Taxonomy of 14042 asteroids from Gaia DR3 reflectance spectra PREPRINT

F. Tinaut-Ruano, B. Carry, M. Galinier, M. Mahlke, A. Sergeyev · 2026

Read Taxonomy of 14042 asteroids from Gaia DR3 reflectance spectra PREPRINT by F. Tinaut-Ruano, B. Carry, M. Galinier, M. Mahlke, A. Sergeyev (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-2605-06838.txt.

Introduction

Fernando Tinaut-Ruano and four co-authors classify over fourteen thousand asteroids by their reflected sunlight spectra from the Gaia mission's third data release. Spectra constrain surface composition, and Gaia observed more than sixty thousand asteroids, but instrumental systematics make direct comparison with older taxonomies risky. The team built a scheme tailored to Gaia, then linked its classes back to established ones.

Strict quality cuts and albedo information shrink the sample before clustering in reduced dimensions. The final taxonomy has thirteen classes, three more objects classes than previous spectral efforts covered, with the largest relative gain in the K class. Near-ultraviolet coverage separates primitive types inside the carbon-rich complex. Geographic trends match expectations across the main belt.

Catalog record

CREATORF. Tinaut-Ruano, B. Carry, M. Galinier, M. Mahlke, A. Sergeyev
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.06838

Browse the full library · About the Tower of Babel

Full text

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

Taxonomy of 14042 asteroids from Gaia DR3 reflectance spectra PREPRINT F. Tinaut-Ruano, B. Carry, M. Galinier, M. Mahlke, A. Sergeyev Abstract Context. Asteroid reflectance spectra provide key constraints on surface composition. Gaia Data Release 3 (DR3) enables the study of 60,518 asteroids through near-ultraviolet (NUV) to visible reflectance spectra. Aims. We aim to classify asteroids using Gaia DR3 spectra and provide a homogeneous taxonomical framework. Owing to systematics affecting Gaia DR3 data, direct comparison with previous taxonomies has to be taken with caution; therefore, we developed a classification scheme tailored to Gaia and linked the resulting taxa to established classes. Methods. We selected the highest-quality spectra using Gaia DR3 quality flags and apply uncertainty thresholds to mitigate spectral artifacts, retaining over one third of the original sample at the least noisy wavelengths. To improve compositional discrimination, we included albedo information, reducing the final sample to about one fourth of its initial size. We then iteratively applied dimensionality reduction and clustering techniques to identify the spectral taxa. Results. We classified 14,042 asteroids into 13 taxonomic classes: A, B, C, D, E, F, G, K, L, M, P, S, and V, representing an increase of three compared to the number of objects classified in previous spectral classifications. The largest relative increase is found for the K class. The inclusion of NUV wavelengths allows the separation of B and F types within the C-complex and facilitates the identification of G types. The dynamical distribution follows expected trends, with S types dominating the inner and middle Main Belt, C-complex asteroids prevalent in the outer Main Belt, and D types beyond. Conclusions. We present a taxonomical classification of 14,042 asteroids based on Gaia DR3 reflectance spectra. NUV coverage is critical for disentangling primitive classes within the C-complex. Although artifacts in Gaia DR3 require caution when comparing median spectra with other datasets, this classification provides a robust reference for future Gaia releases, with larger observed samples. 1 Introduction Asteroid reflectance spectra and spectrophotometry provide information on the composition of asteroid surfaces, and on the processes that modify their properties, such as space weathering ( 78 ) . Historically, the use of photoelectric detectors (or photometers), more sensitive at blue wavelengths (i.e., ¡ 0.5 \mu m), and the development of the standard UBV photometric system ( 49 ) , led to the appearance of the first asteroid taxonomies in the 1970s ( 106 ; 17 ) which contained information at blue-visible wavelengths (near-UV or NUV hereafter, from 0.3 to 0.5 \mu m). The introduction of charge-coupled devices (CCDs) in astronomy in the 1990s and later on, contributed to the “loss” of NUV information, as CCDs are less sensitive at those wavelengths. Therefore, the large majority of the modern spectroscopic and spectrophotometric surveys cover the wavelength range from \sim 0.5 \mu m up to 2.5 \mu m. Nevertheless, there are some exceptions. One of the first large surveys with information in the NUV is the Eight Color Asteroid Survey (ECAS, 105 ). They provided photometry in eight broad band filters between 0.34 to 1.04 \mu m for 589 minor planets, including two filters below 0.45 \mu m. These observations were used to develop a new taxonomy ( 92 , see) . More recent catalogues, like the Sloan Digital Sky Survey (SDSS) Solar System object catalogue ( 84 ) , or the Solar System Objects observations from the SkyMapper Southern Survey ( 83 ) include NUV photometry for 800,000 and 205,515 objects, respectively. However the spectral resolution in this surveys is limited having five and six filters between 0.3 and 1.1 \mu m. On the other hand, the Moving Objects Observed from Javalambre (MOOJa) catalogue from the J-PLUS survey ( 69 ) , has a larger spectral resolution of 12 filters but a limited number of asteroids observed: 3122 . The Gaia space mission ( 77 ) offers in its Data Release 3 (DR3) a Solar system survey with different data products ( 87 ) , including the reflectance spectra of 60,518 objects binned in 16 wavelengths bands between 0.352 and 1.056 \mu m ( 34 ) . It offers a unique opportunity to explore a large amount of asteroids with a moderate spectral resolution and avoiding the atmospheric absorption in the UV. Some systematic effects are, however, present in the Gaia DR3 spectra. These systematics are: (i) The selection of solar analogs by the Gaia team included a large number of the most commonly used stars by the planetary science community ( 34 ) . Such stars can be used to get asteroid reflectance spectra beyond 0.5 \mu m without any problems. However, as it has been shown in 88 and 95 , they do not have a Sun-like spectral behavior at wavelengths below 0.5 \mu m, introducing an artificial reddening in that region. (ii) Gaia reflectance spectra show a reddening between 0.7 and 1 \mu m as different authors found when comparing Gaia DR3 with other ground-based data sources ( 7 ; 31 ; 68 ) , in the first work, Balossi found that this effect is taxonomy dependent. (iii) The three bluest and reddest reflectance points in Gaia DR3 spectra are often overestimated as 34 described. (iv) An artificial band can be found around 0.65 \mu m, due to the merging of the blue and red parts of the spectra in the production of the mean reflectance spectra ( 34 ; 31 ; 94 ) . (v) For the V-types, a shift in band center and band depth is detected by 75 . Comparisons to other spectral databases has to be done with caution due to those artifacts, making the classification of their spectra more difficult. Thus, we decided to classify the reflectance spectra of asteroids in Gaia DR3 by following the steps of building a new taxonomy based on Gaia spectra and make it compatible with previous taxonomies. In this article, we aim to classify a large fraction of the Gaia DR3 asteroids’ reflectance spectra catalog through variance driven dimensional reduction and clustering methods. The base dataset, the cleaning and other methods are presented in Section 2 . The iterative process and the decision tree are extensively described in section Section 3 . We show the results in Section 4 . Finally, the conclusions are presented in Section 5 . 2 Methodology We describe here the selection of the data set, the tools we used, and the iterative clustering algorithm. 2.1 Selection of the data set To build a taxonomy through variance driven dimensional reduction and clustering methods, we needed to use a sample whose variance is dominated by the mineralogical properties of the asteroids, and not by the noise or systematics present in the data. In this section we show how we selected the best quality data from the Gaia DR3 reflectance spectra catalog. 2.1.1 Gaia reflectance spectra Among the Gaia DR3 products for Solar System Objects (SSOs), the mean reflectance spectra of Solar system small objects provided in the Gaia DR3 span the wavelength range from 374 to 1034 nm in 16 bands, and are normalised at 550 nm. To each band was assigned a quality flag, indicative of the estimated quality of the band: 0 for good quality, 1 for lower quality

Read the complete text