Seattle, Washington, USA · He/Him 47.6062° N, 122.3321° W

Portrait of Joachim Moeyens

Joachim Moeyens

Astronomer and software engineer working on asteroid discovery and differentiable astrodynamics.

About

I build software that finds asteroids. For my doctorate, I developed Tracklet-less Heliocentric Orbit Recovery (THOR) with Professor Mario Jurić. THOR was designed to find asteroids in optical survey data not originally intended for asteroid discovery. Now, through the generous support of the Asteroid Institute, my main focus is extending THOR to find near-Earth objects and those that might one day impact Earth.

I also build empyrean, an astrodynamics toolkit for orbit propagation, ephemeris generation, and orbit determination. It is written in Rust, and every routine in the propagator is differentiable, so covariances come out of the same computation that produces the orbit.

Work on THOR has been featured in the New York Times and Wired. For my dissertation, I was awarded a Ph.D. Prize by the International Astronomical Union (IAU) in 2023.

Projects

thorAsteroid Institute · DiRAC

Tracklet-less Heliocentric Orbit Recovery. Classical linking expects a survey built for asteroid discovery: revisit the same patch two or three times a night, chain those detections into a tracklet, then link tracklets across nights. THOR removes that requirement by testing candidate orbits directly against single, unpaired detections, which makes discovery possible in archival and non-traditional survey data. Current work extends it to the near-Earth population and to Vera C. Rubin Observatory data.

Assume a test orbit, then project every night's detections into the frame co-rotating with it. An object near that orbit barely moves there, so its unpaired detections stack into a cluster; one slightly displaced traces a line. Clustering finds both, and each becomes an initial orbit to correct. Schematic — the geometry is illustrative, not a fitted solution.
Source
github.com/moeyensj/thor
Paper
Moeyens, Jurić, Ford, et al. 2021 · The Astronomical Journal
Recognition
IAU Ph.D. Prize, 2023 · covered by the New York Times, Wired and KUOW

ADAMAsteroid Institute · DiRAC

Asteroid Discovery, Analysis, and Mapping is the Asteroid Institute's cloud platform for astrodynamics at scale: discovery, precovery, orbit determination and impact probability pipelines on Google Cloud. My part has been taking THOR from a research code that ran on one machine to a search that runs across tens of thousands of CPUs, and building the libraries the platform runs on, adam_core among them. In 2024 that search found 27,500 asteroid discovery candidates in 5.4 billion archival NOIRLab observations, with no new observations of the sky.

THOR tests candidate orbits one at a time, and every test orbit is an independent job. On ADAM those jobs run side by side on Google Cloud, across tens of thousands of CPUs, which is how 5.4 billion archival NOIRLab observations became 27,500 discovery candidates in 2024. Schematic — the grid stands for 10,000 CPUs in blocks of 16, one block per test orbit; the fan is illustrative; the 2024 numbers are from the release.
Source
github.com/B612-Asteroid-Institute/adam_core · github.com/B612-Asteroid-Institute
Result
27,500 discovery candidates from NOIRLab Source Catalog DR2, April 2024, more than 100 of them near-Earth asteroids
Press
b612foundation.org · geekwire.com

empyrean

Every routine in the propagator is differentiable: force model terms, coordinate transforms and integrator steps each carry exact derivatives. Covariance propagation and orbit determination are then products of the same computation that produces the orbit, rather than a second pass over it. Event detection runs inside the propagation, and the uncertainty method climbs from a first-order covariance to a second-order map, an adaptive mixture or a sample where a close approach demands it. Written in Rust, with one engine binary behind a Python wheel, a Rust crate, a C shared library, and a command-line tool.

One propagation through an Earth close approach, with the 1σ region drawn the way each method carries it: the covariance mapped by the state transition matrix, the second-order correction that bends it, the thirteen sigma points of the unscented set, a Monte Carlo sample, and then the sample's mean and covariance, which is first order again. Around a close approach, Auto moves between first order, second order and an adaptive Gaussian mixture by itself; sigma points and Monte Carlo are chosen. Schematic — the geometry is illustrative, not a fitted solution.
Python
pip install empyrean
Rust
cargo add empyrean cargo install empyrean-cli
Site
empyrean-dynamics.com
Source
github.com/Empyrean-Dynamics/empyrean
Docs
docs.empyrean-dynamics.com · docs.rs/empyrean
Packages
pypi.org/project/empyrean · crates.io/crates/empyrean
Citation
doi:10.5281/zenodo.21318471
Release
v0.10.0, August 2026 · BSD-3-Clause source, proprietary binary

Affiliations

Asteroid Institute

A program of the B612 Foundation.

Research Scientist

DiRAC Institute

Data Intensive Research in Astrophysics & Cosmology.

Research Software Engineer

University of Washington

Department of Astronomy.

Ph.D. Astronomy, 2022

Research

THOR

Tracklet-less Heliocentric Orbit Recovery

A cadence-independent algorithm for linking single detections across nights into orbits, enabling discovery in archival and non-traditional survey data.

ADAM

Asteroid Discovery, Analysis, and Mapping

Cloud-hosted astrodynamics platform for running discovery, precovery, and impact probability pipelines at scale. Built by Asteroid Institute, a program of B612 Foundation.

ATM

Asteroid Thermal Modeling

Open-source thermal modeling for NEOWISE and similar infrared observations, used for size and albedo retrieval.

Rubin

Vera C. Rubin Observatory

Validating and testing the Observatory's Solar System Pipelines as part of commissioning.

Interests

Current threads

Building research-grade software with frontier AI models. Extending THOR to the near-Earth object population and preparing it for discovery on Vera C. Rubin Observatory data. Differentiable astrodynamics, which now ships as empyrean.

Software

thor Lead

Tracklet-less Heliocentric Orbit Recovery — discovers asteroids by linking single detections across nights, without pre-built tracklets.

empyrean Lead

Uncertainty-first orbit propagation, ephemeris generation, orbit determination and event detection for asteroids and comets, powered by automatic differentiation.

adam_core Core

Core astrodynamics primitives shared across the ADAM platform: orbits, ephemerides, coordinate frames and transforms.

difi Lead

Evaluates the findability of simulated small-body populations under a given survey cadence and discovery algorithm.

Publications

First author

  1. 2021
    THOR: An Algorithm for Cadence-independent Asteroid Discovery Moeyens, J., Jurić, M., Ford, J., et al. · The Astronomical Journal
  2. 2020
    ATM: An open-source tool for asteroid thermal modeling and its application to NEOWISE data Moeyens, J., Myhrvold, N., Ivezić, Ž. · Icarus

Co-author

  1. 2026
    Predictions of Imminent Earth Impactors Discovered by LSST Chow, I., Jurić, M., Jones, R. L., et al. · The Astrophysical Journal
  2. 2026
    Predictions of the LSST Solar System Yield: Neptune Trojans Murtagh, J., Schwamb, M. E., Bernardinelli, P. H., et al. · The Astronomical Journal
  3. 2026
    Lightcurves, Rotation Periods, and Colors for Vera C. Rubin Observatory's First Asteroid Discoveries Greenstreet, S., Li, Z., Vavilov, D. E., et al. · The Astrophysical Journal Letters
  4. 2025
    Sorcha: A Solar System Survey Simulator for the Legacy Survey of Space and Time Merritt, S. R., Fedorets, G., Schwamb, M. E., et al. · The Astronomical Journal
  5. 2025
    Predictions of the LSST Solar System Yield: Near-Earth Objects, Main Belt Asteroids, Jupiter Trojans, and Trans-Neptunian Objects Kurlander, J. A., Bernardinelli, P. H., Schwamb, M. E., et al. · The Astronomical Journal
  6. 2025
    Predictions of the LSST Solar System Yield: Discovery Rates and Characterizations of Centaurs Murtagh, J., Schwamb, M. E., Merritt, S. R., et al. · The Astronomical Journal
  7. 2025
    Expected Impact of Rubin Observatory LSST on NEO Follow-up Wagg, T., Jurić, M., Yoachim, P., et al. · The Astronomical Journal
  8. 2022
    Simulated SPHEREx spectra of asteroids and their implications for asteroid size and reflectance estimation Ivezić, Ž., Ivezić, V., Moeyens, J., et al. · Icarus
  9. 2021
    Sifting through the Static: Moving Object Detection in Difference Images Smotherman, H., Connolly, A. J., Kalmbach, J. B., et al. · The Astronomical Journal
  10. 2021
    Generalization of a method by Mossotti for initial orbit determination Gronchi, G. F., Baù, G., Rodríguez, Ó., et al. · Celestial Mechanics and Dynamical Astronomy
  11. 2019
    LSST: From Science Drivers to Reference Design and Anticipated Data Products Ivezić, Ž., Kahn, S. M., Tyson, J. A., et al. · The Astrophysical Journal
  12. 2019
    Fast Algorithms for Slow Moving Asteroids: Constraints on the Distribution of Kuiper Belt Objects Whidden, P. J., Connolly, A. J., Kalmbach, J. B., et al. · The Astronomical Journal
  13. 2019
    The Large Synoptic Survey Telescope as a Near-Earth Object discovery machine Jones, R. L., Slater, C., Moeyens, J., et al. · Icarus
  14. 2018
    APO Time-resolved Color Photometry of Highly Elongated Interstellar Object 1I/‘Oumuamua Bolin, B. T., Weaver, H. A., Fernandez, Y. R., et al. · The Astrophysical Journal Letters

Full list on ADS · ORCID · Google Scholar

Talks, Presentations & Posters

  1. 2025
    The Discovery of Potentially Hazardous Asteroids in the Era of the Vera C. Rubin Observatory University of Washington · Astrobiology Colloquium · Seattle, WA, USA · Slides
  2. 2024
    The Characterization and Discovery of Solar System Small Bodies in Modern Astronomical Surveys IAU XXXII General Assembly · Cape Town, South Africa · Division B Ph.D. Prize talk
  3. 2023
    ADAM / THOR LINCC Tech Talks · with Spencer Nelson · Recording
  4. 2023
    Tracklet-less Discovery in the Cloud Asteroids, Comets, and Meteors · Flagstaff, AZ, USA · Poster
  5. 2023
    Real-time Small Body Precovery Searches in Any Astronomical Dataset Planetary Defense Conference · Vienna, Austria · Poster
  6. 2022
    B612 Foundation and Their Search for Asteroids National Space Society · North Texas · Remote · Slides
  7. 2021
    THOR: Searching for Asteroids Without Tracklets Divion of Planetary Science · Remote · Slides
  8. 2021
    ATM: An Open-Source Tool For Asteroid Thermal Modeling Planetary Defense Conference · Remote · Slides
  9. 2019
    THOR: An Algorithm for Cadence-independent Asteroid Discovery EPSC-DPS 2019 · Geneva, Switzerland · Slides
  10. 2017
    LSST Moving Object Processing System (MOPS) NASA Frontier Development Lab (FDL) · Slides

Press

  1. 2024
    Seattle researchers are hunting for asteroids with the power of THOR Libby Denkmann · Soundside, KUOW
  2. 2022
    Killer Asteroids Are Hiding in Plain Sight. A New Tool Helps Spot Them. Kenneth Chang · The New York Times
  3. 2022
    The First Privately Funded Killer Asteroid Spotter Is Here Ramin Skibba · Wired
  4. 2017
    For Modern Astronomers, It's Learn to Code or Get Left Behind Sarah Scoles · Wired

Awards & Fellowships

  1. 2023
    IAU Ph.D. Prize — Division B (Facilities, Technologies & Data Science) International Astronomical Union · for the doctoral thesis
  2. 2021
    Graduate Research Award Department of Astronomy, University of Washington · for outstanding achievement in research, 2020–2021 (THOR)
  3. 2020–2022
    Asteroid Institute Graduate Student Fellow Ph.D. research on THOR funded in part by donors to the Asteroid Institute.
  4. 2017–2022
    DiRAC Graduate Student Fellow Invited to join the Data Intensive Research in Astrophysics & Cosmology (DiRAC) Institute at the University of Washington. DiRAC is a center focused on interdisciplinary research with the goal of providing scientists and researchers the knowledge and expertise to tackle the ever-growing volume of astronomical datasets.
  5. 2016–2018
    LSSTC Data Science Fellow Inaugural class member of a two year fellowship featuring six week-long sessions focusing on training data science skills such as programming, machine learning, data engineering, statistics, and open-source software development.

Experience

Research Software Engineer

DiRAC Institute, University of Washington

Joint Research Scientist appointment at the Asteroid Institute (B612 Foundation). Developing THOR and the ADAM platform.

2023 – present

Graduate Student

Department of Astronomy, University of Washington

Research-focused graduate studies yielding THOR and ATM. Led discussion, quiz, and lab sections for ~60 students across three academic quarters.

2016 – 2022

LSST Research Assistant

University of Washington

Analyzed LSST's Moving Object Pipeline System (MOPS) performance for NEO detection with Prof. Mario Jurić and Dr. Lynne Jones.

2015 – 2016

Student Researcher & Research Intern

University of Washington

Constrained atmospheric extinction using stellar magnitudes with Prof. Željko Ivezić.

2014 – 2015

Education

Ph.D., Astronomy

University of Washington · Certificate in Advanced Data Science

Thesis: The Characterization and Discovery of Solar System Small Bodies in Modern Astronomical Surveys

2018 – 2022

M.S., Astronomy

University of Washington

2016 – 2018

B.S., Applied Physics & Astronomy

University of Washington · with Honors

2010 – 2015

Contact

Email
moeyensj@uw.edu
GitHub
github.com/moeyensj
ORCID
0000-0001-5820-3925
LinkedIn
linkedin.com/in/joachim-moeyens