Removing correlated noise stripes from the Nancy Grace Roman Space Telescope survey images

TL;DR

The paper addresses the challenge of correlated $1/f$ noise in the Nancy Grace Roman Space Telescope's Wide Field Instrument, which biases weak-lensing measurements. It introduces imDestripe, a destriping algorithm that uses Roman's multiple roll angles to interpolate backgrounds and solves for per-row offsets via conjugate gradient, yielding robust suppression of stripe power. In hybrid simulations combining real detector noise with modeled skies, imDestripe reduces large-scale stripe power by a factor of $10$ to $30$, significantly mitigating additive shear biases. The authors discuss integration with the existing IRRC correction and PyIMCOM coaddition within the WFI pipeline, outline limitations such as edge-SCA behavior, and propose future enhancements (priors, sky-background corrections) to realize near-systematics-free weak-lensing measurements. The work provides open-source tooling that strengthens Roman’s cosmology pipeline and supports high-precision infrared weak lensing in future surveys.

Abstract

Weak gravitational lensing has emerged as a powerful tool for investigating the matter distribution in the Universe and how it has evolved over cosmic time. The Wide Field Instrument (WFI) on the Nancy Grace Roman Space Telescope (Roman) will deliver some of the highest precision measurements of weak lensing ever made. Since weak lensing is based on statistics of faint sources, it can be biased by even tiny instrument systematics, including correlated read noise. Previous works have shown the infrared detectors used in the Roman WFI show correlations in their noise fields at a level significant for weak lensing measurements, even after application of standard reference pixel corrections; of particular concern is 1/f noise, which appears as horizontal banding in the detector frame. In this paper, we present imDestripe: a new Python module utilizing the multiple roll angles in Roman's observing strategy and linear algebra techniques to remove correlated noise stripes from observed images. We test imDestripe in a hybrid simulation by combining real noise realizations (from darks taken during ground testing) with simulated images of the astronomical scene, and find that the power spectrum of the banding can be suppressed by factors of 10--30 on large scales. We briefly discuss plans for further development of imDestripe in the context of the WFI pipeline.

Figures (10)

• Figure 1: An illustration of a Roman SCA layout. Pixels that are read out at the same time in the fast-scan direction show a read noise correlation.
• Figure 2: An example slope image made from a dark integration on SCA 10. Hot pixels, characteristic 1/f noise stripes, and the reference output channel (the dark, rightmost channel) can be seen here.
• Figure 3: Binned power spectrum of the average noise slope image for SCA 10. The frequency axis shows the cycles each mode completes across the image length-- note the peaks at 4K and 8K. Left: Linear scale power spectrum shows stationary noise on most scales. Right: Log-log plot reveals a region of correlated $1/f$ noise at low frequencies.
• Figure 4: Flow chart depicting the main process of imDestripe. Light blue directives are calculated image-by-image for each SCA contributing to the mosaic, while the dark blue steps are executed while considering all SCAs at once. The purple squares show the variables representing the product of each step, to help the reader relate steps described here to equations in the paper.
• Figure 5: Comparison of a masked simulated image (Observation 670 on SCA 10) before and after destriping. Left: The masked image before destriping; horizontal stripes can be seen in the noise. Center: The masked image after destriping; horizontal stripes are no longer visible. Right: The optimal offset parameters for this image.