heliostack : A Novel Approach to Minor Planet Discovery Journal: Planetary Science Journal

Kevin J. Napier, Matthew J. Holman, Hsing Wen Lin (林省文), David W. Gerdes, Thomas R. Ruch · 2026

Read heliostack : A Novel Approach to Minor Planet Discovery Journal: Planetary Science Journal by Kevin J. Napier, Matthew J. Holman, Hsing Wen Lin (林省文), David W. Gerdes, Thomas R. Ruch (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-16731.txt.

Introduction

Napier and colleagues open heliostack, a search tool using nonlinear shift-and-stack for solar system objects. Faint solar system objects light planet origins and growth, yet conventional surveys struggle to detect them. It merges images over longer time spans than ever before, lifting archival and future surveys' find power. Applied to archival Hubble images spanning 15 days in 2003, it fully searches for Cold classic Kuiper Belt Objects. It finds both known sub-threshold objects and adds two new discoveries, the first found in stacks spanning over a day.

Catalog record

CREATORKevin J. Napier, Matthew J. Holman, Hsing Wen Lin (林省文), David W. Gerdes, Thomas R. Ruch
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE43681
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.16731

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

heliostack : A Novel Approach to Minor Planet Discovery Journal: Planetary Science Journal Kevin J. Napier, Matthew J. Holman, Hsing Wen Lin (林省文), David W. Gerdes, Thomas R. Ruch Abstract The study of faint solar system objects is a promising avenue for understanding the origin and evolution of planetary systems. However, such objects are difficult to detect in conventional surveys. Here we introduce heliostack , an algorithm for nonlinear shift-and-stack searches for solar system objects, which enables us to combine images taken over longer time spans than was previously possible. Applying this algorithm to a number of existing archival and forthcoming surveys will allow us to maximize their potential for discovering faint solar system objects. In this work, we apply heliostack to archival Hubble Space Telescope (HST) data, completing an exhaustive search for Cold Classical Kuiper Belt Objects in a set of HST images taken over a 15-day time span in 2003. We successfully recover both of the known sub-threshold objects in the data, and add two new discoveries. These two new objects are the first to ever be discovered in stacks of images taken over a time span longer than about one day. show][email protected] I Introduction Our solar system’s small body populations offer a unique perspective for studying planetary systems. Today the census of minor planets stands at more than one million, which when taken in aggregate, reveal rich and delicate dynamical structures that have been sculpted over billions of years by the gravity of the planets. By studying the dynamical and physical properties of our solar system’s populations of near-Earth objects, asteroids, Jupiter Trojans, centaurs, comets, Neptune Trojans, Kuiper Belt Objects, and beyond, we have gained deep insight about the processes driving the formation and evolution of planetary systems. Still, fundamental questions remain: Is there a massive planet still awaiting discovery in the distant solar system ( K. Batygin et al. 2019 ) ? Are there more populations of small bodies that have thus far eluded detection? To answer these questions we must develop techniques to discover fainter objects so that we can probe smaller and more distant populations. In the two centuries since scientists began working to detect and characterize solar system bodies, progressive improvements in telescope and detector technology have greatly enhanced our capabilities. However, we are reaching the physical limits of detectors and the practical engineering limits of telescopes, and since the objects we are searching for are moving, we cannot simply take longer exposures to increase image depth. One way that we can continue to make progress without improving hardware, is by developing software and algorithms that can better utilize our data. There is a standard procedure for detecting moving objects fainter than the single-image detection threshold, called shift-and-stack, in which one stacks images along the trajectory of some putative object ( J. A. Tyson et al. 1992 ; B. Gladman et al. 1998 ; R. L. Allen et al. 2002 ; G. M. Bernstein et al. 2004 ; M. J. Holman et al. 2004 ; H. Smotherman et al. 2021 ; K. J. Napier et al. 2024 ) . When the signal from the object is combined in a sufficient number of images, the object becomes detectable. The problem, however, is that we do not usually have a priori knowledge of the trajectories of the real objects in our images. In those cases, we must search through all possible orbits of interest. Historically, shift-and-stack searches for solar system objects have not explored a full six-dimensional orbital parameter space, but have instead remained limited to a four-dimensional regime where an object’s apparent motion remains linear in pixel space. In this regime, it is simple to determine which trajectories need to be checked for an exhaustive search (see K. J. Napier et al. 2024 , for example). However, the requirement for linear trajectories limits stacks to data taken over the span of a few hours. When the duration of the exposure sequence gets any longer, the apparent motion of the objects becomes nonlinear in pixel space, increasing the complexity of the problem. Because of the increased complexity over longer time spans, the shift-and-stack technique has thus far been restricted to dedicated shift-and-stack surveys, which typically take of order 100 images of the same patch of sky in a given night ( D. E. Trilling et al. 2024a ; C. A. Trujillo et al. 2024 ) To date, only G. M. Bernstein et al. 2004 has completed a systematic shift-and-stack search for solar system objects in images that were taken over a time span longer than a few hours. There has been recent work in the area of nonlinear digital tracking (see, e.g. N. Golovich et al. 2025 ), but so far no systematic searches have been completed, and no new objects have been discovered. In the two decades since G. M. Bernstein et al. 2004 , with the emergence of GPUs and incredible improvements in computing capacity on CPU clusters, why haven’t any other shift-and-stack searches over long time baselines been completed? The answer may simply be that the problem is difficult from both theoretical and computational perspectives. Just enumerating the trajectories to be searched requires a careful choice of parameters, and even with an efficient parameterization the number of trajectories to search can quickly become unmanageable. Despite its immense difficulties, the problem is worth solving. If we can manage to combine images taken days, months, or even years apart, we will be able to reach the depth of a shift-and-stack search using data from surveys with sparser cadences such as the Dark Energy Survey (DES), or the Legacy Survey of Space and Time (LSST). If we can gain one or two magnitudes of sensitivity over such large survey areas, we will see a substantial increase in the yield of solar system objects, with especially important gains for populations of objects that are small or very distant ( C. A. Trujillo et al. 2024 ; M. Jurić et al. 2019 ) . Furthermore, if we can manage to search for solar system objects in sparser or more irregular image cadences, we can develop much more flexible survey plans, making it easier for the community of solar system observers to obtain time on the most oversubscribed telescopes such as the James Webb Space Telescope (JWST) or the Nancy Grace Roman Space Telescope (Roman). In this paper we introduce heliostack , an algorithm designed to enable nonlinear shift-and-stack searches for solar system objects. One of the key insights that enables heliostack to efficiently search through the space of nonlinear trajectories is inspired by the eponymous HelioLinC algorithm ( M. J. Holman et al. 2018 ) . Given detections of some solar system object, guesses of an object’s unmeasured orbital parameters can be used to fully define an orbit for each detection. If the guess was correct (or at least close enough), those orbits will cluster on the sky when propagated to a common epoch. In retrospect, the idea is obvious—no object can be in two places at once! We take this idea a step further in heliostack , applying the idea not just to catalogs of tracklets, but rather treating each pixel in each image as a

Read the complete text