Icarus 318 (2019) 230–240

Contents lists available at ScienceDirect

Icarus

journal homepage: www.elsevier.com/locate/icarus

Ac-H-11 Sintana and Ac-H-12 Toharu quadrangles: Assessing the large and small scale heterogeneities of Ceres’ surface

∗ M.C. De Sanctis a, , A. Frigeri a, E. Ammannito b, F.G. Carrozzo a, M. Ciarniello a, F. Zambon a,

F. Tosi a, A. Raponi a, A. Longobardo a, J.P. Combe c, E. Palomba a, F. Schulzeck d,

C.A. Raymond e, C.T. Russell f a Istituto di Astrofisica e Planetologia Spaziali, INAF, via del fosso del Cavaliere, 100, 00133, Rome, Italy b Agenzia Spaziale Italiana, Rome, via del Politecnico, 00133 Rome, Italy c Bearfight Institute, 22 Fiddler’s Road, P.O. Box 667, Winthrop, WA 98862, USA d DLR, Institute of Planetary Research, Berlin, Germany e Jet Propulsion Laboratory, Pasadena, CA 91109, USA f Institute of Geophysics and Planetary Physics, University of California, Los Angeles, CA 90095-1567, USA

a r t i c l e i n f o a b s t r a c t

Article history: Mineralogical maps of the Ac-H-11 Sintana and the Ac-H-12 Toharu quadrangles of the dwarf planet Ceres Received 27 April 2017 were produced in order to access the composition of this planetary body. We used data from NASA’s

Revised 25 July 2017 Dawn spacecraft, in particular the spectra returned by VIR, the imaging spectrometer on board. Different Accepted 9 August 2017 spectral parameters in the infrared range have been computed to study the composition of this portion Available online 12 August 2017 of Ceres’ surface and its large and small scale variability. We studied the variation and distribution of the Keywords: phyllosilicate bands at 2.73 μm and 3.07 μm, the reflectance at 1.2 μm, and the overall spectrum in specific Dwarf planet Ceres locations. We did not observe variations of the center of the bands at 2.73 μm and 3.07 μm, with the Surface composition exception of a few pixels in the Kupalo crater. We found that this southern region, extending from 0 ° to

H2 O ice 180 ° and from 21 °S to 66 °S, show an overall increase of phyllosilicate band intensity from the equatorial

Carbonates areas to the southern areas. Superimposed to the large-scale trend, we observe many smaller localized Phyllosilicates variations of band intensity. The observed variations can indicate large and small-scale heterogeneities in Reflectance spectroscopy the abundance of the different species in the Ceres subsurface. However, the small-scale variation that is mostly associated with young craters, can be also due to processes related with impacts, such as de- hydration or delivery of exogenous material that, mixed with the original surface, could change the band intensity. Several craters, such as Kupalo and Juling, show a different composition with respect to the background, displaying water ice and sodium carbonates. © 2017 Published by Elsevier Inc.

1. Introduction Here we will use the data returned by the VIR and FC in- struments in order to study the mineralogical composition of this NASA’s Dawn spacecraft arrived at Ceres on March 6, 2015. It southern region of Ceres. The thermally-corrected reflectance spec- was launched in September 2007 to study the two most massive trum of Ceres ( Fig. 2 ) shows that the 2.6–4.2 μm wavelength region bodies in the asteroid belt: the asteroid Vesta and the dwarf planet is characterized by a broad asymmetric feature, due to H2 O/OH Ceres ( Russell and Raymond, 2011 ). A combined approach of gravi- bearing materials. Within this broad absorption feature are several tational investigation, visible and near-infrared spectroscopic mea- distinct absorptions bands at 2.73, 3.05–3.1, 3.3–3.5, and 3.95 μm. surements (VIR, De Sanctis et al., 2011 ), gamma ray and neutron VIR observations show a strong and narrow absorption centered spectroscopy (GRaND, Prettyman et al., 2011 ) and imaging with at 2.72–2.73 μm. This characteristic absorption feature is distinc-

Dawn’s framing camera (FC, Sierks et al., 2011) is used to study tive for OH-bearing minerals because H2 O-bearing phases show the innermost dwarf planet Ceres. a much broader absorption band that is a poor match for Ceres’ spectrum. The Sintana and Toharu quadrangles are located in Ceres’ south- ern hemisphere between 21–66 °S and 0–180 °E ( Fig. 1 ). The Sintana

∗ quadrangle truncates several large craters of the heavily cratered Corresponding author.

E-mail address: [email protected] (M.C. De Sanctis). Zadeni South Pole quadrangle. The Kerwan quadrangle and the http://dx.doi.org/10.1016/j.icarus.2017.08.014 0019-1035/© 2017 Published by Elsevier Inc. M.C. De Sanctis et al. / Icarus 318 (2019) 230–240 231

Fig. 1. Scheme of 15 quadrangles used for High Altitude Mapping Orbit (HAMO)- and Low Altitude Mapping Orbit (LAMO)-based regional geologic mapping ( Roatsch et al., 2016 ). The Sintana quadrangle and the Toharu quadrangle are highlighted by a red border. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Haulani quadrangle are on its northern border ( Tosi et al., this is- ters can be mapped on the surface of Ceres with a resolution never sue ; Palomba et al., this issue ). The southern hemisphere of Ceres achieved before. is less heavily cratered than its northern hemisphere ( Hiesinger et VIR acquired data of Ceres during all of the mission phases, al., 2016 ; Marchi et al., 2016 ). A low-crater density (LCD)_ terrain starting from the Approach phase, for a total of two years of oper- was identified by Hiesinger et al. (2016) , centered at 54.2 °E/23.3 °S, ations at the moment of writing ( Russell et al., 2016 ). Ceres was close to the boundary of the Haulani quadrangle ( Krohn et al., observed with increasing spatial resolution from Survey orbit to 2016 ) in the north. Geological maps of these two quadrangles are LAMO orbit. Here we used mainly the HAMO dataset that has a described in Williams et al. (2017) and Schulzeck et al. (2017) . nominal spatial resolution of about 450 m/pix. Data of the Framing Camera (FC), which offer the geolog- 2. Data analysis description ical context of these regions, reach spatial resolutions up to ∼140 m/pixel (HAMO) and ∼35 m/pixel (LAMO). A HAMO mosaic 2.1. Dataset with a mean resolution of 140 m per pixel ( Roatsch et al., 2016 ) has been used ( Fig. 3 ) in this study. The comparison of local spec- The dataset used was acquired by VIR, the mapping spectrom- tral characteristics and surface morphology was done using indi- eter of the Dawn mission ( Russell et al., 2011; De Sanctis et al., vidual FC images. The FC data has been used to construct the 2011 ). Images taken by the Dawn Framing Camera ( Sierks et al., Digital Terrain Models ( DTM ) that give the topography of the re- 2011 ) are also used for context, geological and morphological anal- gions. In this paper we use such maps for comparison with the ysis. VIR is an imaging spectrometer derived from VIRTIS-M aboard derived spectral parameter maps. Rosetta and Venus Express ( Coradini et al., 1998 ). It measures spec- tra between 0.25 and 5.1 μm using two channels: the VIS channel, between 0.25 and 1.05 μm, and the IR, between 1.0 and 5.1 μm. The 2.2. Data analysis high spatial (IFOV = 250 μrad/pixel, FOV = 64 × 64 mrad) and spec- tral ( λVIS = 1.8 nm/band; λIR = 9.8 nm/band) performances al- The VIR data were calibrated and processed in order to ana- low for the identification of spectral features at small spatial scale. lyze them in terms of mineralogy of the different terrains. The The instrument uses a scanning mirror, thus the scene is scanned calibration used here has been improved with respect to the data one line at a time through the slit of the spectrometer. Each line is published in De Sanctis et al. (2012) , in which calibration artifacts made up of 256 pixels, each having a spectrum in the 0.25–5.1 μm were present in the region beyond 2.5 μm. The new calibration range. The set of adjacent images is then stacked to form a 3-D ar- procedure is described in Carrozzo et al. (2016) . Photometric ray called a hyperspectral cube. VIR’s imaging capability, combined effects are corrected for both topographic variations and physical with an extensive spectral range, offers the geological context for characteristics of the regolith using the Hapke approach ( Ciarniello mineralogical investigations: spectra and derived spectral parame- et al., 2017 ). 232 M.C. De Sanctis et al. / Icarus 318 (2019) 230–240

large-scale variation in band centers. For this reason, maps of band centers are not shown. The band at 3.06–3.07 μm has been assigned to different species like brucite ( Milliken and Rivkin, 2009 ) and ammonia-bearing species ( King et al., 1992 ). Brucite, in particular, clearly shows this narrow characteristic absorption. However, ammoniated phyllosili- cates are the best fit of the overall Ceres spectrum ( De Sanctis et al, 2015 ). For this reason, we assign this band to ammoniated species.

3. Maps and parameters

3.1. Reflectance maps

The first global image maps in the visible and infrared wave- lengths were derived from the FC data, and VIR data ( Roatsch et al., 2016 ; Ciarniello et al., 2017 ) was made from Survey orbit im- ages. Photometrically corrected reflectance maps in the infrared, at 1.2 μm, are obtained from VIR data taken at different orbital phases and, consequently, at different spatial resolutions ( Fig. 4 ). The maps Fig. 2. Average Ceres IR spectrum derived by VIR data. are corrected for topography illumination/shading effects and for photometric effects due to small-scale structure of the soils using The near-infrared reflectance spectra of minerals contain ab- Hapke photometric modeling ( Ciarnello et al., 2017 ). The two re- sorptions that are diagnostic of mineralogy, grain size and crystal flectance maps are still affected by a non-perfect photometric cor- structure. Several analytical tools have been developed to interpret rection due to the very extreme illumination conditions of the ac- such information from remotely obtained spectra (e.g., Sunshine et quired data. Nevertheless, a gradient from North to South is visi- al., 1990 , 1993 ). Band analysis methods are useful for characteriz- ble, with increasing reflectance going to southern regions in both ing spectral data and can be used as a starting point for deriving the quadrangles. mineralogical information ( Gaffey et al., 2002; Klima et al., 2007 ). The main spectral parameters derived from the VIR spectra and 3.2. Band depth maps used here to gather information on Ceres’ surface are: reflectance, band center, band depth, band area, FWHM, asymmetry, spectral Band parameter maps have been also computed, and the distri- slopes, etc. To analyze a large dataset like the one considered in bution of the 2.72 μm band depth and the 3.07 μm band depth are this work, we have developed an automatic data processor able to shown in Figs. 5 and 6 . The methods used for the computation of return different spectral indicators (continuum levels, absorption these maps are fully described in Frigeri et al. (this issue) . bands properties, spectral slopes and their mutual correlations) from Ceres observations. This algorithm allows us to map the spa- 4. General mineralogy of Ac-H-11 and Ac-H-12 tial distribution of each spectral indicator across the surface. The description of the details of how the band parameters have been 4.1. Global trend computed is reported in Frigeri et al. (2017, this issue) . For the spe- cific purpose of this study, we use the distribution of the phyllosil- The presence and the position of the 2.7 μm absorption are icate bands at 2.73 and 3.07 μm ( Fig. 2 ). The other bands present in globally constant in this southern region, but its intensity shows the Ceres spectra, such as the band at about 3.9 μm and the broad significant variations in the range from 0.226 to 0.270. The spatial band between 3.3 and 3.5 μm, are not considered in this mapping distribution of the 2.7 μm band depth ( Fig. 5 ) shows a general in- procedure due to the uncertainties in the band parameters deriva- crease of band depth from the northern border going to the most tion. However, the full spectra are considered for the description of southern latitudes. Both the quadrangles, Sintana and Toharu, be- specific regions. have in a similar manner, but this increase at southern latitudes is The band center and shape is indicative of the specific mineral not seen in the other two adjacent quadrangles, Urvara and Yalode producing the absorption, while the depth of an absorption band ( Ammannito et al., 2016 ). is mainly determined by the abundance of the absorbing minerals, The distinct absorption feature at about 3.1 μm has been at- the grain size distribution, and the abundance of opaque phases; tributed to ammoniated phyllosilicates ( De Sanctis et al, 2015; King thus, it is a useful parameter to be used in mineralogical mapping. et al., 1992; Ammannito et al., 2016 ). Also in this case, the analysis OH-stretching vibrations occur in the 2.7–2.85 μm range for of this feature shows that its band center position does not vary phyllosilicates, and band center positions vary at different wave- across the surface ( Ammannito et al., 2016 ). As with the 2.7 μm lengths for different species. Al-rich, Fe-rich and Mg-rich phyllosil- absorption, the absence of variability in the central wavelength of icates can be discriminated in reflectance spectra based on diag- the 3.1 μm absorption suggests that the ammoniated phyllosilicate nostic positions of the stretching and bending bands ( Bishop et phase is compositionally homogeneous over most of Ceres’ surface. al., 2008 ). For instance, OH-stretching bands occur at ∼2.75 μm for However, the portion of Ceres’ surface discussed here shows spa- montmorillonite, ∼2.80 μm for nontronite and near 2.72 μm for Mg tial variability in the 3.1 μm band intensity ( Fig. 6 ) that broadly fol- smectites. They occur near 2.69 μm for Al-rich trioctahedral micas lows that of the 2.7 μm band depth distribution: we see a general and near 2.76 μm for Al-rich dioctahedral micas ( Beran, 2002 ). increase in the intensity of the ammoniated phyllosilicates band, Ceres’ average spectrum shows the phyllosilicate band centered going from the mid-equatorial regions to the south ( Fig. 6 ). at 2.72–2.73 μm, indicative of the presence of Mg-phyllosilicates Band depth primarily reflects the abundance of the absorb- ( Fig. 2 ). The global distribution of this band center shows that most ing species. Although the Ceres phyllosilicates are compositionally of the Ceres surface has the same band position with a few ex- homogeneous, the variations in the depth of absorptions in the ceptions, such as the bright material in Occator ( De Sanctis et al., 2.7 μm and 3.1 μm bands indicate a variability in the abundance of 2016 ). The quadrangles under study in this article do not show the associated minerals. The large-scale regional variations shown M.C. De Sanctis et al. / Icarus 318 (2019) 230–240 233 in Fig. 5 suggest latitudinal heterogeneity in phyllosilicate abun- with craters present in the Northeastern part of the region as Jul- dance over a scale of several hundreds of kilometers. In Ac-11, ter- ing ( −35, 168) and Kupalo ( −39, 173) craters and the nearby Fluusa rains associated with higher concentration of phyllosilicates appear crater ( −30, 178) ( Fig. 3 ). close to Zadeni and Mondamin craters, extending eastward roughly Moreover, smaller phyllosilicate bands can be seen which are to the Hamori crater and northward to Sintana. The entire south- associated with an unnamed crater southeast of Kerwan at ( −23; ern part of Ac-12, including craters Chaminuka and Toharu, is char- 138) (a in Fig. 3 b) and two nearby craters (b,c in Fig. 3 b), and acterized by stronger phyllosilicate band intensity. Thus, there is a a small young crater on a topographic high at ( −35; 111) (d in very large region, extending from longitude 0 ° to 180 ° and latitude Fig. 3 b). It is interesting to note that the large southern craters from −45 ° to −65 ° showing deeper phyllosilicate bands ( Fig. 5 ). such as Toharu and Chaminuka do not show any relevant differ- This distribution is in contrast with phyllosilicate band depth dis- ence in band depth with respect to the surroundings. tribution between longitude 180 °–360 °, where smaller depths are The 3.07 μm band depth distribution is broadly correlated with mapped. This pronounced difference in phyllosilicate concentration the 2.7 band depth, showing smaller intensities where the phyl- between these two southern regions may reflect different compo- losilicate band is also less pronounced ( Figs. 5 and 6 ). The area sitions pointing to an inhomogeneity of constituents of the Ceres’ encompassing Juling and Kupalo craters ( Fig. 11 ) shows the largest surface. contrast in terms of band intensity. The topography of this southern region is relatively lower with Craters Juling (19 km diameter; 36 °S, 168.3 °E) and Kupalo respect the equatorial regions ( Buczkowski et al., 2016 ), containing (27 km diameter; 39.6 °S, 173 °E), in the eastern part of the quadran- large and deep carters as Zadeni, Sintana, Darzmat, Chaminuka and gle, display relatively pristine (sharp rims, well-preserved ejecta) Toharu. This broad correspondence between topography and abun- morphologies ( Fig. 11 ). They expose brighter, fresh material in the dance of the phyllosilicate and ammoniated species could indicate crater walls and ejecta. vertical compositional heterogeneities of the subsurface. Moreover, The floor of Juling crater is covered by a melt/fluidized flow, this southern region broadly corresponds with a quasi-circular and and Kupalo shows signs of long run-out landslides ( Schmidt et al., large-scale depression identified by Marchi et al. (2016) , as a rem- 2017 ). The flow seen in Juling suggests efficient material trans- nant of a large-scale ancient impact. port, and the morphology indicates the presence of water ice. In- Superimposed to this general trend, we see many localized ar- deed, water ice has been identified in the spectra acquired by VIR eas that show variations in band depth with respect to the nearby ( Raponi et al., 2017 ; Combe et al., this issue ). Ice signature has regions. Most of them are associated with craters of different size been clearly observed on the northern crater walls where the illu- and shape. mination is low. Juling also shows some small areas on the crater’s southern rim enriched in carbonates with respect to the back- ground material. The larger amount of carbonates can be mapped 4.2. Specific areas/craters using the band depth at about 4 μm ( Carrozzo et al., 2017 ). Rep- resentative spectra of the Juling crater are shown in Fig. 12 (a,b), 4.2.1. Sintana quadrangle where we report some examples of carbonate-rich ( Fig. 12 a) and The 2.7 band depth distribution shows clearly a region from 0 ° water ice-rich pixels ( Fig. 12 a,b) in comparison with background to 25 ° and −21 ° to −45 ° where several localized areas of smaller material. band strength are present. The smaller bands are broadly associ- An even larger amount of carbonates is present in the Kupalo ated with craters present in the region, such as the large Jarimba crater. It shows prominent bright ejecta, sharp rims with bright crater ( −23, 21), several unnamed craters south of Jarimba (a,b, in material exposures, and a very fresh appearance. It stands out in Fig. 8 ), a crater cluster (c in Fig. 8 ), the Doliku crater and several the carbonates distribution maps, both in terms of high band depth craters (d,e) northeast of Daliku, and an unnamed crater at about and in band centers ( Carrozzo et al., 2017 ). Ceres carbonate band ( −43; 18) (f in Fig. 8 ). are mostly centered at 3.95 μm, but a few small localized areas The craters’ characteristics are very different: young and old, show band center at longer wavelengths. Notably Occator’s cen- large and small, with irregular rims and shapes, or almost regu- tral dome shows a band minimum at 4.02 μm, consistent with the lar. These differences indicate that the excavated material of the presence of sodium-carbonate ( De Sanctis et al., 2016 ). The Ku- overall area has lower band intensities. There are two other areas palo area also has a carbonate band center at long wavelengths, showing smaller band depth in the northeast area of the quad- indicating the presence of sodium-carbonate. Representative spec- rangle. These two areas are also associated with craters: the large tra of carbonate-rich pixels are plotted in Fig. 13 , in comparison Tupo crater and the much smaller Braciaca ( Fig. 9 ). Both of these with background material. craters are young and show well-defined rims. The distribution of Some of the spectra show a shift and a flattening of the mini- band depth in Braciaca is unusual, showing deeper bands inside mum of the 2.7 μm band that moves from 2.72–2.73 μm in D and the crater (crater floor) and smaller bands outside (ejecta). Spectra E, to 2.75–2.76 μm in A and B ( Fig. 14 ). Some spectra show a very representative of Braciaca ejecta and the crater floor are shown in flat minimum of the 2.7 band as pixels in C. Also the carbonate Fig. 10 . band minima shift at longer wavelengths, up to 4.01 μm in A and B ( Fig. 14 ). 4.2.2. Toharu quadrangle The band shift at longer wavelengths are associated with a re- The Toharu Quadrangle contains impact craters that exhibit a duction of the 2.7 μm intensity, while the carbonate band at 3.9 μm range of sizes and preservation styles. Smaller craters ( < 40 km) increases its strength. Similar behavior was observed for Occator’s generally appear morphologically “fresh,” and their rims are nearly bright central material ( De Sanctis et al., 2016 ). The shift of the circular and raised above the surrounding terrain. Larger craters, 2.72 μm band is visible in the high spatial resolution data (nominal such as Toharu, appear more degraded, exhibiting irregularly resolution of 95 m/pix) of Kupalo taken in the Low Altitude Map- shaped, sometimes scalloped, rim structures, and debris lobes on ping Orbit, which is difficult to recognize at intermediate and low their floors. Numerous craters ( > 20 km) contain central mounds. resolution, implying that only a few small areas are characterized The quadrangle also contains a number of large ( > 20 km across) by such phyllosilicate minima. The shift of the metal-OH absorp- depressions that are only observable in the topographic data. tion from 2.72 μm to 2.76 μm ( Fig. 14 a) indicates the additional oc- The 2.7 μm band depth distribution shows several localized ar- currence of another phyllosilicate, such as Al-phyllosilicate, possi- eas of smaller band strengths. The smaller bands are associated bly smectite, kaolinite, or illite ( Bishop et al., 2008 ). In fact, the po- 234 M.C. De Sanctis et al. / Icarus 318 (2019) 230–240

Fig. 3. (a) Clear filter mosaic of the Ac-H-11 Sintana quadrangle. (b) FC clear filter mosaics of Ac-H-12 Toharu. sition of OH-absorption give clues to the specific mineral. Typically, This unit has the highest crater density and is the oldest unit on Mg-phyllosilicates show band absorptions at about 2.72–2.73 μm, Ceres ( Williams et al., 2017 ). The chrono-stratigraphy of the Ac-H- while Al-phyllosilicate has absorptions at about 2.75–2.76 μm. 11 quadrangle indicates that the oldest unit is the cratered terrain The surface of Ceres shows a remarkable uniformity in terms of ( Schulzeck et al., 2017 ). The crater material corresponds to units band centers ( Ammannito et al., 2016 ). After Occator’s bright ma- starting from intermediate to younger ages. The oldest morphologi- terial, Ceres’ surface is the second place where the OH absorption cally visible unit of crater material is observed in the Sintana crater band shows a different wavelength of the band minimum. The as- ( Schulzeck et al., 2017 ). Different types of crater material superpose sociation of sodium-carbonate and Al-phyllosilicate suggests chem- cratered terrain, and few areas of these two quadrangles have been ical processes that involve both the chemical species. mapped as crater material. Usually, only craters with an exclusively fresh morphology, like distinct crater rims, display layers of crater 5. Comparison with geological maps and geology material, but the Sintana crater ( Schulzeck et al., 2017 ) is an ex- ception and has been interpreted as remnants of crater material. According to the topographic maps, there are several scarps and In fact, the Sintana crater is not particularly young in the crono- a topographic low at the border of Ac-11 quadrangle that indicate stratigraphy of the quadrangle ( Schulzeck et al., 2017 ). the possible presence of a basin of about 300 km in diameter. The The crater material in Ac-H-11 broadly corresponds with areas putative old basin should have occurred very early in the quad- showing smaller 2.7 μm band depths in the mineralogical maps rangle’s history ( Schulzeck et al., 2017 ). We do not see any rele- (see Fig. 4 a). Specifically, small band depths have been mapped vant variation in band parameters associated with such a degraded that are associated primarily with crater material close to the structure. Jarimba crater, a crater cluster south of Jarimba, Doliku, Tupo and The cratered terrain is the dominant unit of quadrangles Ac- Braciaca. Nevertheless, the Sintana crater has also been mapped as H-11 and Ac-H-12 ( Williams et al., 2017; Schulzeck et al., 2017 ). crater material, but does not show any relevant contrast in band M.C. De Sanctis et al. / Icarus 318 (2019) 230–240 235

Fig. 4. (Left) Reflectance map at 1.2 μm of quadrangle Ac-H-11; (Right) Reflectance map at 1.2 μm of quadrangle Ac-H-12. depth, which may be related to the relative older age of Sintana phyllosilicate and ammoniated-phyllosilicate band intensity going with respect to the previously cited craters. from the sub-equatorial regions to the most southern regions Adding together crater age and band depth distribution, we can ( Figs. 5 and 6 ). This trend is limited to these two quadrangles, affirm that the craters in Ac-11 with smaller band depths are the and it is not seen in the adjacent southern regions extending from youngest of the quadrangle. However, this correlation is not fully 180 ° to 360 °. Thus, the trend is not a general latitudinal trend, confirmed by the analysis of Ac-12. In Ac-H-12, the smallest ab- but seems confined in longitude approximately between 0 ° and sorptions are found in Juling and Kupalo craters. Material in these 180 °. Looking at the topography of the region, we see that part two craters is mapped as crater material ( Williams et al., 2017 ) of it (from 0 ° to about 90 °) broadly corresponds to a topographic even if they are much brighter with respect to the background ma- low, also identified as a possible remnant of a large impact basin terial. Other areas present in the quadrangle mapped as crater ter- (basin C in Marchi et al., 2016 ). The remaining region partially cor- rain also show smaller band depth. Also in this case, there is an responds to large and shallow craters, suggesting a possible rela- apparent trend between crater age and band depth. However, this tionship with topography or with material excavated by these large correlation is not fully valid everywhere. Craters b, c in Fig. 3 b, craters. However, this relationship is not valid elsewhere on Ceres; which show small band depth in the VIR maps ( Fig. 5 b), are clas- in fact, the two nearby quadrangles that host the large Urvara and sified as cratered terrains (older), and conversely, extensive areas Yalode craters show a different trend in terms of band intensity around Toharu and Chaminuka craters e, f, g in Fig. 3 b are mapped ( Ammannito et al., 2016 ). The differences in band intensity on re- as crater terrain (younger), but do not show any decrease in band gional scales indicate large-scale latitudinal and longitudinal het- depth. Thus, the association with small band depth-crater terrain erogeneities in the composition, at least in terms of abundance of cannot be fully confirmed, even if a broad association of younger the phyllosilicates. craters with smaller 2.7 band can be seen. Other correlations with The Ac-H-11 Sintana Quadrangle and the Ac-H-12 Toharu Quan- geological units are not clearly identified. drangle are characterized by a variety of geological features We attempted to understand whether there is a correlation be- ( Williams et al., 2017; Schulzeck et al., 2017 ). Not all of them have tween band intensity and crater diameter/depth, but we easily see a counterpart in the mineralogical maps, but few intermediate to that craters with similar diameter/depth show very different in- small craters are associated with differences in band depths, of- tensities in phyllosilicate bands, both ammoniated and not. A clear ten showing smaller band intensity with respect to the background example is the comparison of the Tupo crater with an unnamed material. These craters are frequently mapped as crater terrain and crater at 53 ° S; 108 °, which are similar in size but very different display a fresh appearance, suggesting a possible correlation be- in terms of band intensity. Thus, we exclude band intensity as be- tween band intensity and age of the feature. However, the corre- ing related only to excavation depths and then to a homogenous lation is not fully confirmed by the behavior of the area north of stratigraphy of the first kilometers of the Cerean subsurface. the Fluusa crater, where an overall reduction of band depth can be seen not associated to any specific young crater, and judging by the 6. Discussion and conclusion occurrences of a few other small areas with smaller phyllosilicate band depth, not mapped as young surface, and vice versa. The main common characteristic of the Ac-H-11 Sintana Quad- Band intensity in a mixture of different minerals is related to rangle and the Ac-H-12 Toharu Quandrangle is the increase of several factors. The most obvious is the abundance of minerals that 236 M.C. De Sanctis et al. / Icarus 318 (2019) 230–240

Fig. 5. Distribution map of 2.72 μm band depth; (left) quadrangle Ac-H-11, and (right) quadrangle Ac-H-12.

Fig. 6. Distribution map of 3.07 μm band depth; (left) quadrangle Ac-H-11, and (right) quadrangle Ac-H-12. M.C. De Sanctis et al. / Icarus 318 (2019) 230–240 237

Fig. 8. FC clear filter image of the areas between Jarimba and Doliku.

Fig. 7. Vertical profiles of band depths taken at longitudes 45 ° (Ac-H-11, red line) and 135 ° (Ac-H-12, blue line). Index 0 corresponds to 21 °S and index 2350 to 65 °S. Fig. 9. FC clear filter image of the areas encompassing Tupo and Braciaca craters. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) produce such a signature, thus, depth-dependent variations in the abundance of materials exposed by craters and/or heterogeneity of the first subsurface layers of the surface must be considered, in addition to other factors such as particle grain size. Moreover, phase changes caused by such impact processes as dehydration and abundance in the quantity of impact-delivered material can be other factors playing a role in the band intensity variation. These possible causes have been considered in explaining the global distribution of band intensity on Ceres ( Ammannito et al., 2016 ). They concluded that the relative abundance and size of opaque phases are not the main cause of the observed variation, because they have a major effect on albedo and the strength of absorption bands ( Clark, 1983; Pieters, 1983 ), but in a way that behaves differently from what is observed on Ceres. Also, in the Sintana and Toharu quadrangles, albedo and band strength are not correlated, thus a variation in opaque content cannot account for Fig. 10. VIR spectra of representative pixels of Braciaca ejecta and crater floor. the observed variations in band strength. Another possible cause is the dehydration of phyllosilicates at increased temperatures generated by impacts. The warming of ent impact velocity, a sort of reduction, although limited, of band phyllosilicates could remove H2 O and OH from the heated material, intensity must be present, associated with all impact structures. reducing the intensity of the OH-stretch band at 2.7 mm ( Frost and Nevertheless, we have seen that not all the craters show varia- Vassallo, 1996; Russell and Farmer, 1964 ) and of the NH4-related tion in band intensity. Moreover, such a process cannot explain the band at 3.1 mm ( Chourabi and Fripiat, 1981; Frost and Vassallo, large-scale variation in band intensity (from equatorial to southern 1996 ). Even if the heating of the target can be different for differ- region) observed in the two quadrangles studied here. 238 M.C. De Sanctis et al. / Icarus 318 (2019) 230–240

Fig. 13. Spectra from selected pixels from areas rich in carbonate in Kupalo, as Fig. 11. FC clear filter images of Juling and Kupalo craters. shown in Fig. 11 .

Fig. 14. (a) Continuum removed spectra of pixels taken in areas A, C, and E of Fig. Fig. 12. Representative spectra of different areas of the Juling crater. (a) Some ex- 11 , showing the shift of the center of the band at 2.7 μm; (b) Continuum removed amples of carbonate- and ice-rich areas in comparison with background and crater spectra of pixels taken in areas A and B of Fig. 11 , showing the shift of the center of floor; (b) Spectra of ice-rich areas. the band at about 4 μm. A spectrum of the background is also plotted as reference. M.C. De Sanctis et al. / Icarus 318 (2019) 230–240 239

Similarly, the observed large-scale trend is difficult to explain Carrozzo, F.G., De Sanctis, M.C., Raponi, A., Ammannito, E., Castillo-Rogez, J., Cia- with a different abundance in the quantity of impact-delivered rniello, M., Ehlmann, B.L., Fonte, S., Formisano, M., Frigeri, A., Giardino, M., Magni, G., Palomba, E., Tosi, F., Stains, N., Zambon, F., Raymond, C.A., Russell, material, while it could explain small-scale, impact-related hetero- C.T., 2017. Distribution of carbonates on Ceres, Sci. Adv., in revision. geneities. Chourabi, J.J., Fripiat, 1981. Determination of tetrahedral substitutions and interlayer Most probably, the observed variation corresponds to variations surface heterogeneity from vibrational spectra of ammonium in smectites. Clays

Clay Miner. 29, 260–268. doi:10.1346/CCMN.1981.02904031.4972256. in the abundance of different minerals present in the shallow Ceres Ciarniello, M. , De Sanctis, M.C. , Ammannito, E. , Raponi, A. , Longobardo, A. , subsurface, even if we cannot exclude that all the factors described Palomba, E. , Carrozzo, F.G. , Tosi, F. , Li, J.-Y. , Schröder, S.E. , Zambon, F. , Frigeri, A. , above can contribute in different ways to the mapped distribution Fonte, S. , Giardino, M. , Pieters, C.M. , Raymond, C.A. , Russell, C.T. ,2017. Spec- of phyllosilicates. For instance, the broad correlation of smaller in- trophotometric properties of dwarf planet Ceres from the VIR spectrometer on board the Dawn mission. Astron. Astrophys. 598, A130 . tensity with young crater age could indicate impact-related pro- Clark, R.N., 1983. Spectral properties of mixtures of montmorillonite and dark car- cesses that fade with the aging of the surface: local de-hydration bon grains: implications for remote sensing minerals containing chemically or mixture of the target material with the less hydrated impactor and physically adsorbed water. J. Geophys. Res. 88, 10635–10644. doi:10.1029/ JB088iB12p10635 . material. Local induced (i.e. impact-related processes) variations of Combe, J.-Ph. , Raponi, A. , McCord, T.B. , Tosi, F. , De Sanctis, M.C. , Ammannito, E. , band intensity are expected to change with the aging of the sur- Byrne, S. , Carrozzo, F.G. , Carsenty, U. , Hayne, P.O. , Hughson, K.H.G. , Johnson, K.E. , face due to the mixing of the altered material with the background Landis, M.E. , Pieters, C.M. , Mazarico, E. , Platz, T. , Ruesch, O. , Schröder, S. , Singh, S.M. , Zambon, F. , Raymond, C.A. , Russell, C.T. , 2017. Exposed H O-rich ar- 2 material (Pieters et al., 2012). eas detected on Ceres with the Dawn Visible and InfraRed mapping spectrome- Among the most interesting local features in terms of mineral- ter. Icarus this issue . ogy of these two southern quadrangles, are craters Juling and Ku- Coradini, A. , Capaccioni, F. , Drossart, P. , Semery, A. , Arnold, G. , Schade, U. , Angrilli, F. , Barucci, M.A. , Bellucci, G. , Bianchini, G. , Bibring, J.P. , Blanco, A. , Blecka, M. , palo. Both craters show smaller intensity of NH4-phyllosicate and Bockelee-Morvan, D. , Bonsignori, R. , Bouye, M. , Bussoletti, E , Capria, M.T. , Carl- Mg-phyllosilicate, but they are also characterized by very different son, R. , Carsenty, U. , Cerroni, P. , Colangeli, L. , Combes, M. , Combi, M. , Crovisier, J. , mineralogies. Juling crater is the most equatorial place where wa- Dami, M. , De Sanctis, M.C. , DiLellis, A.M. , Dotto, E. , Encrenaz, T. , Epifani, E. ,

Erard, S., Espinasse, S., Fave, A., Federico, C., Fink, U., Fonti, S., Formisano, V., ter ice has been detected (Raponi et al., 2017). Ice has been found Hello, Y. , Hirsch, H. , Huntzinger, G , Knoll, R. , Kouach, D. , Ip, W.P. , Irwin, P. , Kach- in the northern part of the rim in the shadows. licki, J. , Langevin, Y. , Magni, G. , McCord, T. , Mennella, V. , Michaelis, H. , Mon- Juling also shows more carbonate than the background mate- dello, G. , Mottola, S. , Neukum, G. , Orofino, V. , Orosei, R. , Palumbo, P. , Peter, G. , rial similar to the nearby Kupalo crater, where a large quantity of Pforte, B., Piccioni, G., Reess, J.M., Ress, E., Saggin, B., Schmitt, B., Stefanovitch, D., Stern, A. , Taylor, F. , Tiphene, D. , Tozzi, G. , 1998. VIRTIS: an imaging spectrometer carbonate has been detected, especially on the rim and ejecta. The for the ROSETTA mission. Planet. Space Sci. 46, 1291–1304 . analysis of the spectra indicate that the carbonate in this area is De Sanctis, M.C. , Coradini, A. , Ammannito, E. , Filacchione, G. , Capria, T. , Fonte, S. , not only (Mg, Ca)-carbonate, as is most of the Ceres surface, but Magni, G., Barbis, A., Bini, A., Dami, M., Fical-Veltroni, I., Preti, G.the VIR Team, 2011. The VIR spectrometer. Space Sci. Rev. 163, 329–369 . also Na-carbonate (Carrozzo et al., 2017). Moreover, we observe a De Sanctis, M.C., et al., 2012. Spectroscopic characterization of mineralogy and shot of the center of the 2.72 μm band that moves to longer wave- its diversity across vesta. Science 336 (6082), 697–700. doi: 10.1126/science. lengths where the sodium carbonate is present. The shift from 2.72 1219270 . De Sanctis, M.C. , Ammannito, E. , Raponi, A. , Marchi, S. , McCord, T.B. , McSween, H.Y. , μ to 2.75–2.76 m can indicate the presence of a phyllosilicate differ- Capaccioni, F. , Capria, M.T. , Carrozzo, F.G. , Ciarniello, M. , Longobardo, A.,F. , ent from the background. A similar shift was observed in Occator, Frigeri, M. , Giardino, G. , Magni, E. , Palomba, D. , Turrini, F. , Zambon, J.-Ph. , where the spectra of the central bright dome show a band center Combe, W. , Feldman, R. , Jaumann, McFaddenL.A. , Pieters, C.M. , Prettyman, T. , Raymond, C.A. , Russell, C.T. , 2015. Ammoniated phyllosilicates with a likely μ at about 2.75 m (De Sanctis et al., 2016; Longobardo et al., this outer Solar System origin on (1) Ceres. Nature 528, 241–244 . issue ) ( Fig. 7 ). De Sanctis, M.C., Raponi, A., Ammannito, E., Ciarniello, M., Toplis, M.J., Mc- Sween, H.Y., Castillo-Rogez, J.C., Ehlmann, B.L., Carrozzo, F.G., Marchi, S., Tosi, F., Zambon, F., Capaccioni, F., Capria, M.T., Fonte, S., Formisano, M., Frigeri, A., Gia- Acknowledgments rdino, M., Longobardo, A., Magni, G., Palomba, E., McFadden, L.A., Pieters, C.M., Jaumann, R., Schenk, P., Mugnuolo, R., Raymond, C.A., Russell, C.T., 2016. Bright

carbonate deposits as evidence of aqueous alteration on (1) Ceres. Nature VIR is funded by the Italian Space Agency and was developed doi: 10.1038/nature18290 . under the leadership of INAF-Istituto di Astrofisica e Planetologia Frigeri, A., De Sanctis, M.C., Ammannito, E., Tosi, F., Ciarniello, M., Zambon, F., Car- Spaziali, Rome, Italy. The instrument was built by Selex-Galileo, rozzo, G., Raponi, A., McCord, T., Raymond, C.A., Russell, C.T., The spectral pa-

Florence, Italy. The authors acknowledge the support of the Dawn rameter maps of Ceres from NASA/DAWN VIR data. Icarus, this issue. Frost, R.L., Vassallo, A.M., 1996. The dihydroxylation of the kaolinite clay miner- Science, Instrument, and Operations Teams. This work was sup- als using infrared emission spectroscopy. Clays Clay Miner. 44, 635–651. doi: 10. ported by the Italian Space Agency ASI-INAF n. I/004/12/1 and 1346/CCMN.1996.0440506 .

NASA . A portion of this work was performed at the Jet Propulsion Gaffey, M.J., Cloutis, E.A., Kelley, M.S., Reed, K.L., 2002. Mineralogy of Asteroids. Uni- versity of Arizona Press, Tucson .

Laboratory under contract with NASA. This research utilizes spec- Hiesinger , et al. , 2016. Cratering on Ceres: implications for its crust and evolution. tra of the NASA RELAB facility at Brown University. We thanks the Science 353, 4759 . Dawn Operation Team and the Dawn Science Team. King, T.V. , Clark, R.N. , Calvin, W.M. , Sherman, D.M. , Brown, R.H. , 1992. Evidence for ammonium-bearing minerals on Ceres. Science 255, 1551–1553 . Klima, R.L. , Pieters, C.M. , Dyar, M.D. , 2007. Spectroscopy of synthetic Mg-Fe pyrox- References enes I: spin-allowed and spin-forbidden crystal field bands in the visible and near-infrared. Meteorit. Planet. Sci. 42 (Nr 2), 235–253 .

Ammannito, E., De Sanctis, M.C., Ciarniello, M., Frigeri, A., Carrozzo, F.G., Combe, J-P., Krohn, K., et al., 2016. The unique geomorphology and structural geology of the

Ehlmann, B.L., Marchi, S., McSween, H.Y., Raponi, A., Toplis, M.J., Tosi, F., Castillo- Haulani crater ofdwarf planet Ceres as revealed by geological mapping of equa-

Rogez, J.C., Capaccioni, F., Capria, M.T., Fonte, S., Giardino, M., Jaumann, R., torial quadrangle Ac-6 Haulani. Icarus.

Longobardo, A., Joy, S.P., Magni, G., McCord, T.B., McFadden, L.A., Palomba, E., Longobardo, A., Palomba, E., Carrozzo, F.G., De Sanctis, M.C., Galiano, A., Tosi, F.,

Pieters, C.M., Polansky, C.A., Rayman, M.D., Raymond, C.A., Schenk, P., Zam- Zambon, F., Raponi, A., Ciarniello, M., Ammannito, E., Stephan, K., Frigeri, A.,

bon, F., Russell, C.T. the Dawn Science Team, 2016. The distribution of phyllosil- Capria, M.T., Fonte, S., Giardino, M., Raymond, C.A., Russell, C.T., Mineralogical

icates on the surface of Ceres. Science 353, 1006. doi: 10.1126/science.aaf4279 . mapping of the Occator quadrangle, Icarus, this issue.

Beran, A. , 2002. Infrared spectroscopy of micas. In: Mottana, A., Sassi, F.P., Thomp- Marchi, S., Emakov, A.I., Raymond, C.A., et al., 2016. The missing large impact craters

son, J.B., Guggenheim, S. (Eds.), Micas: Crystal Chemistry and Metamorphic Petrol- on Ceres. Nature 7, 12257.

ogy . Mineralogical Society of America, pp. 351–370 . Milliken, R.E., Rivkin, A.S., 2009. Brucite and carbonate assemblages from altered

Bishop, J.L., Lane, M.D., Dyar, M.D., Brown, A.J., 2008. Reflectance and emission spec- olivine-rich materials on Ceres. Nat. Geosci. 2, 258–261.

troscopy study of four groups of phyllosilicates: smectites, kaolinite-serpentines, Palomba, E., Longobardo, A., De Sanctis, M.C., Carrozzo, F.G., Galiano, A., Zambon,

chlorites and micas. Clay Miner. 2008. doi: 10.1180/claymin.2008.043.1.03 . F., Raponi, A., Ciarniello, M., Stephan, K., Williams, D., Ammannito, E., Capria,

Buczkowski, D.L. , Schmidt, B.E. , Williams, D.A. , et al. , 2016. The geomorphology of M.T., Fonte, S., Giardino, M., Tosi, F., Raymond, C.A., Russell, C.T., Mineralogical

Ceres. Science 353, 6303 . mapping of the Kerwan quadrangle on Ceres, Icarus, this issue.

Carrozzo, F.G., Raponi, A., De Sanctis, M.C., Ammannito, E., Giardino, M., D’Aversa, E., Pieters, C.M., et al., 2012. Distinctive space weathering on Vesta from regolith mix-

Fonte, S., Tosi, F., 2016. Artifacts reduction in VIR/Dawn data. Rev. Sci. Instrum. ing processes. Nature 491, 79–82. 87, 124501 doi: 10.1063/B . 240 M.C. De Sanctis et al. / Icarus 318 (2019) 230–240

Pieters, C.M., 1983. Strength of mineral absorption features in the transmitted com- Schmidt, B. , Hughson, K. , Chilton, H. , Scully, J. , Platz, T. , Nathues, A. , Sizemore, H. , ponent of near-infrared reflected light: first results from RELAB. J. Geophys. Res. Bland, M. , Byrne, S. , Marchi, S. , O’Brien, D. , Schorghofer, N. , Hiesinger, H. , Jau- 88, 9534–9544. doi: 10.1029/JB088iB11p09534 . mann, R. , Pasckert, H. , Lawrence, J. , Buczkowski, D. , Castillo-Rogez, J. , Sykes, M. , Prettyman, T.H. , Feldman, W. , McSween, H.Y. , Dingler, R. , Enemark, D. , Patrick, D. , Schenk, P. , M De Sanctis, M.C. , Mitri, G. , Formisano, M. , Li, Y.J. , Reddy, V. , Le Storms, S. , Hendricks, J. , Morgenthaler, J. , Pitman, K. , Reedy, R. , 2011. Dawn’s Corre, L. , Russell, C.T. , Raymond-, C.A. , 2017. Geomorphological evidence for gamma ray and neutron detector. Space Sci. Rev. 163, 371–459 . ground ice on dwarf planet Ceres. Nat. Geosci. in press . Raponi, A. , De Sanctis, M.C. , Ciarniello, M. , Ammannito, E. , Frigeri, A. , Combe, J.-P. , Schulzeck, F. , Krohn, K. , Gathen, I.v.d. , Schmedemann, N. , Stephan, K. , Jaumann, R. , Tosi, F. , Zambon, F. , Carrozzo, F.G. , Magni, G. , Capria, M.T. , Formisano, M. , Lon- Williams, D.A. , Buczkowski, D.L. , Mest, S.C. , Scully, J.E.C. , Kersten, E. , Matz, K.-D. , gobardo, A. , Palomba, E. , Pieters, C.M. , Russell, C.T. , Raymond, C.A. , 2017. Water Nass, A. , Preusker, F. , Roatsch, T. , Raymond, C.A. , Russell, C.T. , 2017. Geologic ice on Ceres’ surface as seen by Dawn –VIR: properties retrieval by means of mapping of the Ac-11 Sintana quadrangle. Icarus submitted . spectral modeling. Lunar and Planetary Science Conference, abstract No. 1964, Sierks, H. , Sierks, H. , Keller, H.U. , Jaumann, R. , Michalik, H. , Behnke, T. , Buben- id. 2007 . hagen, F. , Büttner, I. , Carsenty, U. , Christensen, U. , Enge, R. , Fiethe, B. , Gutiér- Roatsch, Th., Kersten, E., Matz, K.-D., Preusker, F., Scholten, F., Jaumann, R., Ray- rez Marqués, P. , Hartwig, H. , Krüger, H. , Kühne, W. , Maue, T. , Mottola, S. , mond, C.A., Russell, C.T., 2016. High-resolution Ceres high altitude mapping or- Nathues, A. , Reiche, K.-U. , Richards, M.L. , Roatsch, T. , Schröder, S.E. , Szemerey, I. , bit atlas derived from Dawn framing camera images. Planet. Space Sci. 129, 103– Tschentscher, M. , 2011. The Dawn framing camera. Space Sci. Rev. 163, 263–327 . 107. doi: 10.1016/j.pss.2016.05.011 . Sunshine, J.M. , Pieters, C.M. , 1993. Estimating modal abundances from the spectra of Russell, J.D., Farmer, V.C., 1964. Infra-red spectroscopic study of the dehydration natural and laboratory pyroxene mixtures using the modified Gaussian model. of montmorillonite and saponite. Clay Miner. Bull. 5, 443–464. doi: 10.1180/ J. Geophys. Res. 98, 9075–9087 . claymin.1964.005.32.04 . Sunshine, J.M. , Pieters, C.M. , Pratt, S.F. , 1990. Deconvolution of mineral absorption Russell, C.T. , Raymond, C.A. , 2011. The dawn mission to vesta and ceres. Space Sci. bands—an improved approach. J. Geophys. Res. 95, 6955–6966 . Rev. 163, 3–23 . Tosi, F., Carrozzo, F.G., Zambon, F., Ciarniello, M., Frigeri, A., Combe, J.-Ph., De Sanc- Russell, C.T. , Raymond, C.A. , Ammannito, E. , Buczkowski, D.L. , De Sanctis, M.C. , tis, M.C., Hoffmann, M., Longobardo, A., Nathues, A., Raponi, A., Ammannito, Hiesinger, H. , Jaumann, R. , Konopliv, A.S. , McSween, H.Y. , Nathues, A. , Park, R.S. , E., Krohn, K., McFadden, L.A., Palomba, E., Pieters, C.M., Stephan, K., Raymond, Pieters, C.M. , Prettyman, T.H. , McCord, T.B. , McFadden, L.A. , Mottola, S. , Zu- C.A., Russell, C.T., Mineralogical analysis of the Haulani quadrangle of the dwarf ber, M.T. , Joy, S.P. , Polanskey, C. , Rayman, M.D. , Castillo-Rogez, J.C. , Chi, P.J. , planet Ceres, Icarus, this issue. Combe, J-P. , Ermakov, A. , Fu, R.R. , Hoffman, M. , Jia, Y.D. , King, S.D. , Lawrence, D.J. , Williams, D.A. , Buczkowski, D.L. , Mest, S.C. , Scully, J.E.C. , 2017. Introduction: the ge- Li, J-Y. , Marchi, S. , Preusker, F. , Roatsch, T. , Ruesch, O. , Schenk, P. , Villarreal, M.N. , ologic mapping of Ceres. Icarus . Yamashita, N. , 2016. Dawn arrives at Ceres: exploration of a small, volatile-rich world. Science 353, 1008–1010 .