Cratering and Tectonic History of the Largest Uranian Satellite, Titania: New Insights Enabled by Image Reprocessing

From heavily cratered Umbriel to extensively tectonized Miranda, Titania is an intermediary of the Uranian system: heavily cratered, yet tectonically modified. An outstanding mystery in Titania's crater population is its apparent relative lack of large (>30 km) craters. However, progress has been limited by the coverage and quality of images available. Here, we present a new map of Titania enabled by reprocessing Voyager images to reduce the effects of motion blur. Of note, we identify a network of fractures, a set of lineaments that may represent a large multi-ring impact structure, and newly identified catenae. These findings suggest Titania's crater population is missing large craters due to viscous relaxation, tectonic resurfacing, and/or planetocentric debris, and does not necessarily require cryovolcanic resurfacing. In preparation for future missions to the Uranian system, this work presents foundations for identifying imaging targets that can contribute to furthering our understanding of the history and evolution of the Uranian system in a broader context of icy satellite evolution.


Introduction
The largest of the Uranian moons, Titaniaʼs surface is both heavily cratered and tectonically modified.A network of fractures extends out from Messina Chasmata, a large fault zone (∼100 km wide and >6 km deep in some locations; Figure 1(A)) with apparent graben and bright scarps (e.g., Croft & Soderblom 1991;Smith et al. 1986;Schenk & Moore 2020).The majority of Titaniaʼs observed surface (∼southern hemisphere) is covered by heavily cratered plains (Figure 1(B); Croft & Soderblom 1991;Smith et al. 1986;Kirchoff et al. 2022).Compared to other Outer Solar System satellites (particularly Oberon and Umbriel, two other heavily cratered regular moons of Uranus), Titaniaʼs surface is relatively lacking in large (>30 km diameter) craters (Figure 2; Croft & Soderblom 1991;Smith et al. 1986;Strom 1987;Moore et al. 2004;Kirchoff & Schenk 2010, 2015;Schenk & Moore 2020;Kirchoff et al. 2022).Three hypotheses have been proposed to explain the dearth of large craters on Titania: (i) Titania experienced a significantly different impactor population than similar worlds in the Saturnian system and the Uranian system (such as debris from the rings and/or planetocentric impactor populations; Smith et al. 1986;Schenk & Moore 2020); (ii) Titania experienced at least one extensive cryovolcanic resurfacing event that overprinted early large craters (Croft & Soderblom 1991;Plescia 1987;Strom 1987;Moore et al. 2004); and (iii) the crater population on Titania is dominated by secondary impact craters from one or more large primary craters yet to be identified (Smith et al. 1986).These observations have led previous workers to posit that Titaniaʼs geologic evolution consisted of an early period of heavy cratering and near-complete cryovolcanic resurfacing, followed by the formation of tensional fractures and graben due to global expansion and possible smaller cryovolcanic eruptions (Croft & Soderblom 1991;Smith et al. 1986;Plescia 1987).
A better understanding of Titaniaʼs geology is crucial to answering outstanding questions not only about Titania itself, but also about the evolution of the Uranian system as a whole.
In particular, open questions include: the formation of the Uranian satellites in relation to the hypothesized giant impact that caused the extreme tilt of Uranus' rotational axis; the early impactor flux in the outer solar system; the long-term dynamics of the Uranian rings; and the existence of relict or modern subsurface oceans on Uranus' moons (e.g., Zahnle et al. 2003;Arridge et al. 2014;Wong et al. 2019;Ahrens 2021;Cartwright et al. 2021;Kirchoff et al. 2022;Rufu & Canup 2022).Modeling suggests that Titania could plausibly preserve an ocean to the present day and the geologic record can illuminate potential links between the surface and past states of Titaniaʼs subsurface ocean (Bierson & Nimmo 2022;Castillo-Rogez et al. 2023).As the Uranian system was listed as the top priority flagship destination in the 2023-2032 Planetary Science and Astrobiology Decadal Survey (National Academies of Sciences, Engineering, and Medicine 2022), it is important to revisit the Voyager 2 data of Titania in order to identify potential imaging targets for future missions to help resolve these fundamental outstanding questions.
Less than half of Titaniaʼs surface was imaged by Voyager 2 (e.g., Smith et al. 1986;Schenk & Moore 2020).The highest resolution images captured, shown in Figures 1(A) and 3(A), were taken at phase angles of 69.9°and 33.9°, with highest resolutions of ∼3.4 km pixel −1 and ∼4.6 km pixel −1 , respectively (Thomas et al. 1987).Thomas et al. (1987)  In the following sections, we describe the techniques used to reprocess Voyager 2 images of Titania to reduce the effects of motion blur and our methods to construct a new geologic map of Titania from the improved images.We highlight the key findings of previously unmapped features, as well as evaluate the three hypotheses for the lack of large craters observed on Titania, and we discuss the implications for the history of Titania and the Uranian system.

Image Reprocessing Methods
Techniques to reduce the effects of camera motion blur from images have significantly advanced since the Voyager 2 flyby of Titania, with the advent of modern post-processing methods that improve recovered image sharpness while reducing noise and artifacts (e.g., Fergus et al. 2006;Levin et al. 2011; Al- ) with officially named features labeled, as well as some additional names adopted from Croft & Soderblom (1991).Unless otherwise specified, names refer to craters.(B) An existing geologic map of Titania.et al. 2012;Yadav et al. 2016).Unlike object motion blur (when an object appears blurred because it is moving fast across a frame), camera motion blur affects every pixel in an image.Blur is typically not uniform across an image due to the motion of the camera relative to a three-dimensional imaging subject, and the problem of recovering a true image from a blurred image is an underconstrained problem (Fergus et al. 2006).Nevertheless, images with low to moderate motion blur and suitable contrast can be treated to reduce the effects of motion blur, even when an unblurred original image is not available (i.e., requiring a blind deconvolution).In particular, we apply an expectation-maximization maximum a priori (EM MAP) probability estimate, a type of nonuniform blind deconvolution (Levin et al. 2011).This EM MAP algorithm initially solves for the best deblurred image using an assumed blur kernel, then finds a best-fitting blur kernel given the image   and its covariance (Levin et al. 2011).Alternating between these two steps iteratively, the algorithm converges to an optimal deblurring of the images (Levin et al. 2011).For interested readers, a guide to using the method of Levin et al. (2011) is provided in the Appendix, along with a tutorial example in Figure A9.

Ameen
Though EM MAPs have been used extensively in other applications (e.g., astronomy, satellite imaging of Earth, traffic monitoring, and medical imaging; e.g., Marrugo et al. 2011;Al-Ameen et al. 2012;Mourya et al. 2015;Khare et al. 2016;Abinaya et al. 2017;Anger et al. 2019), we validate the application of this method to planetary data by comparing reprocessed Voyager images of Jovian satellites to images from more recent missions of similar regions (Figures 3 and A1-2).Following these successful benchmark exercises, we reprocess images of Titania to enable a new mapping of its surface features (Figure 4).

Geologic Mapping Methods
We conduct our mapping on the reprocessed version of the highest resolution, full-disk-coverage image of Titania obtained by Voyager (vg-iss-2-u-c2683649), using higherresolution, lower-coverage images for reference; this enables us to map more of the surface (∼40%) than has previously been mapped (∼30%; Croft & Soderblom 1991;Smith et al. 1986), though ∼60% of Titaniaʼs surface has not been imaged at a resolution suitable for mapping.We distinguish units based on morphology, albedo, and spatial relationships and map only the highest-confidence features, which we further divide into two classes of certainty.In particular, we have only mapped exceptionally bright or topographically welldefined craters larger than ∼25 km in diameter.Thomas et al. (1987) noted that the bright features resolved in Titania imagery encompass craters and crater ejecta, with the exception of brighter patches along scarps, which appeared  A2-A3 and Figures  A4-A5).The diameters and coordinates for all officially named features are listed in Table A1.
to be albedo differences, rather than bright reflections from Sun-facing slopes.Recent work by Kirchoff et al. (2022) noted that the best resolution areas in the imagery of Titania permit completeness in mapping craters as small as 15.5-17.5 km, but in this work, we additionally map the lower-resolution areas of the available imagery and have thus chosen a more conservative threshold.Note also that the stratigraphic relation of contacts between lineaments and fractures is largely uncertain at the available resolution.Although a formal map is beyond the scope of the present work, we work to incorporate established best practices where possible (Skinner et al. 2022).

Results
We reprocessed Voyager images of Titania (Figures 4(B The plains are heavily cratered with smaller craters (e.g., Kirchoff et al. 2022) and appear mottled, possibly because of cratering below the resolution of the available imagery.Larger (>25 km diameter) and/or notable craters (bright ejecta and/or particularly sharply defined topographic detail) are divided into three classes, based on the quality of their characterization.Class 3 craters (e.g., Valeria) are the most well-defined, characterized by sharp topographic detail (including central peaks or pits), bright ejecta rays, or extensive non-ray ejecta material.Class 2 craters (e.g., Imogen) are characterized by moderate topographic detail (though they may still appear to have a central peak or pit), but lack any obvious ejecta.Finally, Class 1 craters, many of which are newly observed in this mapping, are characterized by relatively poor topographic detail, such as a partially visible rim.Example images and further definitions of units can be found in Tables A2-A3.
Gertrude, identified by previous mapping efforts as a candidate impact basin (Croft & Soderblom 1991;Smith et al. 1986;United States Geological Survey 1988;McKinnon et al. 1991), is mapped as its own unit and possesses a central area of rough terrain surrounded by an intermediate zone of smooth terrain, and an outer zone of rough terrain that may be a degraded crater rim (e.g., Smith et al. 1986;Plescia 1987;Moore et al. 2004;Schenk & Moore 2020).An additional terrain unit observed on Titania is found interior to chasma, most notably Messina Chasmata.This terrain is bounded by inward-facing scarps and can contain additional downfaulted blocks.Emanating from these chasmata are a series of fractures (now more clearly observed in their full extent), the most obvious of which (Figure 4(C)) are associated with a clearly visible scarp face and/or high-albedo linear feature.Additional newly mapped fractures are characterized by linear segments that appear straight over ∼50 km or more.The fractures span most of the observed surface.
In addition to the linear segmented fractures, a series of curvilinear features were identified.These newly observed lineaments (Figure 4(C)) are distinguished from the fractures by appearing arcuate over relatively small lengths.Many of these lineaments appear to be approximately concentric, with an inner zone of more irregular and wavy lineaments and an outer zone of less wavy, curved lineaments.

Discussion
This new mapping of Titaniaʼs surface (Figure 4(C)) supports the general findings from the early mapping of Titania (Smith et al. 1986; Figure 1(B)), though there are several noteworthy differences and additions.Both the map presented in this work (Figure 4(C)) and the original maps Croft & Soderblom 1991;Smith et al. 1986) are dominated by cratered plains and largely agree on the boundaries of major features such as Messina Chasmata and Gertrude.The primary new results revealed by the updated mapping are: (1) a more extensive fracture network than previously reported; (2) additional craters and crater chains; and (3) a proposed multi-ring impact structure.
In addition to the fractures and scarps identified by Smith et al. (1986) and Croft & Soderblom (1991; Figure 1(B)) near Messina Chasmata, we mapped a significantly more expansive fracture network extending across the observed hemisphere.These newly identified fractures (e.g., Figure 5(A)) include some that split into multiple parallel branches, similar to those observed to cross Ursula crater.The broad extent of the fracture network supports the idea that these fractures, including larger features like Messina Chasmata, formed in response to global stresses, perhaps caused by the freezing of a subsurface ocean (Smith et al. 1986;Manga & Wang 2007;Berton et al. 2020) and/or tidal stresses (Schenk & Moore 2020).Croft & Soderblom (1991) mapped lineaments extending from Gertrude near the terminator.It is possible that additional lineaments may not be evident in the available imagery, due to the lighting geometry when the images were captured (Thomas et al. 1987).
Our mapping includes all previously mapped and officially named (United States Geological Survey 1988) craters except for: (1) one small unnamed crater (directly adjacent to Ursula in Figure 1(B), but not visible in Figure 4) and one candidate basin (below Gertrude in Figure 1(B)) identified by Smith et al. (1986), as well as (2) three large craters (north of the Bona Branch of Messina Chasmata), four intermediate size craters, and several small craters near the terminator in the Voyager images mapped by Croft & Soderblom (1991).Kirchoff et al. (2022) studied only a portion of the highest resolution imagery and mapped smaller craters than we considered, but there is generally good agreement with the larger craters (∼>30 km) we map in the study area Kirchoff et al. (2022) considered.We note that the challenge of consistently identifying smaller craters on Titania has been longstanding and is resolution-limited (Plescia 1987;Strom 1987;McKinnon et al. 1991;Kirchoff et al. 2022).Outside of the highest-resolution areas mapped by Kirchoff et al. (2022), we additionally mapped many other craters that have not been characterized before.
Of particular note in the craters mapped are several catenae (crater chains; e.g., Melosh & Schenk 1993;Stephan et al. 2013), as well as larger craters that show evidence for central peaks or domes.Catenae (e.g., Figure 5(B)) were identified primarily closest to the equator and have no obvious radial relationship to observed impact basins.Many craters mapped exhibit central peaks (e.g., Figure 5(C)), consistent with a previously determined simple-to-complex crater transition diameter of ∼24 ± 2 km (Croft 1987) and a lack of observed simple craters attributed to resolution limits by Schenk (1989).

A Proposed Multi-ring Impact Structure on Titania
The concentric curvilinear lineaments (Figure 4(C)) resemble a Valhalla-class multi-ring impact structure, defined by an inner irregular zone and an outer more linear series of concentric lineaments making up discontinuous rings (e.g., Schenk & McKinnon 1987).The rings extend outward to a diameter of ∼1300 km, slightly smaller than the Asgard multiring structure on Callisto (Figure 6).Although the ratio of the crater diameter to Titaniaʼs radius (∼1.65) is much larger than the predicted threshold crater size (a minimum ratio of crater diameter to Titaniaʼs radius of ∼0.6; McKinnon et al. 1991) that would lead to the disruption of Titania, and larger than the ratios for other large craters on icy moons (e.g., Herschel and Mimas: ∼0.7), it is similar to the ratio of Valhallaʼs diameter and Callistoʼs radius, ∼1.7 (e.g., McKinnon et al. 1991).If the concentric curvilinear lineaments we identify on Titania do indeed constitute a Valhalla-class multi-ring impact structure, then we argue that, as in the case of Valhalla, the original crater is much smaller than the total structureʼs diameter and is within the central palimpsest (McKinnon et al. 1991).Additionally, the existence of a multi-ring impact structure on Titania could provide a link between global fracture networks and large impact features on other icy moons (e.g., Odysseus on Tethys; McKinnon et al. 1991;Moore et al. 2004).We note that Messina Chasmata and several of its most prominent branches are approximately radial to the proposed multi-ring impact structure.We note that due to the low resolution of the available imagery, it is not possible to completely rule out other features besides multi-ring impact basins that occur as concentric lineaments on icy worlds (e.g., Enceladus' southern curvilinear terrain and Mirandaʼs coronae).
If subsequent imaging of Titaniaʼs surface confirms this multi-ring impact structure, it would represent the first Valhalla-class multi-ring impact structure detected outside the Jovian system.More detailed measurements of ring spacing could enable an estimate of the past thermal and rheological structure of Titania (e.

Paucity of Large Craters
Three hypotheses have been proposed to explain the dearth of large craters on Titania: (i) Titania experienced a significantly different impactor population than similar worlds in the Saturnian system and the Uranian system (such as debris from the rings and/or planetocentric impactor populations; Smith et al. 1986;Schenk & Moore 2020); (ii) Titania experienced at least one extensive cryovolcanic resurfacing event that overprinted early large craters (Croft & Soderblom 1991;Plescia 1987;Strom 1987;Moore et al. 2004); and (iii) the crater population on Titania is dominated by secondary impact craters from one or more large primary craters yet to be identified (Smith et al. 1986).
In addition to these hypotheses, it is possible that older craters may have been deformed or erased by (iv) tectonic resurfacing and/or (v) viscous relaxation.Tectonic resurfacing, as observed on Ganymede (e.g., Head et al. 1997;Pappalardo & Collins 2005;Pappalardo et al. 2023) and proposed for Ariel (e.g., Beddingfield et al. 2022), may have occurred on Titania during periods of time associated with global expansion (such as during ocean freezing) or with stresses from past resonances with other Uranian satellites.Certainly, the presence of widespread tectonic features on Titaniaʼs surface (Figure 4(C); fractures and scarps), as well as the large Messina Chasmata, attest to significant tectonism in Titaniaʼs past that should be considered in interpretations of Titaniaʼs cratering record.Additionally, viscous relaxation on icy bodies erases larger craters more rapidly than smaller craters and may thus be responsible for an apparent deficit of larger craters (Parmentier & Head 1981).For instance, Gertrude (326 km diameter) has a maximum depth of ∼1-3 km, which is significantly less than the maximum depth predicted for a large, complex crater of its size (McKinnon et al. 1991).This strongly suggests that viscous relaxation (or some other flattening/infill process) has modified this crater.
In light of the new geologic map (Figure 4(C)), we revisit the hypotheses previously posed to explain the lack of large craters on Titania (i-v).The identification of the proposed multi-ring impact structure, newly mapped >30 km diameter craters, and the updated crater age dating for heavily cratered terrains at ∼4.5 Ga (Kirchoff et al. 2022) suggest that extensive cryovolcanic resurfacing may not be sufficient to explain the dearth of large craters (>30 km) on Titania.Although the proposed multi-ring impact structure, as well as larger craters like Gertrude, could form ejecta capable of producing secondary impact features, such ejecta is expected to escape Titania (McKinnon et al. 1991).On the basis of these data and observations, we suggest that the most likely explanation for Titaniaʼs observed paucity of large craters may be due to the significant viscous relaxation of larger craters preferentially and overprinting, perhaps by tectonic resurfacing (e.g., Head et al. 1997).Further mapping work with higher resolution imagery than is presently available (i.e., enabling mapping of smaller craters across a greater extent of Titaniaʼs surface) is required to assess whether differences in the impactor population at Titania also contribute to the distribution of crater diameters on Titania (Figure 2).
We propose that Titaniaʼs early history was dominated by heavy crater bombardment, followed by a period of global tectonism, perhaps driven in part by ocean freezing, in agreement with Plescia (1987) and Smith et al. (1986).In contrast with the geologic evolution proposed by Plescia (1987) and Smith et al. (1986), we argue that near-complete cryovolcanic resurfacing is not necessary to explain the observed paucity of large craters on Titania, calling in addition, to the major role of crater viscous relaxation and tectonic resurfacing.

Conclusions
We used a nonuniform blind deconvolution method to reduce motion blur in Voyager-era images of Titania.The improved clarity of the reprocessed images enables updated geologic mapping with improved coverage over previous efforts (Smith et al. 1986).Based on the identification of a feature that we interpret as a multi-ring impact structure, as well as other newly mapped larger (>30 km diameter) craters, we suggest that the long-noted dearth of large craters on Titania is best explained by the tectonic overprinting and viscous relaxation of large craters.It is plausible that planetocentric debris (including interaction with the Uranian rings) could also contribute to the relative abundance of small craters.However, in contrast to previous work, we argue that near-complete cryovolcanic resurfacing of Titania is not required to explain its apparent lack of large craters.Additionally, the detection of a more extensive fracture network supports the idea that chasmata-forming tectonism on Titania was driven by global stresses, such as those that result from the freezing of a subsurface ocean (Smith et al. 1986).
The surface of Titania is a rich scientific target.Future imaging and mapping has the potential to shed light on the early impactor flux in the Uranian system, a key to understanding the formation and evolution of the Uranian satellites in relation to the event that tilted Uranus' rotational axis (e.g., Wong et al. 2019;Cartwright et al. 2021).Further mapping of Titaniaʼs surface may also place constraints on moon-moon resonance interactions between the Uranian satellites (e.g., Tittemore 1990;Ćuk et al. 2020).Moreover, if confirmed, the existence of a multi-ring impact structure as well as an array of complex crater morphologies would make Titaniaʼs crater population ideal for comparative study with those on other icy worlds, including Callisto, Ganymede, Rhea, Tethys, and Pluto.Modern image processing methods such as nonuniform blind deconvolution motion blur correction have the power to improve the science return of imaging data sets across the solar system (e.g., Figures A6-A8) and identify scientific targets for future missions.
Below we present a step-by-step guide for replicating the image processing described in the manuscript.This step-bystep guide is accompanied by a figure (Figure A9) showing the image (starting with Voyager frame vg-iss-2-u-c2683649) after each step listed below, so that this guide may be used as a tutorial.For some steps, we list instructions for either MATLAB or Adobe Photoshop, though MATLAB is the preferred processing method.This methodology does not retain the geospatial information associated with the image.To use these images in a geospatial software, the images must be referenced to other georeferenced images.
1. Download the full resolution desired image frame and load the image into the desired image processing program.The image may also be cropped at this stage, if desired.2. Interpolate the data for the control points.The overall processing method is not sensitive to the interpolation method so long as: (1) the total area interpolated is small compared to the total image area and (2) the interpolation is smooth.A2.A lineament that is ∼linear over >50 km.
provide a complete discussion of the phase curves and opposition effects for Titaniaʼs surface.Early geologic mapping of Titania by Smith et al. (1986) Figure 1(B)) based on the highest resolution image available (Figure 1(A)) recognizes three geologic units and several feature classes: cratered terrain, smooth material, bright material, craters, chasmata (elongate, steep-sided topographic lows), and rupes (scarps).Later mapping work by Croft & Soderblom (1991; Figure 1(C)) subdivided the plains and cratered units and mapped additional craters and tectonic features.Most recently, Kirchoff et al. (2022) mapped craters in a portion of Titaniaʼs surface that was imaged at the highest resolution.Unfortunately, geologic mapping has been limited not only by image coverage and resolution, but also by image quality.Due to the high relative velocity of Voyager 2 during its Titania flyby, the images are characterized by decreased image sharpness due to camera motion blur.

Figure 1 .
Figure 1.(A) Highest resolution Voyager 2 image of Titania (vg-iss-2-u-c2684313; highest resolution: ∼3.4 km pixel −1 ) with officially named features labeled, as well as some additional names adopted from Croft & Soderblom (1991).Unless otherwise specified, names refer to craters.(B) An existing geologic map of Titania. Figure modified from Smith et al. (1986)."Voyager 2 in the Uranian System: Imaging Science Results."Science 233 (4759), p. 59.Reprinted with permission from AAAS.(C) Another geologic map of Titania. Figure modified from Croft & Soderblom (1991).From Uranus, edited by Jay T. Bergstralh, Ellis D. Miner, and Mildred Shapley Matthews.© 1991 The Arizona Board of Regents.Reprinted by permission of the University of Arizona Press.(D) Craters mapped by Kirchoff et al. (2022) in the areas of Titaniaʼs surface imaged at highest resolution.Figure modified from Kirchoff et al. (2022).
Figure 1.(A) Highest resolution Voyager 2 image of Titania (vg-iss-2-u-c2684313; highest resolution: ∼3.4 km pixel −1 ) with officially named features labeled, as well as some additional names adopted from Croft & Soderblom (1991).Unless otherwise specified, names refer to craters.(B) An existing geologic map of Titania. Figure modified from Smith et al. (1986)."Voyager 2 in the Uranian System: Imaging Science Results."Science 233 (4759), p. 59.Reprinted with permission from AAAS.(C) Another geologic map of Titania. Figure modified from Croft & Soderblom (1991).From Uranus, edited by Jay T. Bergstralh, Ellis D. Miner, and Mildred Shapley Matthews.© 1991 The Arizona Board of Regents.Reprinted by permission of the University of Arizona Press.(D) Craters mapped by Kirchoff et al. (2022) in the areas of Titaniaʼs surface imaged at highest resolution.Figure modified from Kirchoff et al. (2022).

Figure 3 .
Figure 3. (A) Voyager frame (vg-iss-1-j-c1640354) taken in 1979 of Shu Crater (the largest bright crater in the image) and the surrounding area on Ganymede, cropped and contrast-adjusted.The black dots are control points.(B) The same as (A), but reprocessed to reduce camera motion blur.Note that even small-scale features (small bright craters, grooves, and crater rays) are now recovered.(C) Juno image (NASA/SwRI/MSSS: JNCE_2021158_34C00002_V01; ∼1-2 km pixel −1 ) from 2021 June 7 of a similar region.

Figure 4 .
Figure 4. (A) Voyager frame (vg-iss-2-u-c2683649; best resolution: ∼4.6 km pixel −1 ) cropped and contrast-adjusted.The black dots are control points.(B) Same as (A), but reprocessed to reduce camera motion blur.Note the improvement in image sharpness.(C) New geologic mapping in the same orientation as (A) and (B), based on reprocessed images of (B) with a legend.Triangles point in the downslope direction for scarps.Black regions were too poorly illuminated to map.More details on the definition of map units and explanation of map symbols, along with examples of terrain types, can be found in the Appendix (Tables A2-A3 and Figures A4-A5).The diameters and coordinates for all officially named features are listed in TableA1.
) and A3) and produced a new geologic map based on these reprocessed images (Figure 4(C)).Titaniaʼs observed surface can be divided into cratered terrains (Figure 4(C); lavender), the smooth and rough terrains in Gertrude (Figure 4(C); bright and dark purple), crater material (Figure 4(C); shades of yellow), and terrains interior to chasmata (Figure 4(C); teal).The majority of Titaniaʼs observed surface is classified as cratered plains.
g., McKinnon & Melosh 1980; Stephan et al. 2013) and constraints on Titaniaʼs possible participation in resonances among the Uranian satellites, including those hypothesized to be responsible for geologic activity on Miranda and Ariel (e.g., Tittemore 1990; Ćuk et al. 2020).

Figure 5 .
Figure 5. Examples of features of interest identified in the updated mapping of Titania.Individual feature images have been contrast-adjusted.Outlines on reference images show the extent and context of the feature images.The base image of Titania is the same as Figure 4(B).Subpanels (A)-(C) are taken from the base image of Titania.(A) Example of a branching fracture on Titania.(B) A crater chain on Titania.(C) Valeria (59 km diameter) is a crater with a central peak.

Figure 6 .
Figure 6.Comparison of proposed multi-ring impact structure identified on Titania to multi-ring impact structures on Callisto.(A) Reprocessed full disk view of Titania contrast-enhanced to highlight features.(B) Close-up view of several ring segments.(C) The subset of lineaments in Figure 4(C) corresponding to the proposed multi-ring impact structure (outer zone diameter ∼1300 km).(D) Voyager image of Callisto.The approximate extent of Asgard, a multi-ring impact structure with a diameter of ∼1600 km.Note that although these structure diameters are similar, Callisto is much larger than Titania.(E) Mapping of Asgardʼs multi-ringed structure from a global cylindrical projection of Voyager data.Figure component modified from Schenk & McKinnon (1987).Reprinted from Icarus, 72 (1), Schenk & McKinnon, "Ring Geometry on Ganymede and Callisto," p. 214, © 1986, with permission from Elsevier.
3. Adjust the black point of the image such that the value of pixels showing the background are zero.4. Deblur the image: (a) Using MATLAB: the code associated with Levin et al. (2011).The downloaded files contain everything needed to run the code and an informative README file.Convert the loaded image file into a double array.Then use the "deconv_diagfe_filt_sps" function with the image double array as the first input and a kernel size as the second and third inputs (for the processing shown in the manuscript, we used 25, but this can be adjusted for images with different degrees of blurriness and contrast).The function has three outputs, the second of which is the processed image.The resultant image array can be worked with further in MATLAB or exported to a JPEG.

Figure A2 .
Figure A2.(A) Voyager frame (vg-iss-1-j-c1639046) of Huo Shen Patera, to the right of (1), Horus Patera (2), Kava Patera (3), and the surrounding area on Io.The black dots are control points.(B) The same as (A), but reprocessed using a nonuniform blind deconvolution to reduce camera motion blur.(C) Unprocessed Voyager frame (vg-iss-1-j-c1639044) with some overlapping coverage with (A) and (B).Common features are numbered across all panels.

Figure A3 .
Figure A3.Test of result sensitivity to specific EM MAP probability estimate nonuniform blind deconvolution used.The results produced in panels (B) and (C) are nearly identical; the primary difference is the treatment of noise from bad pixels in the original image.(A) Voyager frame (vg-iss-2-u-c2683649) cropped and contrastadjusted.The black dots are control points.(B) The same as (A), but reprocessed using the Shake Reduction tool in Adobe Photoshop CC 2018.(C) The same as (A), but reprocessed using the MATLAB codes from Levin et al. (2011).

Figure A4 .
Figure A4.Larger version of the map from Figure 4(C), with named features labeled.

Figure
Figure A6.(A) Voyager frame (vg-iss-2-u-c2684539) of Ariel.The black dots are control points.(B) The same as (A), but reprocessed using a nonuniform blind deconvolution to reduce camera motion blur.

Figure A8 .
Figure A8.(A) Voyager frame (vg-iss-2-n-c1139617) of Triton.The black dots are control points.(B) The same as (A), but reprocessed using a nonuniform blind deconvolution to reduce camera motion blur.

Table A3
Explanation of Map Symbols for Figure 4(C)