Improved Identification of Satellite Trails in ACS/WFC Imaging Using a Modified Radon Transform

TL;DR

This paper introduces the Median Radon Transform (MRT) as a robust detector for linear artifacts, notably satellite trails, in ACS/WFC imaging and implements it in the findsat_mrt pipeline within acstools. By using the median along potential trail paths, the method remains sensitive to faint, extended trails while resisting bright astronomical sources, achieving ~85–88% recovery of visually identified trails with ~2.5% false detections after excluding diffraction spikes. The study provides a thorough sensitivity analysis, validates results against Frontier Fields data, and analyzes two decades of ACS/WFC data to show the satellite-trail incidence has doubled from 2002 to 2022, with no clear evolution in trail brightness. While effective, the approach has limitations in crowded fields and corner regions, and the authors discuss future enhancements including distortion-corrected, multi-chip analyses and faster MRT computation to broaden applicability.

Abstract

We present a new approach to identify satellite trails (or other linear artifacts) in ACS/WFC imaging data using a modified Radon Transform. We demonstrate that this approach is sensitive to features with mean brightness significantly below the background noise level, and it is resistant to the influence of bright astronomical sources (e.g., stars, galaxies) in most cases. Comparing with a set of satellite trails identified by eye, we find a trail recovery rate of 85\% and a false detection rate (after removing diffraction spikes that are easily filtered) of 2.5\%. By performing an analysis using a much larger ACS/WFC data set where false trails are identified by their persistence across multiple images of the same field, we identify the Radon Transform parameter space and image properties where our algorithm is unreliable, and estimate a false detection rate of $\sim10\%$ elsewhere. We apply our method to ACS/WFC data taken between 2002 and 2022 to determine both the frequency of satellite trail contamination in science data and also the typical trail brightness as a function of time. We find the rate of satellite trail contamination has increased by approximately a factor of two in the last two decades, but there is no clear systematic evolution in the typical trail brightness. Our satellite trail identification program is available as part of the \texttt{acstools} package.

Figures (12)

• Figure 1: The coordinates used to parameterize linear paths over which summations are calculated in the RT. In this example the path of interest (thick orange line) is crossing the $+x,+y$ quadrant and is parameterized by the coordinates $\theta$ and $\rho$, where $\theta$ represents the angular orientation of the line measured relative to the $x$-axis and $\rho$ represents the distance between the origin and the line. Note that by convention $\theta$ spans $0^{\circ}$ to $180^{\circ}$. The full range of parameter space is filled by having $\rho$ go from $-\infty$ ($-\infty < y < 0$) to $+\infty$ ($0 < y < +\infty$).
• Figure 2: (top) A model image with five labeled trails. (bottom-left) Corresponding RT of the image with each trail labeled. (bottom-right) Same as bottom-left but using the MRT. The signals in RT/MRT space are elongated because the model trails are several pixels thick.
• Figure 3: A comparison of the standard (bottom-left) and median (bottom-right) Radon Transforms for a single chip of a real ACS/WFC image (rebinned $2\times2$) with a satellite trail (top). Red arrows indicate the location of the signal from the satellite trail in the transformed images. Due to numerous bright astronomical sources, the standard RT is filled with streaks of intensity comparable or brighter than the signal from the satellite trail itself. The MRT significantly reduces the intensity of these additional streaks, making the satellite trail itself the dominant feature in the transformed image.
• Figure 4: Examples of trails identified by both findsat_mrt and previous by-eye inspection. The red bands highlight the trails.
• Figure 5: Examples of trails found via by-eye inspection but missed by findsat_mrt. The red bands highlight the by-eye trails. (top-left) The trail was identified but later rejected due to having persistence $<0.5$. (top-right) The trail was identified but rejected due to being too wide, likely due to the alignment with the bright galaxies in the frame. (bottom panels) Trails passing through the corners of the images. Roughly half of missed trails fall close to the corners.