Open Access
Issue
A&A
Volume 693, January 2025
Article Number A140
Number of page(s) 27
Section Planets, planetary systems, and small bodies
DOI https://doi.org/10.1051/0004-6361/202450662
Published online 15 January 2025

© The Authors 2025

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

Impacts are commonly occurring events in the main asteroid belt spanning from the past to the present. Occasionally, a large impactor collides with a planetesimal, resulting in an energetic impact event that either partially or completely destroys the planetesimal, breaking it into smaller pieces. These resulting fragments are referred to as a collisional family. The Hirayama family, first identified by Hirayama (1918), stands as the earliest recognized collisional family. The collisional family is a part of the dynamical families where members share similar orbital proper elements. Numerous efforts have been made to identify asteroid dynamical families (Zappala et al. 1990; Nesvornỳ et al. 2005; Nesvorný et al. 2015; Milani & Gronchi 2010; Milani et al. 2014). The dynamical families are solely defined by their proper orbital elements, and in some cases multiple collisional families can combine to form a larger dynamical one, as is seen in the Nysa–Polana complex family. Consequently, collisional families may require additional constraints such as albedo or spectral characteristics for accurate identification.

The internal structure of primitive primordial bodies, which serve as the parent bodies of asteroid families, remains poorly understood. An onion-shell structure of these primordial bodies, attributed to an internal temperature difference induced by the short-lived radionuclide 26Al, has been hypothesized (Grimm & McSween Jr 1989). Recently, the Dawn spacecraft provided a unique case study of such a primordial body. The dwarf planet Ceres may not have experienced catastrophic disruption, as is suggested by the absence of an apparent collisional family (Tinaut-Ruano et al. 2022), thereby potentially preserving its original internal structure since formation. The low density of 2.16 g/cm3 and the geological evidence supports the presence of differentiation and subsurface liquids (e.g., Neveu & Desch 2015; Marchi et al. 2016; Park et al. 2019), possibly consistent with the onion-shell model (Castillo-Rogez & McCord 2010; Neumann et al. 2020). However, ground-based observations of asteroids within families often show spectral homogeneity (Tholen 1984; Cellino et al. 2002; Mothé-Diniz et al. 2005; de León et al. 2016), implying minimal compositional variation in a family. This contradicts the onion-shell hypothesis, which predicts more internal compositional diversity. Additionally, recent in situ investigations of the rubble-pile asteroids (162173) Ryugu and (101955) Bennu by the spacecraft Hayabusa2 and OSIRIS-REx, respectively, have revealed homogeneous near-infrared (NIR) spectra on their surface (Kitazato et al. 2019; Simon et al. 2020). This suggests compositional homogeneity in their parent bodies, which have gone through catastrophic disruption and re-accumulation of their fragments. Furthermore, some observational findings of minor bright boulders indicate the presence of a small portion (<1%) of the materials with varying degrees of thermal metamorphism on the surface of Ryugu (Tatsumi et al. 2021; Sugimoto et al. 2021a,b). In addition to these observations, recent model calculations propose the possibility of large-scale mud convection inside the primordial bodies, resulting in a homogeneous internal temperature (Bland & Travis 2017). Thus, the asteroid collisional families provide us with a unique opportunity to investigate the internal compositional distribution of the primordial bodies through the fragmented member asteroids.

In our companion paper (Tatsumi et al. 2023), we explore the relationship between phyllosilicates and near-ultraviolet (NUV) absorption based on the asteroid observations and laboratory spectra of meteorites. We find a strong correlation between the intensity of the 2.7-μm absorption attributed to hydroxyl presence in minerals and the NUV absorption spectral feature. Furthermore, the asteroids exhibiting a 0.7 μm band demonstrate deeper NUV absorption compared to those lacking 0.7 μm absorption, suggesting that the NUV feature is particularly sensitive to interlayer iron within phyllosilicates. This is also supported by the recent Gaia dataset (Tinaut-Ruano et al. 2024). Building upon these findings, we propose measuring NUV absorption below 0.5 μm as an indicator of phyllosilicates, especially Fe-rich phyllosilicates (Tatsumi et al. 2023). In this study, we aim to apply this idea to investigate the phyllosilicate distribution of asteroid families and their primordial bodies. We investigate the spectral characteristics of asteroid families from NUV to NIR using several spectrophotometric datasets, thereby shedding light on their internal homogeneity or heterogeneity.

2 Dataset

Two multiband photometric surveys, the Sloan Digital Sky Survey (SDSS) (Sergeyev & Carry 2021; Ivezić et al. 2001) and the Eight Color Asteroid Survey (ECAS) (Zellner et al. 1985), have been conducted down to wavelengths below 0.4 μm. These surveys complement each other in terms of the absolute magnitude ℋ of the observed objects; SDSS primarily covers objects with 8 < ℋ < 20, while ECAS focuses on objects with ℋ < 10. Therefore, we utilized these two datasets to encompass a broad size range of asteroids. Furthermore, to have wider wavelength information to characterize the asteroid families, we also used NIR spectrophotometry dataset, MOVIS-C (Popescu et al. 2018). In the same way as Tatsumi et al. (2023), we used albedo values from AcuA dataset by AKARI (Usui et al. 2013) and the NEOWISE dataset (Mainzer et al. 2019).

In this study, we used the dynamical families identified by Nesvorný et al. (2015), with the dataset of Nesvorny (2020) accessible via Planetary Data System Small Bodies Node. Although several studies have used SDSS colors to study taxonomic types of asteroid families (Parker et al. 2008; Carvano et al. 2010; DeMeo & Carry 2013), they made their classifications consistent with Bus/Bus-DeMeo taxonomic systems (Bus & Binzel 2002a; DeMeo et al. 2009) that did not include NUV data in their definitions. We used the latest and complete SDSS color catalog, and the shortest wavelength filter “u” to further classify the primitive asteroids based on Tholen’s taxonomy (Tholen 1984).

2.1 SDSS

The SDSS acquired photometric data using five filters, u (0.3557 μm), g (0.4825 μm), r (0.6261 μm), i (0.7672 μm), and z (0.9097 μm) (Fukugita et al. 1996). As the u filter photometry especially has large errors, we selected the data using the errors and the photometry flag noted by Sergeyev & Carry (2021). We used good-quality data with errors smaller than 0.1 mag. The photometric spectra were normalized at the g filter, and the NUV and VIS spectral slopes were calculated using the following equations: SNUV=100.4(gu(gu))1λSDSS,gλSDSS,u,$\[S_{\mathrm{NUV}} =\frac{10^{-0.4\left(g-u-(g-u)_{\odot}\right)}-1}{\lambda_{\mathrm{SDSS}, g}-\lambda_{\mathrm{SDSS}, u}},\]$(1) SVIS=100.4(ig(ig))1λSDSS,iλSDSS,g,$\[S_{\mathrm{VIS}}=\frac{10^{-0.4\left(i-g-(i-g)_{\odot}\right)}-1}{\lambda_{\mathrm{SDSS}, i}-\lambda_{\mathrm{SDSS}, g}},\]$(2)

where λ represents the effective filter wavelength, and the solar colors are (ig) = −0.57, (gu) = −1.40, as was reported by Holmberg et al. (2006). We selected the primitive asteroids with the same criteria in Tatsumi et al. (2023): the bluer group (gr) < 0.55 and (iz) > −0.15 for the C complex, and the redder group (gr) ≥ 0.55 and pV < 0.11. In addition to the data selection criteria described by Tatsumi et al. (2023), we removed the outlier spectra that have extremely large reflectance (>1.5 for u, r and i filters, and >1.7 for z filter) and small reflectance (<0.2 for u filter, and <0.7 for r, i, and z filters) when they are normalized at the g filter.

2.2 ECAS

The Eight Color Asteroid Survey (ECAS) is a photometric survey of 589 asteroids, utilizing eight filters: s (0.337 μm), u (0.359 μm), b (0.437 μm), v (0.550 μm), w (0.701 μm), x (0.853 μm), p (0.948 μm), and z (1.041 μm) (Zellner et al. 1985). Color indices of asteroids were derived to yield mean color indices of zero for four well-characterized solar analogs (Tedesco et al. 1982). Consequently, the spectral reflectance, Rλ, can be obtained by log(Rλ) = ±0.4cλ, where cλ represents the tabulated color index with the negative sign chosen for wavelengths shorter than the v band. Tholen (1984) developed the taxonomic classification based on cluster analysis in the principal component space applied to the ECAS dataset with known geometric albedo. In our study, we focus on primitive asteroid families between 2 to 4 au, classified into C-complex, P (low-albedo X), and D classes based on Tholen’s taxonomy. We calculated the reflectance at the wavelengths corresponding to those of SDSS by interpolating two reflectance values at the nearest wavelengths from the ECAS filters. After interpolation of spectra at the SDSS wavelengths, spectral slopes were computed by Eqs (1) and (2) as well. We defined the NUV absorption strength as the difference of the slopes: SNUVSVIS. Additionally, we measured the indication of a 0.7 μm band absorption, in the following hydration parameter (HYD): HYDSDSS=12RiRr+Rz.$\[\mathrm{HYD}_{\mathrm{SDSS}}=1-\frac{2 R_{i}}{R_{r}+R_{z}}.\]$(3)

Positive values of the HYD may indicate stronger absorption at 0.7 μm.

2.3 MOVIS-C

The MOVIS catalogs (Popescu et al. 2016, 2018) represent the spectrophotometric survey of Solar System objects conducted through the VISTA-VHS (Visible and Infrared Survey Telescope for Astronomy – VISTA Hemisphere Survey) program (McMahon et al. 2013; Sutherland et al. 2015). This survey covered the entire sky observed from the southern hemisphere using the NIR broad-band filters Y (1.020 μm), J (1.252 μm), H (1.645 μm), and Ks (2.147 μm). The corresponding color cat alog is known as MOVIS-C. In this study, we used the latest version of the MOVIS-C catalog (Popescu et al. 2018) to complement the NIR information for the asteroid families. We computed the near-infrared spectral slopes between YJ (1.020–1.252 μm) and JKs (1.252–2.147 μm) as follows: SNIR1=100.4(YJ(YJ))1λVISTA,YλVISTA,J,$\[S_{\mathrm{NIR} 1}=\frac{10^{0.4\left(Y-J-(Y-J)_{\odot}\right)}-1}{\lambda_{\mathrm{VISTA}, Y}-\lambda_{\mathrm{VISTA}, J}},\]$(4) SNIR2=100.4(JKs(JKs))1λVISTA,JλVISTA,Ks,$\[S_{\mathrm{NIR} 2}=\frac{10^{0.4\left(J-K_{s}-\left(J-K_{s}\right)_{\odot}\right)}-1}{\lambda_{\mathrm{VISTA}, J}-\lambda_{\mathrm{VISTA}, K_{s}}},\]$(5)

where the sun color for VISTA is (YJ) = 0.219, (JKs) = 0.336 (Popescu et al. 2018; Casagrande et al. 2012). In Table B.1, we calculated the median and the interquartile range when the family has more than four members from the MOVIS-C catalog. It is worth noting that the concavity of the NIR spectra can be assessed by the difference between two slopes: SNIR2SNIR1; the value indicates a concave shape and the negative value indicates a convex shape.

3 Correlation analysis

The correlation analysis using a Monte Carlo simulation was conducted to measure relationships between variables; that is, NUV absorption, the VIS slope, NIR slopes, and HYDs. We measured the correlation with a nonparametric metric: the Spearman correlation coefficient. The Spearman correlation coefficient does not assume an underlying relationship between two variables, but instead produces a generic measure of positive or negative correlation. The Spearman correlation coefficient, ρ, can take a value between +1 to −1, where +1 or −1 indicates that each of the variables is a perfect monotone function of the other with a positive or negative correlation, respectively. The Spearman correlation coefficient is calculated with the following equation: ρ=16d2N3N,$\[\rho=1-\frac{6 \sum d^{2}}{N^{3}-N},\]$(6)

where d is the difference between the two ranks of each observation, and N is the number of observations. As our analysis has uncertainties of variables to some extent, we need to account for these variations. One simple way to do this is a Monte Carlo simulation. We assumed a random distribution for the variables and simulated 10 000 instances of our dataset and calculated the coefficient for each instance. Finally, we obtained the mean value and standard deviation for the Spearman correlation coefficient.

thumbnail Fig. 1

Spectral classifications used in this study following Tholen’s taxonomy. The classification criteria is described in Tatsumi et al. (2023).

4 Results

The taxonomic classification was done based on the criteria using the visible spectral slope, SVIS, and the NUV absorption, SNUVSVIS (Table 3 in Tatsumi et al. 2023). This criterion follows Tholen’s classification (Tholen 1984): B, C, F, G, P, and D types. The albedo and spectral slopes for each of Tholen’s taxonomic classes are presented in Table 1 in Tatsumi et al. (2023). The B type is the bluest endmember and the D type is the reddest. The F type exhibits the shallowest NUV absorption, while the G type the deepest NUV absorption (Fig. 1). G-type asteroids tend to have the 0.7-μm absorption (Tatsumi et al. 2023). Although we did not use albedo to classify the asteroids, the B type has the highest mean albedo value (Tatsumi et al. 2023).

4.1 NUV absorption versus 0.7-μm absorption, NIR slope, and albedo

Table B.1 presents the numbers of observed asteroids in both photometric surveys across various collisional families. We excluded families with fewer than seven members. Median values of the NUV and VIS slopes, the NUV absorption strength, and the HYD were calculated for each family. Some of these families have undergone spectroscopic studies, allowing us to evaluate the fraction of members exhibiting 0.7 μm band absorption, as has been documented in various articles (Mothé-Diniz et al. 2005; Morate et al. 2016, 2018, 2019, 2021; de León et al. 2016; De Prá et al. 2020b; De Prá et al. 2020a). The fraction of family members exhibiting the 0.7 μm absorption feature, Qhyd, was calculated based on previous spectroscopic studies (the references are indicated in Table B.1). A relationship between HYD and Qhyd was investigated (Fig. 2). The families with positive HYD values show that over 20% of their members display the 0.7 μm band absorption. This strongly indicates the presence of Fe-bearing phyllosilicates within the family members. In the high HYD group, the Astrid and Klio families exhibit relatively low percentage ~20% of their members with the 0.7 μm absorption band. We note that only five members of the Astrid family were spectroscopically observed (Mothé-Diniz et al. 2005). However, examination of the spectra used from SMASSII (Bus & Binzel 2002b), revealed that three out of four spectra classified as C type in their paper exhibit shallow 0.7 μm bands upon visual inspection. This highlights the importance of careful evaluation in determining the presence or absence of this feature, and a standardized evaluation method is required, as is proposed by Morate et al. (2016). This correlation between HYD and the 0.7 μm absorption suggests that the HYD value can serve as a discriminator for the hydration state of other families; that is, we classified a family with positive HYD as a highly hydrated family. The Spearman correlation coefficient between the NUV absorption strength and HYD for the families with more than 20 members is ρ = 0.48 ± 0.08. This correlation between the median values of HYD and the NUV absorption strength of families is observed in Fig. 3, indicating that NUV absorption is also attributed to the intervalence charge transfer (IVCT) of iron in Fe-bearing phyllosilicates.

We observed a potential correlation between NUV absorption and NIR slope within the range of 1.25–2.15 μm, although the large error bars resulted in a significant overlap of values (Fig. 4). The Spearman correlation coefficient is ρ = −0.47 ± 0.12, suggesting a negative correlation. This correlation may be attributed to changes in spectral slope occurring when the material is composed of hydrated silicates. Laboratory measurements show that the highly hydrated meteorites exhibit a shallower slope in their reflectance spectra just before the strong 2.7-μm absorption (e.g., Potin et al. 2020), resulting in convex (concave-down) NIR spectra. Ziffer et al. (2011) first reported two distinct shapes in NIR, concave (concave-up) shapes for the Themis family and convex (concave-down) for the Veritas family. The Veritas family has the highest NUV absorption in our analysis, while the Themis family has a moderate NUV absorption value. Thus, our analysis found a similar tendency to Ziffer et al. (2011) but with more samples. Further investigation of this phenomenon is warranted in future studies.

We found a correlation between albedo and spectral characteristics of asteroid families. Figure 5 illustrates the relationship between albedo and the VIS spectral slope, as well as albedo and the NUV absorption. Notably, the Pallas family exhibits the highest albedo, standing apart from the main cluster, indicating its peculiarity. We removed the Pallas family from the statistical analysis. The Spearman correlation coefficients for NUV absorption and VIS slope are ρ = 0.42 ± 0.08 and ρ = −0.31 ± 0.08, respectively. Within the main cluster, higher albedo values correspond to a bluer VIS slope and increased NUV absorption. This anticorrelation between the VIS slope and albedo may reflect the variations in organic abundance among dark asteroids (Alexander et al. 2012; Beck et al. 2021). Additionally, in the bottom figure, two subclusters are evident: one with negative HYD values and another with positive values. Families with positive HYD values tend to exhibit slightly higher NUV absorptions. Interestingly, the albedo ranges of the two subclusters are similar, consistent with findings from the AKARI survey (Usui et al. 2013), which highlighted that the smaller asteroids with low NUV absorption, (F, P, and D types) are among the darkest within the carbonaceous type, whereas those with moderate to high absorption, (B, C, and G types) are brighter.

thumbnail Fig. 2

Comparison between fraction of family members with the 0.7 μm band absorption (Qhyd) from spectroscopic studies (Mothé-Diniz et al. 2005; Morate et al. 2016, 2018, 2019, 2021; de León et al. 2016; De Prá et al. 2020b; De Prá et al. 2020a) and the HYD value calculated from the SDSS data. Filled circles indicate more than 20 family members and open circles indicate fewer than 20 family members.

thumbnail Fig. 3

Median NUV absorption strengths vs. median HYD values for the primitive asteroid families (gray markers indicate fewer than 20 family members). They show a moderate correlation between these two values (ρ = 0.48 ± 0.08), suggesting IVCT of iron inside of phyllosilicates may cause both absorptions.

thumbnail Fig. 4

NIR slope from 1.25 μm to 2.15 μm, SNIR2, vs. NUV absorption, SNUVSVIS, of asteroid families. The error bars show the interquartile range. There is a correlation between these two values (ρ = −0.47 ± 0.12).

thumbnail Fig. 5

Spectral characteristics correlate with albedo. (top) Albedo versus VIS slope. (bottom) Albedo versus NUV absorption. The yellow symbols show the families with HYD < 0, possibly with the 0.7 μm absorption. The blue symbols are with HYD > 0. The error bars show the standard deviation.

4.2 Overview of primitive asteroid families

Figure 6 displays the median NUV absorption strength and VIS slope for each asteroid family within the different regions of the main asteroid belt, listed in Table B.1. We classified the mean family spectra by Tholen’s taxonomy, more specifically according to the values in Table 3 in Tatsumi et al. (2023). Furthermore, the highly hydrated families, as is indicated by positive median HYD values, are denoted by (h) in Table B.1. Our analysis reveals differences in NUV-VIS distributions among the IMB, MMB, and OMB. OMB families with negative HYD values exhibit a clear negative correlation in Fig. 6. In contrast, the IMB and MMB families do not show a strong correlation.

4.3 Family constitutions

The data on NUV provides an insight into the diversity among the asteroid families, allowing us to discern discrepancies in NUV absorption among family members for the first time. Figures B.1B.3 illustrate the taxonomic structure of families in the IMB, MMB, and OMB, respectively. The detailed spectral properties for each family, also compiling the previous studies, are described in Appendix A.

4.3.1 Inner main belt primitive families

From our dataset, we were able to sample six inner main belt primitive families: the Polana–Eulalia complex, Erigone, Clarissa, Sulamitis, Klio, and Chaldaea families. Despite some families having a small number of members, such as the Clarissa and Sulamitis families, there are three types of family constitutions. One is predominantly composed of F types, exemplified by the Polana–Eulalia complex and Clarissa families. Another type is characterized by a dominance of B, C, and G types (moderately to strongly hydrated types), as is seen in the Erigone and Chaldaea families. The third type consists of numerous P types (with redder visible slope), as is observed in the Sulamitis and Klio families, with little presence of D types among them. DeMeo et al. (2014) estimated there are ~100 inner belt D-types with diameters between 2.5 and 20 km. Their investigation has shown that these have a higher albedo compared to the average one of D-types.

Intensive studies on the IMB families have been conducted using ground-based and space-based spectroscopy (de León et al. 2016; Pinilla-Alonso et al. 2016; Morate et al. 2019; de León et al. 2018; Arredondo et al. 2020, 2021a,b; Tatsumi et al. 2022; Delbo et al. 2023; Harvison et al. 2024). These families have garnered attention as a potential origin of two sample-return mission targets, (162173) Ryugu and (101955) Bennu, explored by Hayabusa2 (JAXA) and OSIRIS-REx (NASA), respectively. Comparing our results with previous studies can show the robustness of our analyses and provide valuable insights at the same time (See Appendix A).

4.3.2 Middle main belt primitive families

We were able to sample 16 middle main belt families: Nemesis, Adeona, Padua, Chloris, Misa, Dora, Astrid, Konig, Hoffermeister, Vibilia, Phaeo, Mitidika, Karma, Postrema, Pallas, and Brucato (Fig. B.2). Notably, the Pallas and Brucato families are primarily distributed in the higher-inclination areas of the MMB, with inclinations ranging from 32–36 degrees and 26–31 degrees, respectively.

The Chloris, Padua, Misa, and Astrid families consist of ~40% of G types, suggesting a high degree of hydration. Similarly, the Adeona and Dora families consist of >30% of B types. The Konig family is predominantly composed of C types. The Hoffmeister and Mitidika families display a similar distribution, with F-type dominating by 60 – 70%. The Phaeo, Karma, and Postrema families host half of the P-type members. Notably, the Phaeo and Postrema families also include some D-type asteroids (~10%). The Pallas family stands out as particularly peculiar among other primitive families, as it is dominated by B-type asteroids, constituting ~60% of its members.

4.3.3 Outer main belt primitive families

We could sample 16 outer main belt families: Hygiea, Themis, Sylvia, Meliboea, Emma, Veritas, Naema, Lixiaohua, Theobalda, Beagle, Fringilla, Ursula, Inarradas, Euphrosyne, Alauda, and Luthera (Fig. B.3). The Euphrosyne, Alauda, and Luthera families are distributed in the higher-inclination area of the outer main belt, 25–28 degrees, 21–24 degrees, and 18–20 degrees, respectively.

In the OMB, several families show a high proportion of P types, such as the Sylvia, Lixiaohua, Fringilla, Ursula, Euphrosyne, and Luthera families. Hygiea, Themis, Meliboa, Naema, and Alauda consist of many taxonomic classes evenly distributed. The Veritas and Inarradas families are dominated by G types, which are possibly highly hydrated. The Theobalda and Beagle families are dominated by F types and B types by ~80% and ~60%, respectively. The Emma family is dominated by the small NUV absorption members; that is, F and P types.

thumbnail Fig. 6

NUV-VIS distribution of primitive asteroid families. The x-axis shows the NUV absorption calculated from the difference in the spectral slope of VIS and NUV and the y-axis shows the VIS spectral slope. The symbols and error bars show the median values and interquartile ranges. The circles are the families with HYD < 0 and the triangles are those with HYD > 0 (highly hydrated). (top) All the primitive asteroid families in this study are shown (blue: IMB, orange: MMB, red: OMB). Tholen’s taxonomic classification, which is applied in Tatsumi et al. (2023), is also shown as divided zones labeled with the names of classes. (bottom) The same plot as the top plot with the three regions of the main belt separately shown. The blue circles are HYD<0 and the orange triangles are HYD>0.

5 Grouping of families and compositional link

Our analysis establishes a classification scheme for 38 primitive asteroid families into ten types based on their NUV-VIS spectral features (summarized in Fig. 7).

  1. Blue-dominant families (B + F > 40%)

    • Polana–Eulalia type: Characterized by a high abundance of F-type asteroids (>40%).

    • Pallas type: F types are less common and have a higher average albedo than 0.1.

    • Themis type: F types are less common (<40%) and have a lower average albedo than 0.1.

  2. High NUV absorption families (C + G > 50%)

    • Veritas type: Dominated by G-type asteroids (>50%).

    • Padua type: G-type asteroids constitute a significant population (30–50%).

    • Erigone type: Minor presence of G type (<30%).

  3. Red-dominant families (P + D > 40%)

    • Lixiaohua type: The very minor presence of B and G types.

    • Euphrosyne type: B and G types are present in minor quantities (>5%).

The combination of NUV, 0.7 μm, and 3 μm absorption features provides valuable insights into the possible compositions and relationships between these primitive families and meteorite types. Table B.2 summarizes the characteristics of families in each group.

The 0.7-μm absorption is a well-known indicator of Fe-rich phyllosilicates (e.g., Vilas & Gaffey 1989). In Fig. 7, the families with positive HYD are underlined. It is easily noticed that many high NUV absorption families show positive HYD values, suggesting 0.7-μm band absorptions. The coexistence of positive HYD and strong absorption is consistent with the presence of Fe-rich phyllosilicates. On the other hand, only one family, Emma, shows a positive HYD value among the red-dominant families, which tend to show low NUV absorptions. The blue-dominant families are a little complicated because of the mix of positive and negative HYD values. The F-dominated Polana–Eulalia type families have negative HYD values, which agrees with the low NUV absorption, while the mix of F and B, Themis type, families sometimes show positive HYD values, indicating compositional complexity.

In particular, the 3 μm absorption feature is key because it strongly indicates an aqueous alteration state. Asteroids that underwent aqueous alteration exhibit absorption features at 2.7–2.8 μm, which is caused by metal-OH phyllosilicates, referred to as the OH band. Additionally, some asteroids display a small absorption feature at 3–3.2 μm (hereafter the 3.1-μm band) alongside the OH band, creating a “w-shaped” spectral profile (Usui et al. 2019). Others only show the OH band (Takir & Emery 2012; Rivkin 2012; Usui et al. 2019, Pallas-type or Sharp-type;). The 3.1-μm band can further be distinguished by its absorption center; the absorption center at 3.05 μm (Cerestype) and the absorption center at 3.1 μm (Europa-type) (Rivkin et al. 2019). In situ observations by the Dawn spacecraft suggest that the 3.061 ± 0.011 μm absorption feature on Ceres (Ammannito et al. 2016) is due to ammoniated phyllosilicates (de Sanctis et al. 2015), which is supported by laboratory measurements (Ehlmann et al. 2018). Kurokawa et al. (2022) propose that a narrow absorption band at less than 3.08 μm might be indicative of ammoniated phyllosilicates rather than ice frost. Conversely, a broad band at 3.08–3.1 μm is more consistent with ice frost on the surface. This confirms the previous laboratory measurements of reflectance spectra of ammoniated saponites (King et al. 1992). Their modeling also suggests that the ammoniated phyllosilicates form readily under high water-rock ratios (W/Rs) when the starting materials contain NH3 and CO2 ice (Kurokawa et al. 2022). The scarcity of the 3.1 μm feature in Ch/Cgh asteroids (potential parent bodies of CM chondrites) implies that asteroids with the Ceres-type 3-μm band likely formed under higher W/R conditions compared to those with the Pallas-type, sharp 3-μm feature. Alternatively, asteroids with the Pallas-type, sharp 3-μm feature might have formed inward of the NH3 snow line, while Ceres types were formed beyond it.

The 3-μm spectral observations from space-based and ground-based telescopes reveal distinct compositional trends among asteroid families. The largest members of high-NUV-absorption families exhibit sharp-type 3-μm absorption features indicative of phyllosilicates. In contrast, the largest members of red-dominant families show the Europa-type 3-μm absorption features, suggesting the presence of water ice.

The blue-dominant families exhibit more complex compositional variations. While ground-based observations of the largest members in the Polana–Eulalia type, such as (142) Polana and (778) Theobalda, have not revealed clear 3-μm absorption features, those in the Themis type show a range of 3-μm absorption shapes, indicating a diverse composition. However, it is note-worthy that families originating from larger parent bodies, like Themis, Hygiea, and Alauda, tend to exhibit Ceres- or Europatype w-shaped features, suggesting a common evolutionary history.

5.1 Polana–Eulalia type

The Polana–Eulalia, Clarissa, Hoffmeister, Mitidika, and Theobalda families belong to the Polana–Eulalia type. These families are predominantly composed of F-type asteroids, characterized by neutral to negative VIS spectral slopes and weak or absent NUV absorption.

The estimated diameters of parent bodies that formed these families are all less than 100 km (Brož et al. 2013; de León et al. 2018). Notably, the Theobalda and Clarissa families are considered young with ages of ~6.9 Myr (Novaković 2010) and ~60 Myr (Bottke et al. 2015), respectively. Despite exhibiting the bluest VIS slopes among all of the studied families, their albedo and NUV absorption strengths differ from those of the Pallas or Themis families. This weak NUV absorption of the Theobalda and Clarissa families suggests a compositional origin rather than space weathering or an aging effect.

Ground-based observations indicate that most members of the Polana–Eulalia, Hoffmeister, and Clarissa families lack the 0.7 μm absorption band (de León et al. 2016; Morate et al. 2018, 2021). Moreover, other NIR spectroscopy of (142) Polana and (778) Theobalda in the 3 μm region has revealed no clear absorption in the observable wavelengths from the ground (Takir et al. 2024; Rivkin et al. 2022). It is important to note that the atmospheric absorption hinders direct observations of the 2.7–2.9 μm range, making it impossible to definitively determine the presence or absence of OH-band absorption. The combined spectral signature from NUV to NIR aligns with either Ryugu, Bennu samples (CI-type) or dehydrated carbonaceous chondrites (ATCC or CY-type).

The densities of the largest members within the Polana–Eulalia type families remain unknown. However, the density estimates for two F-type asteroids, (88) Thisbe and (704) Interamnia, are 2.14 ± 0.42 g/cm3 and 1.84 ± 0.28 g/cm3, respectively (Vernazza et al. 2021). These values are consistent with the CI and CM chondrites, as well as Tagish Lake meteorites (Flynn et al. 2018; Yada et al. 2021). Although Almahata Sitta meteorites share spectral characteristics with F-type asteroids in VIS-NIR (Gayon-Markt et al. 2012), their bulk density was estimated as 2.8–3.1 g/cm3 (Shaddad et al. 2010; Kohout et al. 2010), which is distinct from the densities of F-type asteroids. It is crucial to consider that these F-type asteroids are not members of the Polana–Eulalia type families, and their formation processes might differ, influencing their densities. Further investigations are needed to determine the bulk densities, and directly observe in the 3 μm region with space telescopes of the constituent members within the Polana–Eulalia type families, for a more conclusive compositional link.

thumbnail Fig. 7

Grouping of primitive asteroid families. We classified the primitive asteroid families into eight types based on the taxonomic compositions and average albedos. The asteroid families with HYD > 0, possibly the 0.7-μm absorption, are underlined.

5.2 Pallas type

The Pallas family stands out as the sole member of the Pallas type within our classification scheme. This family is unique due to its high albedo (Alí-Lagoa et al. 2016). The Pallas family members lack the 0.7-μm absorption band, but exhibit strong NUV absorption. Further compositional insights come from the detection of a deep and sharp OH band by AKARI (Usui et al. 2019), indicative of abundant phyllosilicates on the surface. The central wavelength of the OH band at 2.73 μm aligns well with highly aqueously altered chondrites like CI and CR (Potin et al. 2020). Pallas does not, however, exhibit a 3.1-μm absorption band (Usui et al. 2019), which is associated with ammoniated phyllosilicates or structural water. The reflectance factors at a 30° phase angle of CR chondrites are 0.08 ± 0.04, corresponding to a geometric albedo of 19 ± 11% using the phase curve slope parameter by Martikainen et al. (2021). Thus, the CR meteorites are consistent with the observed albedo value of Pallas and its family. This suggests that CR chondrites are a good candidate for the composition of the Pallas family’s parent body. Recently, high-angular-resolution observations of Pallas were obtained using the Very Large Telescope. Marsset et al. (2020) estimated its density of 2.89 ± 0.08 g/cm3, exceeding the density of the differentiated dwarf planet (1) Ceres. The study also revealed its very unique collisional environment due to its large orbital eccentricity (e = 0.23) and inclination (i = 34.8°). Notably, the high bulk density of Pallas is the most consistent with the bulk density of CR meteorites (3.1 g/cm3) compared to CI (1.6 g/cm3) or CM (2.3 g/cm3) chondrites (Macke et al. 2011; Flynn et al. 2018), further supporting the compositional link suggested by spectral features.

5.3 Themis type

The Adeona, Dora, Brucato, Hygiea, Themis, (Beagle), and Alauda families comprise the Themis type within our classification scheme. These families are found in the middle and outer main belt and host a significant population of blue F- and B-type asteroids, along with other taxonomic types, except for D-type.

Observations of the 3 μm region for the largest bodies of those families reveal intriguing details. The spectra of (10) Hygiea, (24) Themis, and (702) Alauda exhibit w-shaped absorption features at 2.7 μm and 3.1 μm (Usui et al. 2019). For (145) Adeona, the 2.7-μm absorption is evident, but the 3.1-μm feature is unclear due to low signal-to-noise (Usui et al. 2019).

Furthermore, some of the members of the Hygiea and Themis families, and most members of the Adeona and Dora families, exhibit the 0.7-μm absorption band (De Prá et al. 2020b; Morate et al. 2021). This implies a compositional mix of Fe-rich and Fe-poor phyllosilicates within these families. More interestingly, the Hygiea, Themis, Beagle, Alauda, and Adeona families are known to host active asteroids (Hsieh et al. 2018), suggesting a volatile-rich composition of their parent bodies.

(10) Hygiea’s composition appears similar to (1) Ceres based also on the NIR and MIR observations (Takir & Emery 2012; Vernazza et al. 2017). Their densities are also compatible: (10) Hygiea of 2.06 ± 0.20 g/cm3 (Vernazza et al. 2021) and (1) Ceres of 2.08 ± 0.04 g/cm3 (Thomas et al. 2005). A Ceres-associated family has not been identified, while the parent body of the Hygiea family suffered a large impact. The similarities between (10) Hygiea and (1) Ceres are puzzling considering their distinct evolutionary histories. The Dawn spacecraft mission revealed the surface composition of (1) Ceres to be dominated by Mg-rich serpentine and ammoniated clay (de Sanctis et al. 2015). So far, there has not been an exact match between Ceres and our meteorite collection on Earth. Additionally, the remotesensing observation on Cerealia Facula suggested the presence of internal salty water (De Sanctis et al. 2020), hinting at the differentiation of (1) Ceres. And the 0.7 μm absorption, possibly Fe-rich phyllosilicates, was detected around the Occator crater (Rizos et al. 2019). This internal layering with distinct lithologies could potentially explain the spectral variations observed among members within the Themis-type families.

The estimated sizes of the primordial bodies of Themistype families are mostly over 200 km, with the exceptions being the Brucato family. This size of primordial bodies is considered important for differentiation and potentially forming a compositional distribution similar to (1) Ceres.

The densities of the largest members within these families are all relatively low: (10) Hygiea is 2.06 ± 0.20 g/cm3 (Vernazza et al. 2021), (24) Themis is 1.31 ± 0.62 g/cm3 (Vernazza et al. 2021), (145) Adeona is 1.52 ± 0.21 g/cm3 (Vernazza et al. 2021), and (702) Alauda is 1.57 ± 0.5 g/cm3 (Rojo & Margot 2011). These densities are compatible with CI, CM chondrites, and Tagish Lake meteorite (Flynn et al. 2018; Yada et al. 2021). The observed density variation can be attributed to differences in water ice content. Again, it is important to note that, similar to Ceres, a perfect match for the compositions of these parent bodies might not exist in our current meteorite collection. Their compositions likely represent a mixture of Fe-rich, Fe-poor, and ammoniated phyllosilicates, along with water ice.

5.4 Veritas type

The Veritas-type families are characterized by strong NUV absorption. Only two families in the outer main belt, Veritas and Inarradas, are categorized into this family type. These Veritas-type families could consist of asteroids associated with the 0.7 μm band based on the positive HYD values. As was discussed above, the 0.7 μm band is a strong indication of Fe-rich phyllosilicates, which is the main assembly of CM chondrites.

The Veritas family is an extremely young asteroid family formed ~8.3 Myr ago (Nesvorný et al. 2003) and the Inarradas family was also estimated to possibly be as young as 20 Myr (Carruba et al. 2018). The strong NUV and 0.7 μm absorption might be associated with the youngness of these families, meaning that the absorption is prominent enough because they are least space-weathered.

5.5 Padua type

The Padua-type families also have strong NUV absorptions, but slightly lower than the Veritas type. Five families, Padua, Chloris, Misa, Chaldaea, and Astrid, are this family type, located in the IMB and MMB. Spectroscopic studies on the Padua family suggested that the members rarely have the 0.7 μm band (Morate et al. 2021). This is consistent with negative HYD value of the Padua family. However, the stronger NUV absorption compared with Polana–Eulalia type may suggest the higher content of phyllosilicates in their parent body.

The largest body (363) Padua and (313) Chaldaea of their namesake families, respectively, have been observed in the 3 μm region by a ground-based telescope and found to be a sharp shape (Rivkin et al. 2022), although (363) Padua has a redder VIS spectral slope categorized as P type. The relatively short band center wavelength of the OH band (Usui et al. 2019) and the absence of the 0.7 μm band (Morate et al. 2021) suggest Fe-poor or Mg-rich phyllosilicates, such as CI chondrites, for (363) Padua. On the other hand, Chloris, Misa, and Astrid families have positive HYD values, suggesting the presence of 0.7-μm band. Also, the previous ground-based study suggests the presence of 0.7-μm band for the Chaldaea family (Morate et al. 2019). Thus, considering the wide variation of taxonomic composition among the family members, the parent body could be composed of mixed mineral phases such as Fe-rich and Fe-poor phyllosilicates. Consequently, further detailed observations in the 0.7 μm and 3 μm regions are necessary to have a solid conclusion on their composition in the future.

The parent body sizes for this family type are smaller than 150 km. The parent body size could explain the difference in compositions between high NUV absorption families and Themis-type families.

5.6 Erigone type

The Erigone type encompasses the Erigone, Nemesis, Meliboea, Klio, Naema, Sulamitis, Konig, and Vibilia families, distributed across various regions of the main belt. This type is characterized by moderately deep NUV absorption.

There are notable taxonomic compositional variations in this family type. Klio, Sulamitis, Meliboea, and Nemesis families are characterized by a significant presence of P types, along with some portions of F and G types. The Konig family stands out as a unique case: the majority of its members are classified as C-type asteroids, with a minority belonging to G-type with strong NUV absorption.

Spectroscopic observations by Morate et al. (2016, 2018, 2019); Mothé-Diniz et al. (2005) have revealed that the largest members and majority of members within the Erigone, Meliboea, and Sulamitis families exhibit the 0.7-um absorption band. In contrast, the Nemesis family members rarely have the 0.7-μm band Morate et al. (2021). Similarly, the only Nemesis family has a negative HYD value, while other families have positive values. Because of this contrast in the 0.7-μm region, the Nemesis family could be an outlier in this type.

The largest members across these families show a remarkably uniform signature: the sharp and deep 3 μm absorption (Rivkin et al. 2015, 2022). However, except for (128) Nemesis, they were observed from the ground and missed the direct observation around 2.7 μm. Further investigations, such as at the band center position, using a space telescope, are needed to fully understand the implications of this shared feature. Based on the prevalence of the 0.7-μm absorption band, the Erigone-type families likely originated from parent bodies composed primarily of Fe-rich phyllosilicates, resembling CM chondrites. The largest body, (128) Nemesis, of its namesake family is the only one in this type observed in the 3 μm region directly by AKARI and was found to have a sharp 3-μm band shape centered at 2.74 μm (Usui et al. 2019). Moreover, the density of (128) Nemesis was estimated as 1.5 ± 0.8 g/cm3 (Vernazza et al. 2021). This range of density is consistent with the CI and CM chondrites (Flynn et al. 2018; Yada et al. 2021).

5.7 Euphyrosyne type

The Euphyrosyne type is distinguished by the red VIS slope, reflecting the dominance of P-type asteroids. However, the Euphyrosyne-type families also encompass members with high NUV absorption, such as B and G types. The Euphrosyne, Ursula, Phaeo, Sylvia, Postrema, and Karma families are classified into this family type. The significant variation in NUV absorption observed across Euphrosyne-type family members suggests a wide range of hydration states within their parent bodies. However, the spectroscopic study on the Ursula family did not detect the 0.7-μm band in any of the seven analyzed samples (De Prá et al. 2020b). Moreover, consistently our analysis suggests all these families have negative HYD values. This absence suggests that the hydrated minerals present might not be Fe-rich.

Observations of (31) Euphrosyne and (375) Ursula in the 3-μm region using ground-based telescopes revealed broad absorptions at 3.1 μm, suggesting the presence of water ice (Takir & Emery 2012; Rivkin et al. 2019, 2022). However, the lack of direct observations at 2.7 μm prevents definitive confirmation of hydrated minerals. The observation of (87) Sylvia, the largest member of its namesake family, has yielded contradictory results regarding its mineralogy. The space-based observation by AKARI failed to detect any absorption feature either at 2.7 μm and 3.1 μm (Usui et al. 2019). However, the ground-based observation suggested a broad absorption at 3.1 μm, indicative of water ice (Rivkin et al. 2022).

The estimated densities of (31) Euphrosyne and (87) Sylvia are 1.67 ± 0.24 g/cm3 (Yang et al. 2020) and 1.2 ± 0.1 g/cm3 (Marchis et al. 2005), respectively. The analysis of the spin period distribution within the Ursula family suggests a much lower density around 0.8 g/cm3 (De Prá et al. 2020b). These low densities align with the CI chondrites and Tagish Lake meteorite, or even fall lower than those values (Flynn et al. 2018; Yada et al. 2021). The red VIS spectral slope exhibited by the Euphrosyne-type families might prefer the possible compositional analog to the Tagish Lake meteorite.

By integrating spectral and physical characteristics, we can infer that the parent bodies of the Euphyrosyne type likely consisted of a complex mixture: unhydrated and hydrated silicates (possibly similar to the Tagish Lake meteorite), water ice, and an organic-rich matrix.

5.8 Lixiaohua type

The Lixiaohua type is characterized by the red VIS slopes, weak NUV absorption, and one of the lowest albedos. These families are primarily found in OMB and include the Lixiaohua, Fringilla, Luthera, and Emma families. The red VIS spectra and low albedo of these families are possibly indicative of an organic-rich composition. The scarcity of members with high NUV absorption suggests a low abundance of hydrated minerals within these families. Indeed, the spectroscopic study did not detect the 0.7-μm band in any members of the Lixiaohua family (De Prá et al. 2020a). However, it should be noted that most families of this type present negative HYD values, consistent with the Lixiaohua family case, except for the Emma family, which has a slightly positive HYD value. While Lixiaohua and Emma show significant portions of F types, the Fringilla and Luthera families are more than 60% dominated by P types.

Two active asteroids, P/2012 T1 and 313P, have been linked to the Lixiaohua family (Hsieh et al. 2013, 2015). This association indicates the volatile-rich composition of the parent body of this family.

Due to the lack of observations in the 3-μm region, it is difficult to constrain the composition of this family type. The low albedo suggests the high contents of organics. Interestingly, these parent bodies are estimated to be ~100 km.

Based on the available data, the Lixiaohua-type family is likely composed of materials resembling cometary nuclei: a mixture of anhydrous silicates, water ice, and an organic-rich matrix. Further investigations, especially in the NIR spectra, are required to refine this hypothesis.

6 Discussions

6.1 NUV-VIS spectral change by aqueous alteration and thermal metamorphism processes

As we discuss in the companion paper, the NUV absorption is closely related to phyllosilicate abundance, especially Fe-rich phyllosilicates (Tatsumi et al. 2023). The correlation between HYD (indication of 0.7 μm absorption) and the NUV absorption of asteroid families reinforces this idea. In addition to this, our analysis of asteroid families shows a clear correlation between the NUV absorption and albedo.

The analysis reveals key correlations: less NUV absorption correlates with less HYD and lower albedo, while more NUV absorption correlates with higher HYD and higher albedo (Figs. 5 and 3). This variation may correspond predominantly to the degree of aqueous alteration in asteroids. In Fig. 6, we may find two trends: one from P to G and another from G to F.

The main trend from P to G could be simple to understand. The presence of 3.1-μm and absence of 0.7 μm and 2.7 μm absorptions of red-dominant families suggest that they have not undergone a strong aqueous alteration process that requires the liquid water. This suggests that the red-dominant families may possess few phyllosilicates. On the other hand, the high NUV absorption families may contain significant amount of phyllosilicates as is discussed in the previous sections. Thus, the transition from P to G can be explained by anhydrous silicates that are transformed into phyllosilicates by water-rock interaction triggered by radiogenic decay of 26Al. However, some of the red-dominant families host G-type asteroids, suggesting a potential aqueous alteration process in some portion of the parent body. The transition trend of P to G indicates that there are gradual differences in the heating temperature among the parent bodies of asteroid families on this trend. The peak temperature variation can be caused by accretion timing and/or the size of parent bodies. Among those families, the Luthera family is the reddest family including a significant number of D types, suggesting the least aqueous alteration of this parent body.

The F-type dominated families exhibit weak NUV absorption, despite evidence from the recent sample return missions suggesting that they possess significant phyllosilicates. The second transition trend from G to F presents a more intriguing puzzle. Our meteorite samples offer valuable clues to understanding the mineralogy of asteroids. Among meteorites, the CI and CM types especially are considered to be analogs of aqueously altered dark primitive asteroids. Some mineral phases of CI and CM chondrites appear to have formed as a result of post-accretion aqueous processes inside the parent bodies. For example, sulfate and carbonate salts, and magnetite in CI chondrites have crystallized from aqueous alteration (Kerridge et al. 1979). Aqueous alteration degree can change from Fe-rich phyllosilicates in CM chondrites to Fe-poor phyllosilicates and magnetites (McSween 1979). The difference in the degree of aqueous alteration between CM and CI meteorites was interpreted as W/R and temperature differences based on the oxygen isotopic ratios (Zolensky et al. 1997; Clayton & Mayeda 1999). More specifically, forming CI chondrites required a higher temperature (~150 °C) and W/R a factor of four or five higher than for CM chondrites (Zolensky et al. 1997; Clayton & Mayeda 1999). By the alteration process under high W/R in CI meteorites’ parent bodies, interlayer Fe in phyllosilicates was transformed into magnetite and replaced by Mg, resulting in the absence of the 0.7 μm band absorption by Fe-rich phyllosilicates. Futhermore, magnetite abundance contributes to the changing overall spectral slope, making it bluer and darker in NUV-VIS (Yang & Jewitt 2010; Cloutis et al. 2011; Izawa et al. 2019). Further discussions of opaque substances are presented in the following subsection. Thus, the trend from G to F types (more specifically, the Veritas family to the Theobalda family) can be explained by the aqueous alteration progress from CM to CI.

Alternatively, the decrease in NUV absorption in ATCCs is known to be caused by thermal metamorphism after aqueous alteration (Cloutis et al. 2012). ATCCs are known to have smaller NUV absorption compared with hydrous CM/CI, due to decomposition of phyllosilicates with increased temperature (Hiroi et al. 1993; Hiroi et al. 1996b). The heated Ivuna (CI1) sample especially shows bluing in the VIS and reddening in the NUV (Hiroi et al. 1996a). Indeed, clear 2.7-μm absorption has not been found on the Polana–Eulalia-type families. To clarify the cause of small NUV absorption can be tested by the 3 μm region observations of these asteroid families.

6.2 NUV variation by opaque compounds

While the strong correlation between NUV absorption and the 0.7 μm band suggests a link with Fe-rich phyllosilicates, the presence of opaque compounds can complicate this relationship. Hendrix & Vilas (2019) demonstrates that increasing abundances of opaque substances, such as amorphous carbon or graphite in phyllosilicates, can weaken the NUV absorption signal as well as decrease the overall reflectance. Similarly, magnetite, another opaque mineral commonly found in CI chondrites as a by-product of aqueous alteration, exhibits a significant blue slope in NUV (Izawa et al. 2019). Therefore, the strength of the NUV absorption might not solely reflect the abundance of Fe-rich phyllosilicates, but could be also influenced by the presence of opaque minerals. While opaque compounds can similarly influence NUV absorption, their effect on the VIS range differs. Magnetites exhibit blue spectra in VIS (Yang & Jewitt 2010; Cloutis et al. 2011; Izawa et al. 2019). In contrast, organic materials, another opaque component in asteroids, are linked to a redder VIS slope. Gradie & Veverka (1980) introduced the concept of organic solids to explain the red slope of the D-type asteroids. They found that the mixture of montmorillonite, magnetite, carbon black, and coal-tar residue matches the spectra of D-type asteroids. Kerogen-like structures are found in carbonaceous meteorites (Kerridge et al. 1987). Laboratory measurements also suggest that the VIS slope increases with an increasing aliphatic-to-aromatic ratio (Moroz et al. 1998, 2004). Furthermore, a red slope in VIS characterizes organic-rich regions on Ceres, showing the spatial correlation with the 3.4 μm detection (Pieters et al. 2018). Also, there is a potential correlation between the H/C value and the redness of the VIS slope of insoluble organic matter in carbonaceous meteorites (Kaplan et al. 2019). This interplay between phyllosilicates and opaque compounds might explain some of the observed trends. F and P types exhibit both weak or no NUV absorption and have low albedo. However, there are variations in VIS slope. F types are abundant both in magnetite and carbon compounds, resulting in weak NUV absorption and blue VIS spectra, while P types are more controlled by the organics, resulting in weak NUV and red VIS spectra. D types are even redder, possibly because of the high H/C ratio. This highlights the need for further studies to explore how the combined presence of phyllosilicates and various opaque compounds influences the overall spectral signature.

6.3 Internal structure and formation of primordial bodies

Asteroids are believed to be the remnants of planetesimals. The accretion age of carbonaceous chondrites at 2–4 Myr is comparable to the core accretion timescale of Jupiter of 1–5 Myr (e.g., Kruijer et al. 2020). Recent discussions raise the necessity of coupling the formation of Jupiter and planetesimals (asteroids). Recently, the pebble accretion model has become a widely accepted hypothesis for the formation of giant planets like Jupiter and Saturn. In this model, planetary embryos grow rapidly by sweeping up millimeter-to decimeter-sized pebbles dropping by efficient gas drag (Ormel & Klahr 2010; Lambrechts & Johansen 2012). In addition to the model, Gundlach & Blum (2014) proposed the efficient pebble formation outside of the snowline, due to the sticking of water-ice particles. In this model, the pressure bumps where the pebbles are trapped are important to draw the formation of asteroids. Atacama Large Millimeter/submillimeter Array (ALMA) observations show that many protoplanetary disks have non-monotonic pebble distributions, with rings or spiral structures being very common (Andrews et al. 2018). It is plausible that these pebble structures are caused by pressure bumps. The origin of pressure bumps is unclear but they can be formed by the presence of relatively large planetesimal in the disk (e.g., Dodson-Robinson & Salyk 2011; Pinilla et al. 2012), the dead zones (e.g., Regály et al. 2012), or the water and CO snowlines (Charnoz et al. 2021; Izidoro et al. 2021). Regardless of the origin of pressure bumps, the pebble accretion model can create planetesimals quite rapidly and efficiently in pressure bumps. Izidoro et al. (2021) propose three condensation lines, silicates, water, and CO, to reproduce the mass distribution and isotropic dichotomy in meteorites and rocky planets. In their model, non-carbonaceous, carbonaceous, and comet-like objects were formed almost simultaneously around three condensation lines. Either type II migration of proto-giant planets (Walsh et al. 2011) or the temperature evolution of the gas disk (Izidoro et al. 2021) could vary the position of pressure bumps inward. Thus, the timing and place of asteroid formation can be arbitrarily varied, unlike the classical formation model where the planetesimals were formed in monotonic order as a function of the semi-major axis (Grimm & McSween 1993). It is known that Calcium–Alminium-rich Inclusions (CAIs) are more abundant for carbonaceous chondrites than non-carbonaceous meteorites, even though CAIs are formed at high temperatures of >1400 K near the Sun (Rubin 2011). This can be also explained by CAI trapped by a pressure bump (Desch et al. 2018).

Heating temperatures in the primordial bodies are important in the formation of various minerals observed in asteroids and meteorites. The short-lived radiogenic heat source of 26Al is the plausible heat source for the early stage of planetesimal formation. After 5 Myr of CAIs, 26Al decreases by <1%. Heating temperature by 26Al highly depends on W/R and formation timing. W/R can be a good indicator of the formation distance from the Sun. The W/R at the CI formation zone was estimated as unity (Lodders 2003). However, it should be noted that W/R can be changed by differentiation process (Wakita & Sekiya 2011; Neumann et al. 2020; Kurokawa et al. 2022). The heterogeneities of the presence or absence of the 0.7 μm band or NUV absorption suggest a possible differentiation process. The taxonomic composition of a family can be a witness of the internal structure of a primordial body, before comprising a collisional family. Based on our analysis of collisional families, the internal structure and formation of primordial bodies are estimated as below:

  • Themis-type families are a mixture of presence or absence of the 0.7 μm bands and high or low NUV absorptions, suggesting the heterogeneous mixture of high and low W/R hydrated minerals. Since most of them are estimated as large parent bodies >200 km, this different W/R can be formed by the differentiation process (Wakita & Sekiya 2011; Neumann et al. 2020). The internal layer of this type could have W/R~0.3–0.6 to form Fe-rich phyllosilicates and the outside layer would be more Fe-poor and Mg-rich phyllosilicates under high W/R ratio. Seeing fewer G types for asteroids >50 km also supports the view that low W/R hydrated lithology was formed inside of the primordial body. If the percentages in Figs. B.1B.3 reflect the internal structure, B type families especially are composed of more than 50% of Fe-poor phyllosilicates (total of F- and B-type members). They could have a subsurface liquid water layer like Ceres (Russell et al. 2016).

  • Veritas- and Padua-type families are dominated by asteroids with the 0.7 μm band and deep NUV absorption, strongly suggesting Fe-rich phyllosilicates. Their parent bodies are smaller than Themis-type families. Noticeably, the Veritas and Inarradas families are >50% composed of G types. The primordial bodies of the G-type families may be formed in a low-W/R environment. Alternatively, they might lose their water subsequent to their differentiation, in which they need to be accreted fast and large to differentiate the water and rock, and consequently lose the water to the space by evaporation or impacts.

  • Polana–Eulalia-type families comprise more than 60% of F-type members, dominated possibly by Fe-poor phyllosilicates and magnetites. G-type members are very rare, comprising <5%. This type shows a more homogeneous distribution than other types. The Polana-Eupalia family, especially, comprises ~80% of F-type members. The primordial bodies of F-type families might be formed under high-W/R conditions. Since they are relatively small, <100 km, and taxonomically homogeneous, they might not experience a strong differentiation process during aqueous alteration. They might be formed farther beyond Jupiter compared with Veritas and Padua types suggested by Fe-poor phyllosilicate abundance, and need to be accreted fast enough to melt ice and trigger aqueous alteration <2 Myr (Wakita & Sekiya 2011).

  • Red-dominant asteroid families are a mixture of various types; some host B and G types and some do not. This suggests that a minor portion of some primordial bodies of red-dominant families have been aqueously altered. In that case, the cores of some primordial bodies could be warm enough to induce liquid water, while most parts are kept below the melting point. In contrast, some parent bodies have not undergone aqueous alteration and possibly retain the original compositions such as organics and volatiles.

There is no D-type-dominated family in our samples. Thus, it is difficult to deduce anything about the internal structure based on our analysis. Some D types are associated with the Lixiaohua-type P-dominant families as well as a minor portion in the high NUV absorption families, such as the Milboea families. It may be important to expand the taxonomic survey to the outer Solar System using large telescopes in the future.

6.4 Polana–Eulalia complex family as the potential origin of Bennu and Ryugu

The Polana–Eulalia complex family is an intriguing family because it is the potential origin for the sample-return targets (101955) Bennu by OSIRIS-REx and (162173) Ryugu by Hayabusa2. This complex is possibly formed from two families, the older New Polana family and the younger Eulalia family (Walsh et al. 2013). There have been several spectral studies of these two families. Ground-based spectral observations did not find significant differences in these families from NUV to NIR (de León et al. 2016; Pinilla-Alonso et al. 2016; Tatsumi et al. 2022). In contrast, recent analysis using the Gaia DR3 catalog suggests that the two families are statistically different spectral groups; that is, they can be distinguished by the wavelength region 370–550 nm (Delbo et al. 2023).

We tested if two families can be distinguished using our dataset. Figure 8 shows the albedo, NUV absorption, and VIS slope distributions for these two asteroid families. Notably, the distribution of albedo differs in two families; the New Polana family has bimodality, while the Eulalia family has one Gaussian-like distribution. This bimodality of the New Polana family might be related to the bimodal albedo distribution found in the boulders on Bennu (DellaGiustina et al. 2020), while the albedo distribution of Ryugu is more unimodal (Sugita et al. 2019). Moreover, the NUV absorption of New Polana family slightly shifts to a higher value, which is consistent with Delbo et al. (2023). The distribution of VIS slope is also different between the two. Both families have a negative average value in the VIS slope, which is not consistent with the flat average spectra of Delbo et al. (2023). Even though it is possible to statistically distinguish the two families, it is still difficult to address which family Bennu or Ryugu associate with because the two family distributions overlap and the variations inside the family members are large.

6.5 Pallas family as the potential origin of Phaethon

(3200) Phaethon was selected as the target asteroid of the DESTINY+ mission by JAXA (Ozaki et al. 2022). Phaethon is the first asteroid associated with a meteor shower, as Whipple has pointed out that Phaethons has a link to the Geminids meteor stream (reported in Marsden 1983). Initially, de León et al. (2010) suggested the Pallas family as the potential origin of the near-Earth asteroid (3200) Phaethon due to similarities in their inclinations and spectral shapes. A recent dynamical study further supports its origin from the Pallas family through 8:3 or 5:2 mean motion resonances (Todorović & Todorović 2018).

Our analyses suggest that the Pallas family is a unique family, characterized by members with a significantly high albedo. The albedo of Phaethon has been estimated to vary largely between 8% and 12% (Kareta et al. 2018; Hanuš et al. 2016; Ito et al. 2018), which falls at the lower end compared to the Pallas family members studied here. There could be differences between the Pallas family and Phaethon in terms of their hydration state. While the Pallas family shows moderate NUV absorption strength, Phaethon exhibits a flat spectrum down to a wavelength of 0.38 μm (Lazzarin et al. 1996). Furthermore, Takir et al. (2020) found no evidence for the 3 μm absorption with different rotational phases of Phaethon. This observational evidence suggests that Phaethon may have a less hydrated or thermally metamorphosed surface compared to the average Pallas family. Nevertheless, our analysis suggests the presence of F types in the Pallas family, comprising up to ~30% of its members, indicating that Phaethon may preserve its original color as a heterogeneous interior of the Pallas primordial body.

Alternatively, the flat NUV spectrum of Phaethon can be explained by heating of its surface during close encounters with the Sun. The similarity in albedo and 3-μm-band suggests a potential connection between CM and CR chondrites and the Pallas family. Heating experiments on a CM chondrite have shown darkening for moderate heating temperatures up to 800°C and brightening beyond 900°C (Hiroi et al. 1993). This is consistent with the thermal model predicting temperatures of up to 1000 K on the surface of Phaethon at its perihelion (MacLennan et al. 2021).

It is important to note that there are other asteroid families with possible dynamical connections; that is, high inclination angles, such as the Brucato family in MMB and the Alauda and Euphrosyne families in OMB. These families also host a significant portion of F-type asteroids, suggesting a potential origin of Phaethon in terms of spectral similarity. Specifically, the Alauda and Brucato families exhibit similar albedos if the albedo of Phaethon is at the lower end (8%). However, further dynamical investigation is needed to assess the probabilities for the origin of Phaethon.

thumbnail Fig. 8

Spectral characteristics of New Polana and Eulalia family members. (top) Albedo, (middle) NUV absorption, and (bottom) VIS slope.

6.6 Space weathering

Nesvornỳ et al. (2005) showed the correlation of aging and bluing using the average colors of dark asteroid families. In this study, we increased the number of asteroids. Thus, we have obtained the colors of more primitive dark families. Unlike Nesvornỳ et al. (2005), we did not find a strong correlation between the color and the age of primitive asteroid families. However, it is intriguing that two very young families of <10 Myr (Nesvorný et al. 2003; Tsiganis et al. 2007; Novaković 2010), Veritas and Theobalda, are located as the endmembers in the NUV-VIS space (Fig. 6). The Veritas family shows the highest NUV absorption, while the Theobalda family is the bluest and has low NUV absorption. They are located in the OMB, meaning that their spectra possibly reflect the pristine material that escaped from the severe space weathering. Among the Polana–Eulalia-type families, the Theobalda family is the bluest, suggesting that this family type may have a reddening trend by space weathering. Similarly, the Veritas family has the highest NUV absorption, suggesting that other families may be suffering from space weathering reducing the NUV absorption. Thomas et al. (2021) suggest two distinct trends of space weathering observed in the large asteroid families. The Hygiea-type trend shows a bluing with increasing object size until a minimum slope value, while the Themis-type has a reddening trend with increasing object size until a maximum slope value is reached. They found that eight families out of nine that they investigated have the Hygiea-type trend. Although the Hygiea and Themis families were classified in the same type in our study, they show different outcomes of space weathering. This indicates the complexity of space weathering. Recent space missions to Ryugu and Bennu also show the complexity of space weathering on carbonaceous asteroids (Sugita et al. 2019; Clark et al. 2023). Yumoto et al. (2024) suggests that the surface physical conditions can change the weathering trend even if these two asteroids are the same composition. In summary, the C-complex spectral variation cannot be explained only by a simple space weathering trend but the substantial difference in composition and/or surface physical condition may play an important role.

7 Conclusions

We have investigated the spectral characteristics of 38 primitive asteroid families. We used the combined dataset of ECAS and SDSS for NUV-VIS reflectance spectra of these families to classify the family members based on Tholen’s taxonomy. Besides NUV absorption and the VIS slope, we have also investigated HYD as the proxy of 0.7 μm absorption, albedo, and NIR slopes for further characterization.

In addition to Tatsumi et al. (2023), a link between NUV absorption strength and the presence of Fe-bearing phyllosilicates is suggested, further supported by a correlation between NUV absorption and the HYD. Furthermore, a potential negative correlation between NUV absorption and the NIR slope (1.25–2.15 μm) is observed, potentially due to changes in the spectral slope caused by hydrated silicates.

A correlation is found between albedo and spectral characteristics. Asteroids with a higher albedo tend to have a bluer VIS slope and stronger NUV absorption. Interestingly, families with positive HYD (more hydrated ones) exhibit slightly higher NUV absorption despite having a similar albedo range to those with negative HYD.

There is also regional variation inside of the asteroid main belt. The distribution of NUV absorption and VIS slope differs across the main asteroid belt. OMB families show a clear negative correlation, while IMB families show a possible positive correlation.

We established a classification scheme for separating primitive asteroids into eight types based on their taxonomic constitutions. The families in the same types share some common features, suggesting similar compositions. The 3-μm absorption feature, especially, can provide us further information on phyllosililicates, water ice, and organics. The high-NUV-absorption families and the red-dominant families have common 3-μ absorption features among them, while the blue-dominant families clearly have distinctive characteristics among the sub-types, suggesting complex compositions:

  • Polana-Eulalia-type families (Polana–Eulalia, Clarissa, Hoffmeister, Mitidika, and Theobalda) are primarily composed of F-type asteroids with a neutral or negative VIS slope and weak NUV absorption. They may be linked to CIs or dehydrated carbonaceous chondrites;

  • The Pallas type (only Pallas) is unique due to its high albedo and strong NUV absorption. It likely originated from a CR chondrite-like parent body with abundant phyllosilicates;

  • Themis-type families (Themis, Hygiea, Adeona, Dora, Brucato, Alauda, and Beagle) are found in MMB and OMB. They show a mix of blue F and B types and contain members with w-shaped absorption features, indicating hydrated phyllosilicates with water ice or ammoniated phyllosilicates. Their parent bodies may be similar to Ceres in composition;

  • Veritas-type families (Veritas and Inarradas) are young and characterized by strong NUV absorption, possibly due to less space weathering. They might be linked to CM chondrites;

  • Padua-type families (Padua, Chloris, Misa, Chaldaea, and Astrid) also have strong NUV absorption but slightly lower than the Veritas type. They likely contain Fe-poor or Mg-rich phyllosilicates;

  • Erigone-type families (Erigone, Nemesis, Meliboea, Klio, Naema, Sulamitis, Vibilia, and Konig) exhibit diverse taxonomic classifications. They are likely linked to CM chondrites;

  • Euphyrosyne-type families (Euphyrosyne, Ursula, Phaeo, Sylvia, Karma, and Postrema) are dominated by P types but also encompass members with high NUV absorption. They likely have a complex composition including unhydrated and hydrated silicates, water ice, and an organic-rich matrix;

  • Lixiaohua-type families (Lixiaohua, Fringilla, Luthera, and Emma) exhibit red spectral slopes, weak NUV absorption, and low albedo. They might be similar to cometary nuclei in composition.

Overall, the spectral features and densities of these families provide insights into their possible compositions and relationships with various meteorite types. This analysis highlights that NUV absorption in particular is a powerful tool for constraining the abundance of phyllosilicates and Mg/Fe in the phyllosilicates. However, some families lack a perfect match among known meteorites, suggesting the possible sampling bias of our meteorite collection.

There are space mission targets that have potential links to the asteroid families: Ryugu and Bennu to the Polana–Eulalia complex family, and Phaethon to the Pallas family. NUV absorption is an accessible discriminator for making further connections between the targets and the families.

Acknowledgements

ET, FTR, JdL, and JL acknowledge support from the Agencia Estatal de Investigación del Ministerio de Ciencia e Innovación (AEI-MCINN) under the grant “Hydrated minerals and organic compounds in primitive asteroids” (PID2020-120464GB-I00/doi:10.13039/501100011033). ET acknowledges support from the JAXA Hayabusa2# International Visibility Enhancement Project. JdL acknowledges financial support from the Spanish Ministry of Science and Innovation (MICINN) through the Spanish State Research Agency, under Severo Ochoa Programe 2020–2023 (CEX2019-000920-S). MP was supported by the grant of the Romanian National Authority for Scientific Research – UEFISCDI, project number PN-III-P2-2.1-PED-2021-3625. SH was supported by the Hypervelocity Impact Facility (former name: The Space Plasma Laboratory), ISAS, JAXA.

Appendix A Spectral characteristics of each asteroid family

Figure A.1 shows the SDSS spectra of each asteroid family normalized at the g filter. The black lines show the median spectrum for each family.

thumbnail Fig. A.1

SDSS spectra for each family. The black lines indicate the median spectra of families. Note that the spectra are without offset to combine with ECAS dataset, i.e., original SDSS spectra.

A.1 Inner main belt

Polana–Eulalia complex family. The Polana–Eulalia complex family comprises ~80% of F types in our 376 samples. This result is consistent with the results reported by Zellner et al. (1985); Tatsumi et al. (2022); Delbo et al. (2023), which highlighted very flat spectra in the NUV region. Interestingly, Bennu and Ryugu also exhibit flat NUV reflectance spectra, corroborating their association with the Polana–Eulalia complex family (Tatsumi et al. 2022). Further discussions will be provided in Sec. 6.4. Our analysis suggests the negative HYD value (−0.01) and a concave shape in NIR, both suggesting the less hydrated state of this family. Although Morate et al. (2016, 2018, 2019) focused solely on VIS wavelength and employed the Bus taxonomy for asteroid classification, we can still compare the visible reflectance spectra.

Pinilla-Alonso et al. (2016) reported the average NIR slope of 0.092 ± 0.060 (μm−1) in the wavelength range of 0.9–2.2 μm, having concave shapes. This agrees well with our analysis showing 0.06 (μm−1) in JY and 0.14 (μm−1) in KsJ, suggesting a concave shape in average.

Erigone family. The Erigone family comprises a variety of taxonomic types, with the primary constituents being C and G types. Morate et al. (2016) found that 58% of the primitive members show the 0.7 μm absorption, which is consistent with our moderate NUV absorption, 0.8 ± 0.5 (μm−1), and the slight positive median HYD value (0.003). Additionally, they reported a VIS slope for this family as 0.18 ± 0.20 (μm−1), which is consistent with our value, 0.010.12+0.08$\[0.01_{-0.12}^{+0.08}\]$ (μm−1). The NIR slopes of Erigone family members were measured by Harvison et al. (2024), which reporting the average of 0.08 (μm−1) having a wide variety from -0.069 to 0.184 (μm−1) in the wavelength between 1.1 to 2.2 μm mostly with convex shapes. Our analysis, based on the MOVIS data, on 14 members also suggests the convex NIR shape of the Erigone family. The Erigone family was estimated to have formed 130 Myr ago from a parent body with a diameter of ~110 km (Bottke et al. 2015).

Clarissa family. Although only four members of the Clarissa family were found in our sample, all were classified as F types. Morate et al. (2018) observed 33 members and found that the majority exhibited no signs of hydration, consistent with our results of shallow or no NUV absorption and the negative HYD value (−0.02). Additionally, they reported the visible spectral slope for this family as 0.02 ± 0.20 (μm−1), which is consistent with our value of 0.280.04+0.20$\[-0.28_{-0.04}^{+0.20}\]$, similar to that of the Polana–Eulalia complex family. Near-infrared spectroscopy of the Clarissa family has not been conducted due to its faintness, but only the largest body (302) Clarissa was observed in near-infrared (Arredondo et al. 2021b). They noted that the near-infrared spectrum of (302) Clarissa was very similar to those of the Polana–Eulalia family members, having a concave shape (Arredondo et al. 2021b). The similarity of the taxonomic constitution is in agreement with the previous VIS and NIR spectroscopic observations. The Clarissa family was estimated to have formed ~60 Myr ago (Bottke et al. 2015) from a parent body diameter of ~40 km (Brož et al. 2013). Despite the differences in parent body diameters of the Polana–Eulalia complex and Clarissa families by a factor of 2, they may share similar compositions based on NUV-NIR spectra.

Sulamitis family and Klio family. The Sulamitis and Klio families exhibit similar taxonomic compositions, with both showing comparable visible spectra and NUV absorption characteristics. Furthermore, both families show positive HYD values, suggesting the presence of Fe-rich phyllosilicates. They both display moderate NUV absorptions and redder spectral slopes in VIS than other primitive families in the IMB. Additionally, the visible and near-infrared spectroscopic studies suggest similarities between these two families (Morate et al. 2018, 2019; Arredondo et al. 2020, 2021b). However, there is a difference in the proportions of members exhibiting the 0.7 μm absorption, with 60% for the Sulamitis family and 23% for the Klio family (Morate et al. 2018, 2019).

Chaldaea family. The Chaldaea family shares a similar taxonomic constitution to the Erigone family, and its NUV absorption level is also comparable to that of the Erigone family. Our spectral slope range of 0.020.12+0.09$\[0.02_{-0.12}^{+0.09}\]$ (μm−1) is consistent with the previous work, 0.09 ± 0.13 (μm−1) (Morate et al. 2019). While the previous spectroscopic study reported a high presence 79% of members with the 0.7-μm absorption feature (Morate et al. 2019), our analysis suggests a slightly negative HYD value (−0.003). The convex NIR slope of the Chaldaea family, 0.085 ± 0.042 (μm−1) in the wavelength between 0.95 and 2.3 μm (Arredondo et al. 2021a), agrees with that of the Erigone family. However, it should be noted that this NIR slope value is not unique for the Chaldaea and Erigone families, but also other primitive inner main belt families within error (Harvison et al. 2024). The NUV-NIR similarities between the Erigone and Chaldaea families suggest a compositional similarity between the two parent bodies of these families.

A.2 Middle main belt

Nemesis, Padua, Chloris, and Misa families. The Nemesis, Padua, Chloris, and Misa families primarily consist of G-type members with high NUV absorption. However, the groundbased observations of the Nemesis and Padua family members indicate the absence of 0.7 μm absorption (Morate et al. 2021). Nevertheless, the observation of 3 μm region of (128) Nemesis by AKARI revealed a sharp and clear OH absorption centered at 2.74 μm (Usui et al. 2019). While 0.7 μm was not detected, the NUV absorption could indicate OH absorption. This implies that the parent body of the Nemesis family might be composed of Mg-rich phyllosilicates. Similarly, the parent body of the Misa family may also be composed of Mg-rich phyllosilicates. However, observations of (569) Misa, the largest body among the Misa family, in VIS have yielded contradictory results, with one showing 0.7 μm absorption (Bus & Binzel 2002b) and the other not (Lazzaro et al. 2004). In contrast, the Padua and Chloris families could be different compositions due to their distinct taxonomic constitutions and the redder VIS slopes than the Nemesis and Misa families. Furthermore, observations of the largest bodies in these families, (363) Padua and (410) Chloris, reveal no 0.7 μm band (Bus & Binzel 2002b) and a clear 0.7 μm band (Sawyer 1991; Bus & Binzel 2002b), respectively. Ground-based observations suggest that while (363) Padua displays a sharp-type 3-μm absorption shape (Rivkin et al. 2015), the central wavelength was not measured. Consequently, further detailed observations in the 0.7 μm and 3 μm regions are necessary to have a solid conclusion on their composition in the future.

Adeona and Dora families. The Adeone and Dora families are characterized by their higher proportion of bluer members, with F and B types comprising ~50% of their members. The spectroscopic studies reported that the Adeona and Dora families predominantly consist of members with the 0.7-μm absorption (Mothé-Diniz et al. 2005; Morate et al. 2021), consistent with our analyses indicating high NUV absorptions, positive HYD values, and convex shapes in their NIR spectra.

Furthermore, the largest member (145) Adeona within its namesake family was found to have a deep and sharp OH absorption feature centered at 2.76 μm, which is slightly longer in wavelength compared to (128) Nemesis. This difference between the Adeona-Dora and Nemesis families might stem from the compositional variations in the Fe-Mg phases of phyllosilicates, suggesting that the Adeona family is more Fe-rich while the Nemesis family is more Mg-rich.

The Adeona family was formed 0.8 ± 0.2 Gyr ago (Milani et al. 2017) from a parent body estimated to be 170 – 190 km in diameter (Durda et al. 2007; Brož et al. 2013). The Dora family, formed through a catastrophic disruption event, is slightly younger with an estimated age of 0.5 ± 0.2 Gyr (Brož et al. 2013; Spoto et al. 2015). Its parent body is thought to have been 90 – 170 km in diameter (Durda et al. 2007; Brož et al. 2013). Notably, the Adeona family hosts the active asteroid P/2016 G1 (Hsieh et al. 2018) and may be linked to 259P/Garradd (Xin et al. 2024).

Brucato family. The Brucato family is characterized by a negative VIS slope, and high NUV absorption. The taxonomic constitution is similar to those of the Adeona and Dora families mentioned earlier. This family is predominantly composed of bluer members, B and F types, whereas a significant population of C types was also found. While no prior spectroscopic studies exist for comparison, the spectral characteristics revealed by our analyses point towards a high degree of hydration of this family.

Astrid and Konig families. The Astrid and Konig families are primarily composed of C types, although they have relatively small numbers of family members. The Asterid family exhibits one of the highest HYD values, although due to its limited sample size, there may be considerable uncertainty associated with this analysis. The largest member, (1128) Astrid, within its namesake family displays a hint of the 0.7 μm band (Bus & Binzel 2002b). The parent bodies of these two families were estimated as 30 – 40 km (Brož et al. 2013). This relatively small size may contribute to the homogeneity of their composition without differentiation.

Hoffmeister and Mitidika families. The Hoffmeister and Mitidika families are primarily composed of ~60% F type members with flat to slightly blue spectra in VIS. The spectroscopic study suggested the absence of members with the 0.7 μm band within the Hoffmeister family (Morate et al. 2021)while the Mitidika family has not yet been observed spectroscopically. The NIR2 slope value is higher compared with the other F-type dominant families. The dynamical simulation suggested the Hoffmeister family is 22040+60$\[220_{40}^{+60}\]$ Myr old (Carruba et al. 2016) and formed from a parent body diameter of 90–130 km (Brož et al. 2013). Masiero et al. (2013) suggested (272) Antonia as the largest body of the Hoffmeister family. Nonetheless, its albedo of 0.13–0.18 (Mainzer et al. 2019), and taxonomic type, X, of (272) Antonia, differ from the average profile of other members, indicating (272) Antonia may not belong to the Hoffmeister family. Similarly, (2262) Mitidika’s high albedo of ~0.2 (Mainzer et al. 2019) suggests that it may not belong to the namesake family, and instead, (404) Arsinoe or (5079) Brubeck could be the proper largest member of the Mitidika family (Nesvorný et al. 2015). Due to their limited spectral information, estimating their composition is challenging. However, the spectral and parent-body-size similarities with the Polana–Eulalia complex family suggest possible similar compositions between the Hoffmeister, Mitidika, and Polana–Eulalia complex families.

Vibilia family. The Vibilia family displays a taxonomic constitution similar to the Erigone family, with moderate NUV absorption and an almost neutral HYD value (0.001). The largest body, (144) Vibilia, exhibits the 0.7 μm absorption (Bus & Binzel 2002b). This family is lately considered to be the same as the (2782) Leonidas family (Nesvorný et al. 2015; Vinogradova 2019).

Phaeo and Karma families. These families contain a significant number of P types. Particularly, the Phaeo and Postrema families have ~10% of D-type members. These families are intriguing because they include both D types and G types, which exhibit very different spectral shapes. The largest body, (322) Phaeo, within its namesake family was classified as X type with a hint of the 1-μm band (Bus & Binzel 2002b).

The Karma family also comprises predominantly P types but lacks D type. It includes one B-type and one G-type member out of nine samples. The reflectance spectrum of the largest member, (3811) Karma, has not been observed yet.

The diameters of (322) Phaeo and (3811) Karma are estimated as ~70 km and ~30 km, respectively (Mainzer et al. 2019). Therefore, the parent bodies of these families are not likely to exceed 100 km in diameter. The limited information makes it challenging to estimate the composition at the moment, highlighting the need for additional observations.

Postrema family. The Postrema family consists primarily of P types and no/little NUV absorption members, such as F and D types. While the majority of the members exhibit red spectra, the largest member, (1484) Postrema, within its namesake family, displays a flat to blue spectrum classified as B type in the Bus taxonomy (Bus & Binzel 2002b), possibly B or F type in the Tholen taxonomy. This discrepancy might indicate potential differences in compositions between the core and crust of the parent body.

Pallas family. The Pallas family stands out due to its predominantly B type composition and high albedo. (2) Pallas, the largest member, has a diameter of ~513 km (Marsset et al. 2020) and is known to have a blue spectrum in VIS wavelength (e.g., Vilas et al. 1993). Spectroscopic analysis of nine other members in this family has confirmed similarly blue in the VIS wavelength region (de León et al. 2010). Our results indicate that except one out of 21 members in the Pallas family display negative VIS slopes. Furthermore, the NUV absorption observed in the Pallas family members is relatively strong compared with the Polana–Eulalia complex and Theobalda families which also have negative VIS slopes. Additionally, the HYD value of the Pallas family is neutral (−0.000). These spectral features suggest a potential presence of hydrated minerals. Near-infrared (0.8 – 2.5 μm) reflectance spectra of five family members were observed, revealing that the Pallas family is concentrated in the blue, negative slope, group (G4 and G5), which are similar to unusual CM and CV3 meteorites (de León et al. 2012). However, the space telescope AKARI revealed its sharp and deep OH absorption, suggesting the presence of phyllosilicates (Usui et al. 2019). Furthermore, the active asteroid P/2017 S8 (PANSTARRS) was found to belong to the Pallas family suggesting the water sublimation (Xin et al. 2024).

A.3 Outer main belt

Hygiea family. Our analysis identified 306 objects belonging to the Hygiea family. This family is dominated by F and B types, comprising over 60%, characterized by negative VIS slopes. A notably minor fraction of the family consists of P and D types, comprising less than 10% in total. Mothé-Diniz et al. (2001) also reported a predominance of members with negative VIS slopes, although they only observed the VIS part of spectra Furthermore, our results on the albedo and the visible spectral slope agree with the previous study by De Prá et al. (2020b). The largest family member (10) Hygiea is the 4th most massive main-belt asteroid. The Hygiea family was estimated to be formed 1 − 2 Gy ago (Spoto et al. 2015) from the parent body of 410 − 490 km in diameter (Brož et al. 2013; Durda et al. 2007; De Prá et al. 2020b). They investigated the 0.7-μm absorption and found ~15% of members exhibit this band. Our analysis suggests a moderate NUV absorption strength and a negative HYD value and is consistent with a small fraction of the members displaying the 0.7-μm band. Additionally, our analysis indicates a concave shape in the NIR spectra of this family, aligned with the NIR spectra of (10) Hygiea observed by DeMeo & Binzel (2008).

The largest family member (10) Hygiea is the 4th most massive main belt asteroid with a diameter of 433 km (Vernazza et al. 2021). The Hygiea family was estimated to have formed 1 – 2 Gy ago (Spoto et al. 2015) from the parent body of 410 490 km in diameter (Brož et al. 2013; Durda et al. 2007; De Prá et al. 2020b). Because of the spherical shape of (10) Hygiea, the numerical simulation supported the hypothesis that the familyforming collision was caused by an impactor of 100 km in size (Vernazza et al. 2021). It has been suggested that this impact could have fully fragmented the parent body, allowing the material to behave as a fluid and subsequently reaccumulate to form (10) Hygiea. If this scenario holds, the composition of the parent body of the Hygiea family would have been thoroughly mixed onto (10) Hygiea. While (10) Hygiea is classified as C type by Tholen (1984), most members of its family exhibit bluer spectra, consistent with B and F types. The presence of these bluer members may indicate that they escape from the intensive process during the family-forming impact and subsequent reaccumulation. Moreover, the bluer composition may have been inherited internally by (10) Hygiea.

In contrast, (10) Hygiea is known to have a similar composition to the dwarf planet (1) Ceres based on the NIR and midinfrared (MIR) observations (Takir & Emery 2012; Vernazza et al. 2017). The density of (10) Hygiea is 1.94 ± 0.25 g/cm3 (Vernazza et al. 2021) which also aligns with the density of (1) Ceres, estimated as 2.08 ± 0.04 g/cm3 (Thomas et al. 2005). While the parent body of Hygiea family likely experienced a large impact event, no family associated with (1) Ceres has been confirmed yet. The similarity between (10) Hygiea and (1) Ceres is puzzling considering their very different evolutionary histories. The surface composition of (1) Ceres was investigated by the Dawn spacecraft (e.g., de Sanctis et al. 2015) has not revealed an exact match with known meteorite collections on Earth. The surface of (1) Ceres was primarily composed of the Mg-rich serpentine and the ammoniated clay (de Sanctis et al. 2015). Furthermore, the remote-sensing observation on Cerealia Facula suggested the presence of salty water internally to this day (De Sanctis et al. 2020), indicating ongoing differentiation of (1) Ceres. Interestingly one active asteroid (62412) 2000 SY178 is linked to the Hygiea family, suggesting the presence of ice in the parent body (Sheppard & Trujillo 2015). Similar internal layering with different phases could potentially explain the spectroscopic difference between (10) Hygiea and other members.

Themis family and Beagle sub-family. The Themis family is distinct from The Hygiea family in its taxonomic composition. The Themis family members are predominantly B types, with a small proportion of F types compared with the Hygiea family. Our previous spectroscopic work also revealed a stronger NUV absorption in Themis compared to the Polana–Eulalia complex family (Tatsumi et al. 2022). This abundance of Themis family members has made it one of the most well-studied families. The Themis family was estimated to have formed > 2 Gyr ago (Marzari et al. 1995; Spoto et al. 2015) from a parent body 270 – 450 km in diameter(Brož et al. 2013; Durda et al. 2007; De Prá et al. 2020b).

The early spectroscopic observation detected the 0.7-μm band in some members, potentially indicating aqueous alteration (Florczak et al. 1999). However, a more recent study found that only 13% of members exhibit this band and the majority of the members do not (De Prá et al. 2020b). Among the 15 family members Marsset et al. (2016) found diversity in NIR spectra, mostly with concave-up shapes consistent with chondritic porous interplanetary dust particles. The concave-up shapes are consistent with another observation by Ziffer et al. (2011). Thermal infrared spectra 5 – 14 μm of eight Themis family members revealed similarity to Trojan asteroids and cometary dust (Licandro et al. 2012). These MIR spectra suggest either small silicate grains embedded in a relatively transparent matrix, or a fairy-castle-like surface structure (Licandro et al. 2012). Overall, the NIR and MIR observations imply a very primitive, unaltered surface for the Themis family, contrasting with the NUV-VIS spectra suggesting a moderately hydrated surface.

The 3.1 μm rounded absorption of (24) Themis, firstly reported by Campins et al. (2010), was interpreted as water-ice-coated pyroxene grains and amorphous carbon (Rivkin & Emery 2010). The space telescope AKARI confirmed both 2.7-μm and 3.1-μm absorptions in (24) Themis. While the 3.1-μm feature in Ceres and (10) Hygiea is attributed to ammoniated phyllosilicates, the longer central wavelength (3.08 μm) of the absorption feature points towards water ice than (Kurokawa et al. 2022). The bulk density of (24) Themis was estimated as 1.31 ± 0.62 g/cm3 (Vernazza et al. 2021). This density is slightly lower than those of Ceres and (10) Hygiea, further supporting the presence of abundant water. Additionally, multiple active asteroids 133P/Elst-Pizarro, 176P/LINEAR, 288P/(300163), and 433P/(248370) are possibly belong to the Themis and its sub-families (Novaković et al. 2022), supporting volatile-rich nature of their parent body. Despite potential water abundance differences between the Hygiea family and the Themis family, their NUV-NIR spectral profile and the parent body sizes are similar. The key difference lies in the F-type abundance, possibly indicating a lower water content for the Hygiea family’s parent body than for the Themis family’s parent body.

The Beagle family presents an intriguing case as a sub-family within the larger Themis family (Nesvorný et al. 2008). It likely originated from a catastrophic disruption event much younger than the Themis family itself. Estimates suggest a formation time of less than 100 Myr for Beagle sub-family (Nesvorný et al. 2008; Carruba 2019), compared to over 2 Gyr for the Themis family (Marzari et al. 1995; Brož et al. 2013; Spoto et al. 2015). The spectroscopic observations by Fornasier et al. (2016) reveal a slightly bluer spectral signature for the Beagle family compared to the Themis family, which is consistent with our B-type dominant configuration of the Beagle family. Despite this drastic difference in taxonomic composition, the Beagle family’s spectral variation (~60% B type asteroids) falls within that was observed within the Themis family itself.

Theobalda family. The Theobalda family stands out for its unique taxonomic composition: a dominance of F types (85%) and all members except one exhibit blue VIS slopes, making this family the bluest family on average in our study. The HYD value for this family is negative. While NUV absorption is weak and the HYD value is negative, the Theobalda family shows a concave shape in its NIR spectra. This suggests the potential presence of the 3 μm absorption band. Furthermore, this family is linked to multiple active asteroids, P/2016 J1-A/B (PANSTARRS), 427P/ATLAS, 455P/PANSTARRS, and P/2019 A3 (PANSTARRS) (Hsieh et al. 2018, 2023; Xin et al. 2024), hinting at a volatile-rich composition for the parent body. Notably, this family formed extreme recently, ~6.9 Myr ago (Novaković 2010). This recent formation could explain the presence of ice on the surface and its ongoing sublimation.

Sylvia, Lixiaohua, Fringilla, and Luthera families. The Lixiaohua, and Fringilla families share a distinct taxonomic composition: a dominance of P types (> 50%) alongside members with no or little NUV absorption, F and D types. Notably, the largest bodies of these families are also classified as P type. This low NUV absorption suggests a low degree of hydration in their parent bodies. Among these families, only the Lixiaohua family has been spectroscopically studied in detail. De Prá et al. (2020a) found no evidence of the 0.7-um absorption band in any members, aligning well with our low NUV absorption. This family is known to host two active asteroids, 313P/Gibbs and 358P/PANSTARRS (P/2012 T1) (Hsieh et al. 2013, 2018). Moreover, the Luthera was recently also found to link with an active asteroid 426P/PANSTARRS (Xin et al. 2024). This connection suggests a potentially volatile-rich nature of the Lixiaohua and Luthera families’ parent bodies.

Emma family. The Emma family comprises the half of P type and half of F type. The Emma family is also unique because of a positive HYD value (0.002), even though it includes mostly low NUV absorption members. These unique characteristics may indicate a possible rare composition of the Emma family’s parent body. Further observations are needed.

Ursula, Euphrosyne, and Sylvia families. The Ursula, Euphrosyne, and Sylvia families have a significant presence of P types, but unlike the Lixiaohua, Fringilla, and Luthera families discussed earlier, these families also host members with strong NUV absorptions (B and G types). This suggests a potential for hydration in some members of these families. However, De Prá et al. (2020b) found no evidence of the 0.7-μm absorption band in any members of the Ursula family studied, which could be due to the limited sample size. The Euphrosyne family is associated with an active asteroid P/2016 P1(PANSTARRS), suggesting the presence of ice (Xin et al. 2024).

Veritas and Inarradas families. The Veritas and Inarradas families are composed of over 50% G-type asteroids, indicating a high degree of hydration within these families. This conclusion is also supported by our analysis, which identified these two families as having the highest HYD values. Furthermore, Mothé-Diniz et al. (2005); De Prá et al. (2020b) have investigated the Veritas family spectroscopically and reported that ~ 75% of its members exhibit the 0.7-μm absorption band, which aligns well with our analysis. Ziffer et al. (2011) investigated the NIR spectra of the Veritas family, reporting the NIR slope of 0.13 ± 0.31 μm−1 between 0.8 and 2.4 μm. Notably, they observed convex (concave down) NIR spectral shapes for all six members studied. Our analysis also suggests an average convex shape of the Veritas family.

The Veritas family is one of the youngest known asteroid families, formed approximately 8.3 Myr ago from a parent body ~140 km in diameter (Nesvorný et al. 2003). Due to this young age, the Veritas family members are tightly clustered in the proper element space, reducing the likelihood of interlopers contaminating the population. Thus, the minor presence of less hydrated members (F, P, and D) is also inherited from the Veritas family’s parent body. The Inarradas family is a more recent discovery, identified as a small asteroid family (Milani et al. 2014). Further studies investigating the spectral properties of both families may reveal compositional similarities, further solidifying the connection.

Alauda family. The Alauda family is characterized by a negative VIS slope and moderate NUV absorption strength on average. This family is composed of various taxonomic types, whereas B and F types occupy ~50% of its members. No previous spectroscopic work focused on this family, the largest member (702) Alauda has been investigated well in spectroscopy. (702) Alauda does not exhibit the 0.7-μm absorption band (Bus & Binzel 2002b; Lazzaro et al. 2004), but displays an absorption around 3.1-μm, potentially indicating the presence of water ice(Rivkin et al. 2022). The finding of the active asteroid 324P/La Sagra within the family (Hsieh et al. 2018) also supports the volatile-rich composition of this family. Our analysis indicates the straight to convex shape of NIR spectra on average, suggesting potential hydration features at the 3 μm region. The absence of 0.7-μm and the NIR spectral shape may imply Mg-rich phyllosilicate composition. The Alauda family was estimated to form 640 ± 50 Myr ago (Carruba et al. 2016) from a parent body 220 – 330 km in diameter (Brož et al. 2013; Durda et al. 2007).

Meliboea family. Like the Alauda family, the Meliboea family is also characterized by a negative VIS slope and moderate NUV absorption strength on average. However, this family is dominated by C and P types. The positive HYD value suggests a high degree of hydration of this family, consistent with the previous study reported 78% of the studied members exhibit the 0.7-μm absorption band, indicative of hydration (Mothé-Diniz et al. 2005). Our analysis indicates the convex shape of NIR spectra on average, pointing also to hydration. Furthermore, (137) Meliboea, the largest body in its namesake family, shows a sharp-type absorption feature in the 3 μm region (Rivkin et al. 2015). The Meliboea family was estimated to form 640 ± 10 Myr ago (Carruba et al. 2016) from a parent body 170 – 290 km in diameter (Brož et al. 2013; Durda et al. 2007).

Naema families. The Naema family shows a mixed taxonomic constituent similar to the Erigone family. Our analysis suggests a moderate NUV absorption (0.9) and positive HYD (0.01), pointing towards a high degree of hydration. While spectroscopic data for the Naema family is limited, the largest member, (845) Naema, displays the 0.7-μm absorption band and strong NUV absorption (Bus & Binzel 2002b), which align well with our analysis.

Appendix B Additional Materials

Table B.1

Spectral characteristics of primitive asteroid families.

thumbnail Fig. B.1

Inner main belt primitive families.

thumbnail Fig. B.2

Middle main belt primitive families.

thumbnail Fig. B.3

Outer main belt primitive families.

Table B.2

Spectral and physical characteristics of primitive asteroid families and the largest members.

Table B.3

Links between family types and compositions (See Sec. 5 for further discussions.)

References

  1. Alexander, C. M. O., Bowden, R., Fogel, M. L., et al. 2012, Science, 337, 721 [NASA ADS] [CrossRef] [Google Scholar]
  2. Alí-Lagoa, V., Licandro, J., Gil-Hutton, R., et al. 2016, A&A, 591, A14 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Ammannito, E., DeSanctis, M. C., Ciarniello, M., et al. 2016, Science, 353, aaf4279 [NASA ADS] [CrossRef] [Google Scholar]
  4. Andrews, S. M., Huang, J., Pérez, L. M., et al. 2018, ApJ, 869, L41 [NASA ADS] [CrossRef] [Google Scholar]
  5. Arredondo, A., Lorenzi, V., Pinilla-Alonso, N., et al. 2020, Icarus, 335, 113427 [NASA ADS] [CrossRef] [Google Scholar]
  6. Arredondo, A., Campins, H., Pinilla-Alonso, N., et al. 2021a, Icarus, 354, 114028 [NASA ADS] [CrossRef] [Google Scholar]
  7. Arredondo, A., Campins, H., Pinilla-Alonso, N., et al. 2021b, Icarus, 358, 114210 [NASA ADS] [CrossRef] [Google Scholar]
  8. Beck, P., Schmitt, B., Potin, S., Pommerol, A., & Brissaud, O. 2021, Icarus, 354, 114066 [NASA ADS] [CrossRef] [Google Scholar]
  9. Bland, P. A., & Travis, B. J. 2017, Sci. Adv., 3, e1602514 [NASA ADS] [CrossRef] [Google Scholar]
  10. Bottke, W. F., Vokrouhlickỳ, D., Walsh, K. J., et al. 2015, Icarus, 247, 191 [CrossRef] [Google Scholar]
  11. Brož, M., Morbidelli, A., Bottke, W. F., et al. 2013, A&A, 551, A117 [Google Scholar]
  12. Bus, S. J., & Binzel, R. P. 2002a, Icarus, 158, 146 [Google Scholar]
  13. Bus, S. J., & Binzel, R. P. 2002b, Icarus, 158, 106 [CrossRef] [Google Scholar]
  14. Campins, H., Hargrove, K., Pinilla-Alonso, N., et al. 2010, Nature, 464, 1320 [NASA ADS] [CrossRef] [Google Scholar]
  15. Carruba, V. 2019, Planet. Space Sci., 166, 90 [Google Scholar]
  16. Carruba, V., Novaković, B., & Aljbaae, S. 2016, MNRAS, 465, 4099 [Google Scholar]
  17. Carruba, V., De Oliveira, E., Rodrigues, B., & Requena, I. 2018, MNRAS, 479, 4815 [NASA ADS] [CrossRef] [Google Scholar]
  18. Carvano, J., Hasselmann, P., Lazzaro, D., & Mothé-Diniz, T. 2010, A&A, 510, A43 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  19. Casagrande, L., Ramírez, I., Meléndez, J., & Asplund, M. 2012, ApJ, 761, 16 [NASA ADS] [CrossRef] [Google Scholar]
  20. Castillo-Rogez, J. C., & McCord, T. B. 2010, Icarus, 205, 443 [NASA ADS] [CrossRef] [Google Scholar]
  21. Cellino, A., Bus, S. J., Doressoundiram, A., et al. 2002, Asteroids III (Tucson: University of Arizona Press), 633 [CrossRef] [Google Scholar]
  22. Charnoz, S., Avice, G., Hyodo, R., Pignatale, F. C., & Chaussidon, M. 2021, A&A, 652, A35 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Clark, B., Sen, A., Zou, X.-D., et al. 2023, Icarus, 400, 115563 [NASA ADS] [CrossRef] [Google Scholar]
  24. Clayton, R. N., & Mayeda, T. K. 1999, Geochim. Cosmochim. Acta, 63, 2089 [CrossRef] [Google Scholar]
  25. Cloutis, E. A., Hiroi, T., Gaffey, M. J., Alexander, C. M. O. D., & Mann, P. 2011, Icarus, 212, 180 [NASA ADS] [CrossRef] [Google Scholar]
  26. Cloutis, E. A., Hudon, P., Hiroi, T., & Gaffey, M. J. 2012, Icarus, 220, 586 [NASA ADS] [CrossRef] [Google Scholar]
  27. de León, J., Campins, H., Tsiganis, K., Morbidelli, A., & Licandro, J. 2010, A&A, 513, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. de León, J., Pinilla-Alonso, N., Campins, H., Licandro, J., & Marzo, G. 2012, Icarus, 218, 196 [CrossRef] [Google Scholar]
  29. de León, J., Pinilla-Alonso, N., Delbo, M., et al. 2016, Icarus, 266, 57 [CrossRef] [Google Scholar]
  30. de León, J., Campins, H., Morate, D., et al. 2018, Icarus, 313, 25 [CrossRef] [Google Scholar]
  31. De Prá, M., Licandro, J., Pinilla-Alonso, N., et al. 2020a, Icarus, 338, 113473 [CrossRef] [Google Scholar]
  32. De Prá, M. N., Pinilla-Alonso, N., Carvano, J., et al. 2020b, A&A, 643, A102 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  33. de Sanctis, M. C., Ammannito, E., Raponi, A., et al. 2015, Nature, 528, 241 [Google Scholar]
  34. De Sanctis, M. C., Ammannito, E., Raponi, A., et al. 2020, Nat. Astron., 4, 786 [CrossRef] [Google Scholar]
  35. Delbo, M., Avdellidou, C., & Walsh, K. J. 2023, A&A, 680, A10 [Google Scholar]
  36. DellaGiustina, D., Burke, K., Walsh, K., et al. 2020, Science, 370, eabc3660 [NASA ADS] [CrossRef] [Google Scholar]
  37. DeMeo, F., & Binzel, R. P. 2008, Icarus, 194, 436 [NASA ADS] [CrossRef] [Google Scholar]
  38. DeMeo, F., & Carry, B. 2013, Icarus, 226, 723 [NASA ADS] [CrossRef] [Google Scholar]
  39. DeMeo, F. E., Binzel, R. P., Slivan, S. M., & Bus, S. J. 2009, Icarus, 202, 160 [Google Scholar]
  40. DeMeo, F. E., Binzel, R. P., Carry, B., Polishook, D., & Moskovitz, N. A. 2014, Icarus, 229, 392 [NASA ADS] [CrossRef] [Google Scholar]
  41. Desch, S. J., Kalyaan, A., & O’D Alexander, C. M. 2018, ApJS, 238, 11 [NASA ADS] [CrossRef] [Google Scholar]
  42. Dodson-Robinson, S. E., & Salyk, C. 2011, ApJ, 738, 131 [CrossRef] [Google Scholar]
  43. Durda, D. D., Bottke, W. F., Nesvorný, D., et al. 2007, Icarus, 186, 498 [NASA ADS] [CrossRef] [Google Scholar]
  44. Ehlmann, B. L., Hodyss, R., Bristow, T. F., et al. 2018, Meteorit. Planet. Sci., 53, 1884 [NASA ADS] [CrossRef] [Google Scholar]
  45. Florczak, M., Lazzaro, D., Mothé-Diniz, T., Angeli, C., & Betzler, A. 1999, A&ASS, 134, 463 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  46. Flynn, G. J., Consolmagno, G. J., Brown, P., & Macke, R. J. 2018, Geochemistry, 78, 269 [NASA ADS] [CrossRef] [Google Scholar]
  47. Fornasier, S., Lantz, C., Perna, D., et al. 2016, Icarus, 269, 1 [CrossRef] [Google Scholar]
  48. Fukugita, M., Ichikawa, T., Gunn, J. E., et al. 1996, AJ, 111, 1748 [Google Scholar]
  49. Gayon-Markt, J., Delbo, M., Morbidelli, A., & Marchi, S. 2012, MNRAS, 424, 508 [NASA ADS] [CrossRef] [Google Scholar]
  50. Gradie, J., & Veverka, J. 1980, Nature, 283, 840 [NASA ADS] [CrossRef] [Google Scholar]
  51. Grimm, R. E., & McSween, H. Y. 1993, Science, 259, 653 [NASA ADS] [Google Scholar]
  52. Grimm, R. E., & McSween Jr, H. Y. 1989, Icarus, 82, 244 [NASA ADS] [CrossRef] [Google Scholar]
  53. Gundlach, B., & Blum, J. 2014, in Asteroids, Comets, Meteors 2014, eds. K. Muinonen, A. Penttilä, M. Granvik, A. Virkki, G. Fedorets, O. Wilkman, & T. Kohout, 191 [Google Scholar]
  54. Hanuš, J., Delbo’, M., Vokrouhlický, D., et al. 2016, A&A, 592, A34 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  55. Harvison, B., De Prá, M., Pinilla-Alonso, N., et al. 2024, Icarus, 412, 115973 [NASA ADS] [CrossRef] [Google Scholar]
  56. Hendrix, A. R., & Vilas, F. 2019, Geophys. Res. Lett., 46, 14307 [NASA ADS] [CrossRef] [Google Scholar]
  57. Hirayama, K. 1918, AJ, 31, 185 [Google Scholar]
  58. Hiroi, T., Pieters, C. M., Zolensky, M. E., & Lipschutz, M. E. 1993, Science, 261, 1016 [NASA ADS] [CrossRef] [Google Scholar]
  59. Hiroi, T., Pieters, C. M., Zolensky, M. E., & Prinz, M. 1996a, in Lunar and Planetary Science Conference, 27, 551 [Google Scholar]
  60. Hiroi, T., Zolensky, M. E., Pieters, C. M., & Lipschutz, M. E. 1996b, Meteorit. Planet. Sci., 31, 321 [NASA ADS] [CrossRef] [Google Scholar]
  61. Holmberg, J., Flynn, C., & Portinari, L. 2006, MNRAS, 367, 449 [NASA ADS] [CrossRef] [Google Scholar]
  62. Hsieh, H. H., Kaluna, H. M., Novaković, B., et al. 2013, ApJ, 771, L1 [NASA ADS] [CrossRef] [Google Scholar]
  63. Hsieh, H. H., Hainaut, O., Novaković, B., et al. 2015, ApJ, 800, L16 [NASA ADS] [CrossRef] [Google Scholar]
  64. Hsieh, H. H., Novaković, B., Kim, Y., & Brasser, R. 2018, AJ, 155, 96 [NASA ADS] [CrossRef] [Google Scholar]
  65. Hsieh, H. H., Micheli, M., Kelley, M. S., et al. 2023, Planet. Sci. J., 4, 43 [NASA ADS] [CrossRef] [Google Scholar]
  66. Ito, T., Ishiguro, M., Arai, T., et al. 2018, Nat. Commun., 9, 2486 [Google Scholar]
  67. Ivezić, Ž., Tabachnik, S., Rafikov, R., et al. 2001, AJ, 122, 2749 [Google Scholar]
  68. Izawa, M. R. M., Cloutis, E. A., Rhind, T., et al. 2019, Icarus, 319, 525 [NASA ADS] [CrossRef] [Google Scholar]
  69. Izidoro, A., Dasgupta, R., Raymond, S. N., et al. 2021, Nat. Astron., 6, 357 [Google Scholar]
  70. Kaplan, H. H., Milliken, R. E., Alexander, C. M. O., & Herd, C. D. 2019, Meteor. Planet. Sci., 54, 1051 [CrossRef] [Google Scholar]
  71. Kareta, T., Reddy, V., Hergenrother, C., et al. 2018, AJ, 156, 287 [NASA ADS] [CrossRef] [Google Scholar]
  72. Kerridge, J. F., Mackay, A. L., & Boynton, W. V. 1979, Science, 205, 395 [NASA ADS] [CrossRef] [Google Scholar]
  73. Kerridge, J. F., Chang, S., & Shipp, R. 1987, Geochim. Cosmochim. Acta, 51, 2527 [NASA ADS] [CrossRef] [Google Scholar]
  74. King, T. V. V., Clark, R. N., Calvin, W. M., Sherman, D. M., & Brown, R. H. 1992, Science, 255, 1551 [NASA ADS] [CrossRef] [Google Scholar]
  75. Kitazato, K., Milliken, R., Iwata, T., et al. 2019, Science, 364, 272 [NASA ADS] [Google Scholar]
  76. Kohout, T., Jenniskens, P., Shaddad, M. H., & Haloda, J. 2010, Meteor. Planet. Sci., 45, 1778 [NASA ADS] [CrossRef] [Google Scholar]
  77. Kruijer, T. S., Kleine, T., & Borg, L. E. 2020, Nat. Astron., 4, 32 [Google Scholar]
  78. Kurokawa, H., Shibuya, T., Sekine, Y., et al. 2022, AGU Adv., 3, e00568 [CrossRef] [Google Scholar]
  79. Lambrechts, M., & Johansen, A. 2012, A&A, 544, A32 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  80. Lazzarin, M., Barucci, M., & Doressoundiram, A. 1996, Icarus, 122, 122 [NASA ADS] [CrossRef] [Google Scholar]
  81. Lazzaro, D., Angeli, C. A., Carvano, J. M., et al. 2004, Icarus, 172, 179 [NASA ADS] [CrossRef] [Google Scholar]
  82. Licandro, J., Hargrove, K., Kelley, M., et al. 2012, A&A, 537, A73 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  83. Lodders, K. 2003, ApJ, 591, 1220 [Google Scholar]
  84. Macke, R. J., Consolmagno, G. J., & Britt, D. T. 2011, Meteorit. Planet. Sci., 46, 1842 [NASA ADS] [CrossRef] [Google Scholar]
  85. MacLennan, E., Toliou, A., & Granvik, M. 2021, Icarus, 366, 114535 [NASA ADS] [CrossRef] [Google Scholar]
  86. Mainzer, A., Bauer, J., Cutri, R., et al. 2019, NASA Planetary Data System, https://doi.org/10.26033/18S3-2Z54 [Google Scholar]
  87. Marchi, S., Ermakov, A. I., Raymond, C. A., et al. 2016, Nat. Commun., 7, 12257 [NASA ADS] [CrossRef] [Google Scholar]
  88. Marchis, F., Descamps, P., Hestroffer, D., & Berthier, J. 2005, Nature, 436, 822 [CrossRef] [Google Scholar]
  89. Marsden, B. G. 1983, in IAUC No. 3881 [Google Scholar]
  90. Marsset, M., Vernazza, P., Birlan, M., et al. 2016, A&A, 586, A15 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  91. Marsset, M., DeMeo, F. E., Binzel, R. P., et al. 2020, ApJS, 247, 73 [NASA ADS] [CrossRef] [Google Scholar]
  92. Martikainen, J., Muinonen, K., Penttilä, A., Cellino, A., & Wang, X. B. 2021, A&A, 649, A98 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  93. Marzari, F., Davis, D., & Vanzani, V. 1995, Icarus, 113, 168 [NASA ADS] [CrossRef] [Google Scholar]
  94. Masiero, J. R., Mainzer, A. K., Bauer, J. M., et al. 2013, ApJ, 770, 7 [Google Scholar]
  95. McMahon, R. G., Banerji, M., Gonzalez, E., et al. 2013, The Messenger, 154, 35 [NASA ADS] [Google Scholar]
  96. McSween, H. Y. 1979, Geochim. Cosmochim. Acta, 43, 1761 [NASA ADS] [CrossRef] [Google Scholar]
  97. Milani, A., & Gronchi, G. F. 2010, Theory of Orbital Determination (Cambridge University Press) [Google Scholar]
  98. Milani, A., Cellino, A., Knežević, Z., et al. 2014, Icarus, 239, 46 [NASA ADS] [CrossRef] [Google Scholar]
  99. Milani, A., Knežević, Z., Spoto, F., et al. 2017, Icarus, 288, 240 [NASA ADS] [CrossRef] [Google Scholar]
  100. Morate, D., de León, J., De Prá, M., et al. 2016, A&A, 586, A129 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  101. Morate, D., de León, J., De Prá, M., et al. 2018, A&A, 610, A25 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  102. Morate, D., de León, J., De Prá, M., et al. 2019, A&A, 630, A141 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  103. Morate, D., de León, J., Licandro, J., et al. 2021, in European Planetary Science Congress, EPSC2021-424 [Google Scholar]
  104. Moroz, L., Arnold, G., Korochantsev, A., & Wäsch, R. 1998, Icarus, 134, 253 [NASA ADS] [CrossRef] [Google Scholar]
  105. Moroz, L., Baratta, G., Strazzulla, G., et al. 2004, Icarus, 170, 214 [CrossRef] [Google Scholar]
  106. Mothé-Diniz, T., Di Martino, M., Bendjoya, P., Doressoundiram, A., & Migliorini, F. 2001, Icarus, 152, 117 [CrossRef] [Google Scholar]
  107. Mothé-Diniz, T., Roig, F., & Carvano, J. 2005, Icarus, 174, 54 [CrossRef] [Google Scholar]
  108. Nesvorny, D. 2020, NASA Planetary Data System, EAR-A-VARGBDET-5-NESVORNYFAM-V3.0 [Google Scholar]
  109. Nesvorný, D., Bottke, W. F., Levison, H. F., & Dones, L. 2003, ApJ, 591, 486 [CrossRef] [Google Scholar]
  110. Nesvornỳ, D., Bottke, W. F., Vokrouhlickỳ, D., Morbidelli, A., & Jedicke, R. 2005, Proc. Int. Astron. Union, 1, 289 [CrossRef] [Google Scholar]
  111. Nesvorný, D., Bottke, W. F., Vokrouhlický, D., et al. 2008, ApJ, 679, L143 [CrossRef] [Google Scholar]
  112. Nesvorný, D., Brož, M., & Carruba, V. 2015, in Asteroids IV, 297 [Google Scholar]
  113. Neumann, W., Jaumann, R., Castillo-Rogez, J., Raymond, C. A., & Russell, C. T. 2020, A&A, 633, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  114. Neveu, M., & Desch, S. J. 2015, Geophys. Res. Lett., 42, 10197 [NASA ADS] [CrossRef] [Google Scholar]
  115. Novaković, B. 2010, MNRAS, 407, 1477 [CrossRef] [Google Scholar]
  116. Novaković, B., Pavela, D., Hsieh, H. H., & Marčeta, D. 2022, MNRAS, 516, 757 [CrossRef] [Google Scholar]
  117. Ormel, C. W., & Klahr, H. H. 2010, A&A, 520, A43 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  118. Ozaki, N., Yamamoto, T., Gonzalez-Franquesa, F., et al. 2022, Acta Astronaut., 196, 42 [NASA ADS] [CrossRef] [Google Scholar]
  119. Park, R. S., Vaughan, A. T., Konopliv, A. S., et al. 2019, Icarus, 319, 812 [NASA ADS] [CrossRef] [Google Scholar]
  120. Parker, A., Ivezić, Ž., Jurić, M., et al. 2008, Icarus, 198, 138 [Google Scholar]
  121. Pieters, C., Nathues, A., Thangjam, G., et al. 2018, Meteor. Planet. Sci., 53, 1983 [NASA ADS] [CrossRef] [Google Scholar]
  122. Pinilla, P., Benisty, M., & Birnstiel, T. 2012, A&A, 545, A81 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  123. Pinilla-Alonso, N., de León, J., Walsh, K., et al. 2016, Icarus, 274, 231 [NASA ADS] [CrossRef] [Google Scholar]
  124. Popescu, M., Licandro, J., Morate, D., et al. 2016, A&A, 591, A115 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  125. Popescu, M., Licandro, J., Carvano, J. M., et al. 2018, A&A, 617, A12 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  126. Potin, S., Beck, P., Usui, F., et al. 2020, Icarus, 348, 113826 [CrossRef] [Google Scholar]
  127. Regály, Z., Juhász, A., Sándor, Z., & Dullemond, C. P. 2012, MNRAS, 419, 1701 [Google Scholar]
  128. Rivkin, A. S. 2012, Icarus, 221, 744 [CrossRef] [Google Scholar]
  129. Rivkin, A. S., & Emery, J. P. 2010, Nature, 464, 1322 [NASA ADS] [CrossRef] [Google Scholar]
  130. Rivkin, A. S., Thomas, C. A., Howell, E. S., & Emery, J. P. 2015, AJ, 150, 198 [CrossRef] [Google Scholar]
  131. Rivkin, A. S., Howell, E. S., & Emery, J. P. 2019, J. Geophys. Res. (Planets), 124, 1393 [NASA ADS] [CrossRef] [Google Scholar]
  132. Rivkin, A. S., Emery, J. P., Howell, E. S., et al. 2022, Planet. Sci. J., 3, 153 [NASA ADS] [CrossRef] [Google Scholar]
  133. Rizos, J. L., de León, J., Licandro, J., et al. 2019, Icarus, 328, 69 [NASA ADS] [CrossRef] [Google Scholar]
  134. Rojo, P., & Margot, J. L. 2011, ApJ, 727, 69 [NASA ADS] [CrossRef] [Google Scholar]
  135. Rubin, A. E. 2011, Icarus, 213, 547 [NASA ADS] [CrossRef] [Google Scholar]
  136. Russell, C. T., Raymond, C. A., Ammannito, E., et al. 2016, Science, 353, 1008 [NASA ADS] [CrossRef] [Google Scholar]
  137. Sawyer, S. R. 1991, PhD thesis, The University of Texas at Austin, USA [Google Scholar]
  138. Sergeyev, A. V., & Carry, B. 2021, A&A, 652, A59 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  139. Shaddad, M. H., Jenniskens, P., Numan, D., et al. 2010, Meteor. Planet. Sci., 45, 1557 [NASA ADS] [CrossRef] [Google Scholar]
  140. Sheppard, S. S., & Trujillo, C. 2015, AJ, 149, 44 [NASA ADS] [CrossRef] [Google Scholar]
  141. Simon, A. A., Kaplan, H. H., Hamilton, V. E., et al. 2020, Science, 370, eabc3522 [Google Scholar]
  142. Spoto, F., Milani, A., & Knežević, Z. 2015, Icarus, 257, 275 [NASA ADS] [CrossRef] [Google Scholar]
  143. Sugimoto, C., Tatsumi, E., Cho, Y., et al. 2021a, Icarus, 369, 114529 [NASA ADS] [CrossRef] [Google Scholar]
  144. Sugimoto, C., Tatsumi, E., Cho, Y., et al. 2021b, Icarus, 369, 114591 [NASA ADS] [CrossRef] [Google Scholar]
  145. Sugita, S., Honda, R., Morota, T., et al. 2019, Science, 364, eaaw0422 [Google Scholar]
  146. Sutherland, W., Emerson, J., Dalton, G., et al. 2015, A&A, 575, A25 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  147. Takir, D., & Emery, J. P. 2012, Icarus, 219, 641 [NASA ADS] [CrossRef] [Google Scholar]
  148. Takir, D., Kareta, T., Emery, J. P., et al. 2020, Nat. Commun., 11, 2050 [NASA ADS] [CrossRef] [Google Scholar]
  149. Takir, D., Emery, J. P., Bottke, W. F., & Arredondo, A. 2024, Sci. Rep., 14, 15965 [NASA ADS] [CrossRef] [Google Scholar]
  150. Tatsumi, E., Sakatani, N., Riu, L., et al. 2021, Nat. Commun., 12, 5837 [NASA ADS] [CrossRef] [Google Scholar]
  151. Tatsumi, E., Tinaut-Ruano, F., de León, J., Popescu, M., & Licandro, J. 2022, A&A, 664, A107 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  152. Tatsumi, E., Vilas, F., de León, J., et al. 2023, A&A, 672, A189 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  153. Tedesco, E. F., Tholen, D. J., & Zellner, B. 1982, AJ, 87, 1585 [NASA ADS] [CrossRef] [Google Scholar]
  154. Tholen, D. J. 1984, PhD thesis, University of Arizona, Tucson, USA [Google Scholar]
  155. Thomas, P. C., Parker, J. W., McFadden, L. A., et al. 2005, Nature, 437, 224 [CrossRef] [Google Scholar]
  156. Thomas, C. A., Trilling, D. E., Rivkin, A. S., & Linder, T. 2021, AJ, 161, 99 [NASA ADS] [CrossRef] [Google Scholar]
  157. Tinaut-Ruano, F., de Leon, J., Tatsumi, E., et al. 2022, A&A, 658, A64 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  158. Tinaut-Ruano, F., de León, J., Tatsumi, E., et al. 2024, A&A, 686, A76 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  159. Todorović, N., & Todorović, N. 2018, MNRAS, 475, 601 [Google Scholar]
  160. Tsiganis, K., Knežević, Z., & Varvoglis, H. 2007, Icarus, 186, 484 [NASA ADS] [CrossRef] [Google Scholar]
  161. Usui, F., Kasuga, T., Hasegawa, S., et al. 2013, ApJ, 762, 56 [Google Scholar]
  162. Usui, F., Hasegawa, S., Ootsubo, T., & Onaka, T. 2019, PASJ, 71, 1 [NASA ADS] [CrossRef] [Google Scholar]
  163. Vernazza, P., Castillo-Rogez, J., Beck, P., et al. 2017, AJ, 153, 72 [NASA ADS] [CrossRef] [Google Scholar]
  164. Vernazza, P., Ferrais, M., Jorda, L., et al. 2021, A&A, 654, A56 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  165. Vilas, F., & Gaffey, M. J. 1989, Science, 246, 790 [NASA ADS] [CrossRef] [Google Scholar]
  166. Vilas, F., Larson, S. M., Hatch, E. C., & Jarvis, K. S. 1993, Icarus, 105, 67 [NASA ADS] [CrossRef] [Google Scholar]
  167. Vinogradova, T. A. 2019, MNRAS, 484, 3755 [NASA ADS] [Google Scholar]
  168. Wakita, S., & Sekiya, M. 2011, Earth Planets Space, 63, 1193 [NASA ADS] [CrossRef] [Google Scholar]
  169. Walsh, K. J., Morbidelli, A., Raymond, S. N., O’Brien, D. P., & Mandell, A. M. 2011, Nature, 475, 206 [Google Scholar]
  170. Walsh, K. J., Delbó, M., Bottke, W. F., Vokrouhlickỳ, D., & Lauretta, D. S. 2013, Icarus, 225, 283 [NASA ADS] [CrossRef] [Google Scholar]
  171. Xin, Y., Shi, J., & Ma, Y. 2024, MNRAS, 527, 10309 [Google Scholar]
  172. Yada, T., Abe, M., Okada, T., et al. 2021, Nat. Astron., 6, 214 [Google Scholar]
  173. Yang, B., & Jewitt, D. 2010, AJ, 140, 692 [Google Scholar]
  174. Yang, B., Hanuš, J., Carry, B., et al. 2020, A&A, 641, A80 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  175. Yumoto, K., Tatsumi, E., Kouyama, T., et al. 2024, Icarus, 420, 116204 [NASA ADS] [CrossRef] [Google Scholar]
  176. Zappala, V., Cellino, A., Farinella, P., & Knezevic, Z. 1990, AJ, 100, 2030 [CrossRef] [Google Scholar]
  177. Zellner, B., Tholen, D. J., & Tedesco, E. F. 1985, Icarus, 61, 355 [NASA ADS] [CrossRef] [Google Scholar]
  178. Ziffer, J., Campins, H., Licandro, J., et al. 2011, Icarus, 213, 538 [NASA ADS] [CrossRef] [Google Scholar]
  179. Zolensky, M. E., Mittlefehldt, D. W., Lipschutz, M. E., et al. 1997, Geochim. Cosmochim. Acta, 61, 5099 [NASA ADS] [CrossRef] [Google Scholar]

All Tables

Table B.1

Spectral characteristics of primitive asteroid families.

Table B.2

Spectral and physical characteristics of primitive asteroid families and the largest members.

Table B.3

Links between family types and compositions (See Sec. 5 for further discussions.)

All Figures

thumbnail Fig. 1

Spectral classifications used in this study following Tholen’s taxonomy. The classification criteria is described in Tatsumi et al. (2023).

In the text
thumbnail Fig. 2

Comparison between fraction of family members with the 0.7 μm band absorption (Qhyd) from spectroscopic studies (Mothé-Diniz et al. 2005; Morate et al. 2016, 2018, 2019, 2021; de León et al. 2016; De Prá et al. 2020b; De Prá et al. 2020a) and the HYD value calculated from the SDSS data. Filled circles indicate more than 20 family members and open circles indicate fewer than 20 family members.

In the text
thumbnail Fig. 3

Median NUV absorption strengths vs. median HYD values for the primitive asteroid families (gray markers indicate fewer than 20 family members). They show a moderate correlation between these two values (ρ = 0.48 ± 0.08), suggesting IVCT of iron inside of phyllosilicates may cause both absorptions.

In the text
thumbnail Fig. 4

NIR slope from 1.25 μm to 2.15 μm, SNIR2, vs. NUV absorption, SNUVSVIS, of asteroid families. The error bars show the interquartile range. There is a correlation between these two values (ρ = −0.47 ± 0.12).

In the text
thumbnail Fig. 5

Spectral characteristics correlate with albedo. (top) Albedo versus VIS slope. (bottom) Albedo versus NUV absorption. The yellow symbols show the families with HYD < 0, possibly with the 0.7 μm absorption. The blue symbols are with HYD > 0. The error bars show the standard deviation.

In the text
thumbnail Fig. 6

NUV-VIS distribution of primitive asteroid families. The x-axis shows the NUV absorption calculated from the difference in the spectral slope of VIS and NUV and the y-axis shows the VIS spectral slope. The symbols and error bars show the median values and interquartile ranges. The circles are the families with HYD < 0 and the triangles are those with HYD > 0 (highly hydrated). (top) All the primitive asteroid families in this study are shown (blue: IMB, orange: MMB, red: OMB). Tholen’s taxonomic classification, which is applied in Tatsumi et al. (2023), is also shown as divided zones labeled with the names of classes. (bottom) The same plot as the top plot with the three regions of the main belt separately shown. The blue circles are HYD<0 and the orange triangles are HYD>0.

In the text
thumbnail Fig. 7

Grouping of primitive asteroid families. We classified the primitive asteroid families into eight types based on the taxonomic compositions and average albedos. The asteroid families with HYD > 0, possibly the 0.7-μm absorption, are underlined.

In the text
thumbnail Fig. 8

Spectral characteristics of New Polana and Eulalia family members. (top) Albedo, (middle) NUV absorption, and (bottom) VIS slope.

In the text
thumbnail Fig. A.1

SDSS spectra for each family. The black lines indicate the median spectra of families. Note that the spectra are without offset to combine with ECAS dataset, i.e., original SDSS spectra.

In the text
thumbnail Fig. B.1

Inner main belt primitive families.

In the text
thumbnail Fig. B.2

Middle main belt primitive families.

In the text
thumbnail Fig. B.3

Outer main belt primitive families.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.