to leak charge from their saturated neighbors, causing the signal to jump up in a non-linear way. It's like if you filled a bathtub to the brim, and the water didn't just spill over the side, but actually pushed the water in the next bathtub over the edge too, even before that second tub was full. The team observed that this "charge leakage" happens immediately when a neighbor saturates, creating a dynamic, shifting pattern that changes with every single read of the sensor.

The Ghosts: Fading Away Quickly

Next, they checked for the "ghosts." After shining the bright lights, they turned them off and watched the sensors in the dark to see how long the lingering signal lasted. They found that for all the stars they tested, from the dimmest to the brightest, the ghostly signal faded away surprisingly fast. Within about 20 minutes, the signal dropped down to the normal background level of the detector (less than 0.05 electrons per second).

This is great news for the telescope's future. It means that even if the camera gets blinded by a super-bright star, the "afterimage" won't ruin the next few hours of observations. The persistence decays to the background level in a timeframe that is manageable for the mission's schedule. Interestingly, they also noticed something weird with the brightest stars: a faint, glowing halo of light that seemed to spread out from the center, but with dark shadows cast by tiny defects in the camera material. They believe this is caused by light generated inside the camera material itself, traveling sideways and getting blocked by flaws, creating a shadow effect.

What This Means for Space Exploration

The paper concludes that while the Roman telescope's cameras will definitely get saturated by the brightest stars in the universe, the resulting "leakage" and "ghosts" are well-behaved enough to be managed. The saturation zones grow predictably, and the persistence fades quickly enough that it won't ruin the telescope's ability to study the cosmos. The team has shared all their data and analysis tools with the public, inviting other scientists to dig deeper and help build better models for how to handle these bright stars when the telescope finally starts its mission. They didn't find a way to stop the overflow entirely, but they proved that the camera can handle it without breaking, and that the ghosts don't last long enough to be a major problem.

Technical Summary: Nancy Grace Roman Space Telescope Wide Field Instrument: Bright Point Source Saturation Response and Persistence Properties from Thermal-Vacuum Testing

Problem Statement
The Nancy Grace Roman Space Telescope's Wide Field Instrument (WFI) is designed to conduct large-scale astrophysical surveys, including the Galactic Bulge Time Domain Survey (GBTDS), which will observe dense stellar fields containing hundreds of thousands of bright stars. Sources brighter than approximately 17th magnitude will saturate WFI detector pixels during typical survey exposures, with the brightest stars (down to ~4th magnitude) causing deep saturation across large pixel regions. While previous characterization of Roman's H4RG-10 detectors explored persistence under flat-field illumination up to ~15–16th magnitude, the regime of deep point-source saturation and the resulting persistence signals were not previously characterized. Understanding these effects is critical for predicting detector response in flight and ensuring the accuracy of subsequent observations potentially impacted by persistence.

Methodology
To address this gap, the authors conducted a "Bright Star Saturation" test during the WFI's second Thermal Vacuum test campaign (TVAC2) at BAE Space & Mission Systems in Boulder, CO. The test utilized the Stimulus of Ray Cones (SORC) telescope simulator to project nine in-focus point sources through the F146 filter onto two Sensor Chip Assemblies (SCAs), specifically SCA 4 and SCA 11.

Test Configuration: The SORC projected point sources with fluxes tuned to approximate stellar magnitudes ranging from ~4 to ~18. The illumination was continuous wave (CW) to mimic flight observations.
Exposure Sequence: For each source magnitude, the team acquired one illuminated exposure consisting of 56 frames (1 reset-read frame + 55 science frames), resulting in an integration time of approximately 170 seconds. This was followed immediately by an interleaved dark exposure to capture persistence decay on a timescale of minutes. This sequence was repeated for all nine magnitudes on both SCAs.
Data Analysis:
- Saturation: The authors applied Improved Roman Reference Correction (IRRC) and superbias subtraction. They defined four saturation regimes (not saturated, diagonal neighbor, row/column neighbor, and saturated) to analyze pixel slope behaviors. They compared instantaneous slopes against mean slopes derived from non-saturated frames to identify non-linearities and charge leakage.
- Persistence: Persistence was calculated by deriving slopes from interleaved dark frames and subtracting dark current. The authors developed specific methods to correct for structured stray light and fiber contamination from the SORC projector, which inadvertently leaked light during dark exposures. They flagged bad pixels and saturation masks (defined by electron counts between 100,000 and 130,000 $e^-$ ) to isolate persistence signals.

Key Results

Saturation Growth: The saturated region of a ~4 magnitude source expanded to approximately 150 pixels in diameter after ~170 seconds of illumination, compared to ~15 pixels for a ~12 magnitude source.
Charge Leakage and Non-Linearity: Pixels adjacent to the expanding saturation front exhibited pronounced non-linear behavior. Specifically, the instantaneous slope of a pixel increased immediately after a neighboring pixel saturated, consistent with charge leakage from saturated neighbors. This effect was dynamic, changing with each read, rather than remaining constant until classical saturation effects dominated.
Persistence Decay: For all source magnitudes tested (~4 to ~17), the median persistence signal in the first post-illumination dark exposure was broadly consistent. Crucially, persistence decayed to detector background levels ( $\lesssim 0.05 \, e^- s^{-1}$ ) within approximately 20 minutes post-illumination. This decay time is consistent with flat-field persistence measurements from the same TVAC2 campaign.
Flux Halo and Defect Shadows: In the deepest saturation cases (~4 mag), a radially extended flux halo was observed surrounding the saturated PSF. Intriguingly, this halo was partially blocked by detector defects, casting "shadows" on the far side of the defects. The authors attribute this to spontaneous radiative emission near the material bandgap edge, where saturated HgCdTe material temporarily loses absorption capability, emitting photons that travel laterally through the detector layer.
Spatial Variation: Persistence performance varied spatially across the detectors. SCA 4 showed significant spatial variation, which impacted the measured persistence values depending on the projection location, whereas SCA 11 was more uniform.

Significance and Claims
The paper claims that these pre-flight characterization results provide a baseline assessment of WFI detector response to deep point-source saturation, directly informing the astronomical community's preparation for Roman science operations.

Performance Improvement: The authors note that Roman's persistence performance (median signal $\lesssim 0.1 \, e^- s^{-1}$ at 10 minutes post-saturation for SCA 11) represents a $\sim 10\times$ improvement over Euclid's NISP arrays (operating at a similar temperature) and is comparable to JWST's NIRCam (which operates 50 K colder). This improvement is attributed to specific design changes in the H4RG-10 detectors driven by Roman's persistence requirements.
Independence from Magnitude: The initial persistence signal appears largely independent of the source magnitude for the range tested, provided the detector is saturated.
Data Availability: The authors emphasize the value of the dataset for the community, making analysis products (IRRC and gain corrected data cubes, animations, and plots) publicly available via Zenodo and the Roman Research Nexus to encourage further development of saturation and persistence models.

The paper concludes that while the persistence decays rapidly to background levels, the dynamic nature of charge leakage near saturation fronts and the spatial variation of persistence across the focal plane must be accounted for in future calibration models and survey planning.

More like this