The Atacama Cosmology Telescope: Release of A databaSe of millimeTeR ObservatioNs of Asteroids Using acT (ASTRONAUT)

We present A databaSe of millimeTeR ObservatioNs of Asteroids Using acT (ASTRONAUT) hosted on Amazon Web Services, Inc. (AWS) in the form of a public Amazon Simple Storage Service (S3) bucket. This bucket is an Amazon cloud storage database containing flux measurements for a group of asteroids at millimeter (mm) wavelengths. These measurements were collected by the Atacama Cosmology Telescope (ACT) from 2017 to 2021 in frequency bands centered near 90, 150, and 220 GHz. The ASTRONAUT database contains observation times, normalized flux values, and associated error bars for 170 asteroids above a signal-to-noise ratio of 5 for a single frequency band over the stacked co-added maps. We provide an example in generating light curves with this database. We also present a Jupyter notebook to serve as a reference guide when using the S3 bucket. The container and notebook are publicly available in a GitHub repository.

💡 Research Summary

The paper introduces ASTRONAUT, a publicly accessible database of millimeter‑wave observations of asteroids obtained with the Atacama Cosmology Telescope (ACT) between 2017 and 2021. ACT, originally designed for Cosmic Microwave Background (CMB) studies, repeatedly scanned large fractions of the sky, inadvertently capturing thermal emission from a substantial number of Solar System bodies. By exploiting this serendipitous coverage, the authors extracted flux measurements for 170 asteroids that achieve a signal‑to‑noise ratio of at least 5 in at least one of the three ACT frequency bands centered near 90 GHz, 150 GHz, and 220 GHz (more precisely 98, 150, and 228 GHz).

The raw data consist of “depth‑1” maps, which are essentially single‑pass, ~5‑minute exposure images produced for each detector array (pa4, pa5, pa6). These maps are matched‑filtered to enhance point‑source detection, thereby maximizing the S/N for asteroid signals. Using the JPL Horizons API, the precise ephemerides of each asteroid at the exact observation times are computed, allowing the authors to locate the asteroid on each depth‑1 map and read out the filtered flux value at that pixel. The associated uncertainty is derived from a flux‑uncertainty map generated during the ACT data‑reduction pipeline.

To enable meaningful comparisons across different observing epochs, detector arrays, and viewing geometries, the authors normalize each measured flux to a standard reference using the Rayleigh‑Jeans approximation of the Standard Thermal Model (STM). The normalization accounts for the Earth‑asteroid distance (d_earth), Sun‑asteroid distance (d_sun), and the Sun‑asteroid‑Earth phase angle (α). The formula applied is:
 F₀ = F_i · (d_earth,i · 1 AU)² · (d_sun,i · 1 AU)^{½} · 10^{0.004 α_i} ,
where F_i is the raw flux, and the factor 10^{0.004 α_i} approximates the phase‑angle dependence of thermal emission. The same scaling is applied to the flux uncertainties, and a geometry‑based weighting factor W_i is also stored for each measurement.

All processed products—normalized flux (F₀), flux error, weighting factor, and observation time (Unix timestamp)—are packaged as FITS files named in the pattern name_lc_arr_freq.fits, where the user specifies the asteroid name, ACT array, and frequency. These files reside in a public Amazon S3 bucket that is registered in the AWS Open Data Registry, allowing unrestricted programmatic access via the boto3 Python library or any S3‑compatible client. Invalid array‑frequency combinations (e.g., requesting 220 GHz from array pa6) trigger a ClientError, ensuring users receive immediate feedback on malformed queries. For bulk users, the same data are mirrored on the Legacy Archive for Microwave Background Data Analysis (LAMBDA) website as tar archives, facilitating offline analysis on high‑performance computing resources.

To lower the barrier to entry, the authors provide a comprehensive Jupyter Notebook tutorial hosted on a dedicated GitHub repository. The notebook walks users through installing required Python packages (listed in requirements.txt), constructing S3 queries, loading the FITS files, and generating light curves. An example light curve for asteroid (705) Erminia at 90 GHz is reproduced in the notebook, demonstrating that the workflow can be replicated with minimal effort. The notebook also includes helper functions for searching the asteroid list, filtering by array and frequency, and plotting multi‑frequency light curves.

Scientifically, millimeter‑wave thermal emission probes the top few millimeters to centimeters of an asteroid’s regolith, a regime that is difficult to access with optical or infrared observations. While the Rayleigh‑Jeans approximation predicts a nearly black‑body spectrum, deviations arise from variations in surface composition, grain size, porosity, and temperature gradients. By providing a homogeneous, multi‑epoch, multi‑frequency dataset, ASTRONAUT enables statistical studies of these subtle effects across a sizable asteroid sample. Potential applications include: (1) constructing broadband spectral energy distributions by combining ACT data with measurements from other CMB experiments such as the South Pole Telescope (SPT) and the Simons Observatory; (2) deriving phase curves to infer rotation periods and surface heterogeneity; (3) calibrating thermophysical models that link millimeter flux to regolith properties; and (4) cross‑validating with high‑resolution observations from facilities like ALMA.

The paper concludes that ASTRONAUT represents a significant step toward open, reproducible science in the niche field of millimeter asteroid observations. By leveraging existing CMB survey data, providing cloud‑based storage, and offering ready‑to‑use analysis tools, the authors lower the entry threshold for planetary scientists and encourage broader exploitation of these data. Future work may expand the catalog to include additional years of ACT data, incorporate observations from other telescopes, and develop community‑driven pipelines for thermophysical modeling and joint multi‑instrument analyses.