Regional Geology of the Hypanis Valles System, Mars

We present a geomorphic map of the Hypanis Valles watershed and a geomorphic map of the Hypanis deposit region at its terminus. We mapped these two regions at different scales: 1:2,000,000 for the catchment map (−5° to 10°N and 300°–315°E) and 1:500,000 for the Hypanis deposit map (10°–13.0°N and 313°–316.5°E). Our mapping provides new morphologic insights beyond previous efforts which used lower spatial resolution data. We defined units based on morphology, albedo, thermal inertia, elevation, and spectral parameters. We propose that episodic volcanism and aqueous activity filled the Chryse basin from the early Noachian. Hypanis Valles was active during the Noachian, forming the Hypanis terminal deposits in the southern Chryse region. Hundreds of kilometer‐sized mounds and cones stratigraphically post‐date Hypanis fluvial deposition as these features appear to have erupted or effused through all other major map units. We propose sedimentary diapirism or mud volcanism may be responsible for these features, a hypothesis consistent with the compressional wrinkle ridge tectonism in a sedimentary basin. Future work could further investigate the formation of these cones and mounds and better assess their astrobiologic importance.


Hypanis Deposit Geologic Context
The Chryse escarpment is the modified and potentially warped rim (Citron et al., 2018) of the Chryse basin, which separates the Noachian highlands from Hesperian lowlands in the region (Figure 1). The Chryse basin currently has no topographic boundary in the north, and blends into the larger northern lowlands basin of Vastitas Borealis, although evidence exists defining the extent of the ancient impact structure (Pan et al., 2019).
Two regions of the Chryse basin have been explored by rover and lander spacecraft. The Viking I lander and Pathfinder lander with Sojourner rover have each explored the basin fill and superficial deposits above the Chryse basin. Other basin fill materials of the northern lowlands closer to the pole were investigated by the Phoenix lander mission and from orbital studies Smith et al., 2009) which documented buried ice and surface ice. Orbital studies indicate Chryse and the northern plains contain sufficient buried ice to form rampart craters with flow-like (or slurry-like) ejecta lobes, and this morphology is observed as far south as the Noachian highlands in our mapped catchment region (Reiss et al., 2006).
The Chryse basin is filled with sedimentary and volcanic units. The sources of basin fill likely originate from valley network activity (4.0-3.2 Ga), a proposed series of northern oceans (Carr, 1996(Carr, , 2000Carr & Head, 2009;Clifford & Parker, 1999;Head et al., 1998;Parker, 1994; and references therein), Hesperian volcanism (3.6 Ga), outflow channels (3.2 Ga), and other events. Previous geologic and geomorphic maps of the region are at a much larger scale than those produced in this work and are based on various imaging data sets at lower spatial resolution (D. Scott & Tanaka, 1986;D. H. Scott & Carr, 1978;Rotto & Tanaka, 1995;Tanaka et al., 2014).

Previous Work on the Hypanis Deposit
The first work providing a detailed analysis of the Hypanis deposit was Hauber et al. (2009). The authors found it difficult to determine whether the Hypanis and Sabrina deposits were alluvial or deltaic based on a lack of clear evidence for a standing water level such as terraces, shorelines, or a delta front (Hauber et al., 2009). They also interpret that the Hypanis deposit was fed by the combination of both Hypanis Valles and Nanedi Valles, which were connected at that time, thus explaining the large volume of the Hypanis deposits. Subsequently, Kleinhans et al. (2010) presented a model for forming the neighboring Sabrina deposit as an alluvial fan, with a short timescale of formation. Later, Adler et al. (2019) and Fawdon et al. (2018) have provided measurements in support of the deltaic formation hypothesis for Hypanis, and argue that Sabrina was likely similarly formed. These works found a progressive basinward advance of the channel-lobe transition, plausible avulsion node migration, and stacks of roughly horizontal sediment layers indicating a subaqueous environment. These studies noted a lack of evidence of the characteristic alluvial fan architecture (e.g., fixed avulsion node, larger clasts in young channels), but also did not identify a clear topset-foreset slope transition, a feature diagnostic of deltas. It could be because 10.1029/2021JE006994 4 of 22 of the unprecedented size of the deposit and high rates of subsequent erosion that this transition would not be preserved (Adler et al., 2019).
The timing of the Hypanis deposition was dated both from Context Camera (CTX)-derived impact crater statistics in the catchment on an ejecta blanket that superposes Hypanis Valles, and from buffered crater counts on Hypanis Valles (Fawdon et al., 2018). Both methods were consistent with previous age constraints and place the Hypanis deposit as older than 3.6 Ga. The presence of surrounding units including dark flows, wrinkle ridged plains, mesas and mounds, were noted but their stratigraphy and timing relative to Hypanis were not examined in detail. Adler et al. (2019) posited that the plethora of rounded buttes observed mostly in the northern portion of the study region could potentially represent former crater rim remnants, thus allowing for a land dam to enclose Hypanis in a local basin. This current work reassesses that possibility as detailed mapping has led to an improved interpretation of surrounding units.
In this work, we mapped the surroundings of the Hypanis deposit and the units that Hypanis and Nanedi Valles carved through. As a result, we have created a comprehensive list of the materials potentially present in the deposit, and present a better understanding of the regional geologic history surrounding the largest potential delta on Mars.

Geomorphic Mapping
We divided mapping into two regions, a deposit region and a catchment region. The deposit region maps serve to characterize the geology and geomorphology of the units immediately surrounding the Hypanis fan-shaped layered deposit (Figure 2). Two maps of the deposit region were produced: a major unit map to display the floor units and large scale trends ( Figure 3) and a detailed map including features as small as ∼500 m in size ( Figure 4). The deposit maps cover the area from 10° to 13.5°N, 313°-316.5°E, and are 1:500,000 in scale. We used a CTX global mosaic (6 m/pix) as a basemap for drawing contacts (Dickson et al., 2018). The global CTX map was streamed in JMARS (Java Mission-planning and Analysis for Remote Sensing; Christensen et al., 2009), and individual 4° × 4°CTX mosaic tiles were also uploaded to JMARS for enhanced control of color stretching. Units were defined based on a variety of strict quantitative and qualitative guidelines from other data sets which were processed, projected, overlayed, and aligned to the basemap (e.g., albedo, thermal inertia [TI], spectral parameters, morphology, elevation, and stratigraphic relationships) described further in the description of map units (Tables 1-3) and Figures 5-8. A map of the catchment region was also produced at 1:2,000,000 scale, containing features down to a size of 5 km in a similar quantitative and qualitative fashion (Figure 7). The 5 km size limit was chosen because of the large area of the catchment. The catchment map contains the major units that comprise the floor materials that may have traveled down Hypanis Valles and into the Hypanis deposit and deposit region. Several units are discussed within the context of the Tanaka et al. (2014) global geologic map of Mars. For consistency with similar work, some unit names and morphologies follow the same mapping conventions presented in Goudge et al. (2015) of Jezero crater and its watershed. All unit mapping was performed in JMARS v.4.0.20.
To define the paleo-catchment boundary (Figure 1), we used ArcMap (ESRI, 2011) Hydrology tools in a similar fashion to Fawdon et al. (2018). We utilized the High Resolution Stereo Camera (HRSC) and Mars Orbiter Laser Altimeter (MOLA) Blended Global Digital Elevation Model (DEM) v2 created by the USGS (R. L. Fergason et al., 2018) the initial catchment surface, whereas the prior model was based on the MOLA Global DEM at a lower resolution. A standard ArcGIS hydrology workflow was used including filling sinks (erroneous cells in the DEM with an undefined drainage direction), defining a pour point (the outlet where a watershed pours), and calculating streams. The pour point was chosen on the eastern side of Hypanis to include channels relevant to

of 22
the main lobe and braided channels. Both the Hypanis and Nanedi systems were likely connected at the time of formation of the Hypanis deposit, and this was included in the model. The watershed was thus defined as the area that topographically drains into the pour point. Areas in the catchment that postdate Hypanis Valles are marked in Figure 1. Despite the increased resolution of the catchment DEM, the resulting stream network is roughly identical to the one previously presented in Fawdon et al. (2018).
In the deposit map region, higher resolution DEMs were available. We utilized multiple products from High Resolution Imaging Science Experiment (HiRISE) and CTX stereopairs that were previously described in Adler et al. (2019) including our derived regional DEM mosaic (Adler, 2022) for perspective views of units, height and volume measurements, and stratigraphic contact investigations (Malin et al., 2007;McEwen et al., 2007). The DEM mosaic of the Hypanis region utilized in this study and others is described in Adler et al. (2019) and is Several units are defined by their Thermal Emission Imaging System (THEMIS) TI and daytime and nighttime temperature (Christensen et al., 2004;R. L. Fergason et al., 2006). The THEMIS Global Day, Night, Night/Day, and TI mosaics were often used to assist with locating unit contacts, and were useful for capturing the extent of   Figure 3. Features as small as 500 m were mapped using albedo, morphology, spectral parameters, and thermal inertia as quantitative and qualitative factors. Examples of some mapped features at this scale include crater units, phyllosilicate bearing material, cones, mounds, plateaus, and flows. Map scale is 1:500,000. Image is 200 km across and spans from −5° to 10°N and 300° to 315°E. impact crater ejecta. Higher resolution THEMIS Day, Night, and Night/Day mosaics were utilized in the deposit map region (R. Fergason et al., 2007).
The interpreted cross section includes the topographic profile derived from our DEM mosaic (Adler, 2022) of the deposit region in JMARS. For clarity visualizing the stratigraphic sequence, the original high-resolution noisy profile was smoothed in areas using the simplify function in Inkscape and deleting outlier values.

Mineralogic Mapping
Study of the mineralogy of the Hypanis deposit is limited due to the widespread cover of dust, dunes, and TARs (Adler et al., 2019). The presence of dust is shown in Figure 9 where dustier than an index of 0.96 (Ruff & Christensen, 2002). Furthermore, high-resolution targeted CRISM (Compact Reconnaissance Imaging Spectrometer for Mars) observations are limited to only a few locations in the deposit and catchment region. CRISM footprints of IR images with resolution <100 m/pix are shown in Figure 9 (Murchie et al., 2007). To assess the potential compositions over such a large region of Mars, we compiled previously published observations which were based on CRISM and TES (Thermal Emission Spectrometer) data ( Figure 9; Christensen et al., 2001).
TES derived mineral maps were published at ∼14 km/pixel resolution (4 ppd) and previously published CRISM observations are based mostly on targeted images at ∼20 m/pixel resolution or on lower resolution mapping products. Previously published analysis of Multispectral Survey (MSP) images (200 m/pixel) and Full Resolution Targeted (20 m/pixel) observations in the deposit map region were also incorporated (Adler et al., 2019). In this Unit map symbol Unit name Unit description Unit interpretation

Crater units
C Crater (D < 10 km) Crater ejecta, rims, and floor material of all craters with a diameter 1-10 km Impact crater and associated deposits of diameter 1-10 km Cpa Crater pitted apron Eroded, pitted unit within crater rampart ejecta (Cr) where presumably mantling fan deposits (Fd) Eroded and pitted impact crater ejecta where presumably mantling fan deposits Cr Crater rampart ejecta (D > 10 km) Crater ejecta, rims, floor material, and central peaks of all craters with a diameter >10 km. Rampart craters with lobate, flow-like ejecta Impact crater and associated deposits of diameter >10 km. All craters of this size appear to be rampart craters implying subsurface ice at the time of formation   Figure 4) previous work, we found that the High-Calcium Pyroxene and the D2300 parameters (Viviano-Beck et al., 2014) were useful mapping indicators in the deposit region.

Chryse Basin Units-"Deposit Map"
Within the southern Chryse basin area covered by the deposit map (Figures 3 and 4), major units range from the mid-Noachian to early Hesperian and possibly younger. The floor material to the south is generally older, and material to the north progressively younger. In the Tanaka   To the north of mNh, the Chryse skirt unit (Cs) lies along the Chryse escarpment. This unit reappears again along the Chryse escarpment to the northwest of the mapped region, circumscribing the Chryse basin. This unit has a strong slope component (topographic profiles show a steady basinward slope over tens of kilometers) and connects the highlands elevations to the lower plains. Cs is also marked by a high THEMIS nighttime temperature and high TI. In Adler et al. (2019), regionally mapping of CRISM MSP images showed Fe/Mg phyllosilicate signatures in the deposit region. From our mapping, most of these phyllosilicate bearing materials occur in the Cs unit. The teal color within the Cs and Lr units in Figure 4, shows the locations of the phyllosilicate-bearing material.
Cs is found elsewhere along the edge of the Chryse basin outside of the mapped deposit region, extending up to 400 km to the northwest along the Chryse escarpment. Outside of the mapped region to the northwest, Cs appears to be the only major unit dividing the Noachian highlands from the Hesperian plains.  Table 2.
The 90 km Lederberg crater in the deposit map region is heavily modified by surface deposits and volcanism. We mapped three units within this complex: the Lederberg rim (Lr) and two Lederberg floor units (Lf1 and Lf2). Features in Lederberg crater are investigated in detail in Brož and Hauber (2013). In our mapping, we distinguished Lr as having a high elevation, following the curved crater rim shape, having ridges with high relief, darker valleys, and a multitude of small mounds. Lf1 is lighter-toned than its surroundings and contains dark patches in topographic lows associated with quasi-circular crater/cone rims. Lf2 is a smooth plains unit that is lightly cratered.
The plains fill unit (Pf) is similar to Lf2 but occurs outside of Lederberg crater and appears on top of mNh, Cs, and lNh. Pf is marked by flat lying plains with few small craters and sometimes a raised circular perimeter. The channel deposits (Cd) are light-toned, layered, low TI, fan-shaped units discussed in previous works. Finally, the late-Noachian highlands (lNh) unit is reassessed for this work and mixes some of the characteristics from each of the Tanaka et al. (2014) lNh and eHt units. We map lNh as a mostly plains forming unit containing grabens, wrinkle ridges, cones, mounds, and mesas.
Within the southern Chryse basin covered by the detailed deposit map (Figure 4), geomorphic units appear to belong to four categories: Channel, Volcanic, Crater, and Regional. The regional units were described above and are the major units mapped in Figure  Here, we describe in detail the Volcanic units (Df, Ca, Co, Mo, Md, and Pl) mapped in Figure 4. The term volcanism is used in the general sense, as we consider a variety of formation mechanisms including mud volcanism, igneous volcanism, and volatile release. Examples of each of these units are shown in Figures 5 and 6. Dark flows (Df) have a low albedo and higher TI than their surroundings. Df is a crater-retaining unit often found as a capping unit or in topographic lows and with flow-like margins. Calderas (Ca) are cone shaped or oblong and have a large central crater with a low or collapsed rim. Most calderas are within the Lederberg floor unit (Lf1) as summits of presumed volcanic edifices, or within the late-Noachian highlands (lNh) plains.
Cone (   mounds (Md) are large mounds with a roughly radial dissection, faulting, or collapse. Plateaus (Pl) are mesas larger than 2 km with non-round shapes and usually taller than cones or mounds.

Hypanis Deposit Units-"Catchment Map"
Within the Hypanis watershed region in Xanthe Terra covered by the catchment map, major units span from the early Noachian to the Hesperian. Most of the catchment floor is composed of mid-Noachian highlands (mNh) and early Noachian highlands (eNh), described Tanaka et al. (2014) and re-mapped for this work. Early Noachian highlands occur as isolated regions that are more heavily cratered and with greater relief than mNh. Large craters with a diameter greater than 50 km are labeled as unit (C) where their ejecta and morphology are not majorly degraded or buried. One difference from previous maps is the extent of ejecta from the "Unnamed crater" in Within the mNh, eNh, and other major catchment units mapped and described above, several crater morphologies are observed. Craters with slurry ejecta and craters with ejecta and central pits are common throughout the catchment region. Craters with floor fill appear significantly buried or modified and are also ubiquitous in the catchment region. Craters with a chaos floor are found only in few locations and near the eastern border of the catchment. Craters with pedestal style ejecta blankets are much smaller in diameter than the previously mentioned catchment craters and are widespread. A few smaller diameter craters have either dark ejecta or a young appearance with rays.

Deposit Region
While some units in the deposit region exactly match the description and interpretation of units in Tanaka et al. (2014) (e.g., mNh and eHt), others do not (Table 1). For example, a departure from the global map is the unit we designate lNh, which merges relevant parts of the descriptions and interpretations from the global [lNh] and [eHt]. This is likely a result of mapping at a finer scale within a "transition zone." Higher resolution images show mounds and mesas within the wrinkle ridged plains, not separated. The unit we designate Cs is also not present within the global map, yet can be identified at both our map scale and at the global map scale. We assume from the slope of the Cs unit and the cross cutting of Hypanis Valles that Cs underlies the Hypanis deposit, and filled the southern Chryse region prior to Cd and lNh ( Figure 10). Thus, Cs may have perhaps been part of the Chryse impact basin wall or floor, and where exposed may be synonymous with the Chryse escarpment. If the Chryse basin were filled with water in the past, perhaps the present day albedo, thermal signature, and gentle regional slope are related to a former ocean or sea. It is interesting that the Cs unit has very few cones or mounds compared to the adjacent lNh floor unit (Figure 4). This could mean that Cs is a less volatile-rich unit, or that it is more difficult to fracture or form wrinkle ridges. The abundance of cones and mounds to the north of the Cs unit may also be a clue to the past water level. As Cs roughly follows elevation contours, its relation to water level or an original Chryse impact structure is somewhat expected. Cs occurs above the water levels proposed for Sabrina and Hypanis deposits in Fawdon et al. (2018) and Adler et al. (2019). Cs also contains the bulk of the phyllosilicate-bearing units (Figure 4). The relative lack of phyllosilicate detections in lNh may be due to a number of factors, including that lNh formed after this period of alteration or that lNh material is less conducive to this style of alteration. Figure 9. Mineralogy previously identified in the Hypanis catchment and deposit region compiled from multiple publications. Regions dustier than an index of 0.96 are mapped (Ruff & Christensen, 2002). High-resolution CRISM IR observation footprints are outlined in orange. Previously published mineralogy includes hydrated silica, kaolinite, and feldspar Pan et al., 2021;Wray et al., 2013); hydrated sulfate, phyllosilicate (Adler et al., 2019;Carter et al., 2014;Ehlmann & Edwards, 2014 and references therein); hematite (Thermal Emission Spectrometer [TES] Hematite Abundance >0.10, Bandfield, 2002); olivine (TES Olivine Abundance >0.10, Koeppen & Hamilton, 2008); high-Ca pyroxene (TES HCP Abundance >0.10, Bandfield, 2002); phyllosilicate/hi-Si glass (TES High-Si Glass Abundance >0.10, Bandfield, 2002); and plagioclase (TES Plagioclase Abundance >0.10, Bandfield, 2002). Where TES pixels have more than one mineral abundance above the detection limit, pixels are colored according to the legend. The southern catchment is more clearly identified as basaltic in composition. This may either be due to dust cover or true differences in the underlying composition. Proximal, middle, and distal catchment regions described in the text are separated by dashed lines at 300 and 700 km from the Hypanis fan deposit (catchment pour point). The Lederberg crater region has previously been interpreted by Brož and Hauber (2013). We agree with their interpretations that floor unit (Lf1) contains explosive and effusive volcanic features including breached cones (perhaps from hydrovolcanic eruptions), volcanos, and dark flows. We interpret the higher relief ridges, mounds, and dark valleys in the Lederberg rim unit (Lr) to be erupted material that later followed fractures and conduits from the Lederberg impact. The plains Lederberg floor unit (Lf2) closely resembles Pf and other mantling plains in the region, thus we assume formation occurred by a similar mechanism.
Crater and channel units within the detailed deposit map follow descriptions and interpretations as presented in other works in greater detail including Goudge et al. (2015), Hynek et al. (2010), Fawdon et al. (2018), Knade et al. (2017), and Adler et al. (2019). One exception is the Crater pitted apron unit (Cpa) with a unique morphology that is only present where possibly mantling a Hypanis delta lobe (Fawdon et al., 2018, Figure 7a).
We interpret six units as volcanic in origin. Dark flows (Df) are interpreted as volcanic due to crater-retention, lower albedo, flow-like margins, widespread occurrence in topographic lows, as plains, or as a flat capping unit. Plateaus or mesas (Pl) we also interpret as possibly volcanic material due to similar morphologies as (Df) as the top material of some of these features. We interpret these isolated plateaus as highly eroded due to their shape and separation from one another, which suggests a high-standing continuous layer now broken apart.
The interpretation of regional volcanic activity extends more widely than in the mapped deposit region. Cone and mound shaped features are widespread in the Chryse Planitia, Acidalia Planitia, Arabia Terra, and Chryse outflow regions. Mounds and cones with varying morphometric properties in the broader region have been interpreted as either igneous eruption features or sedimentary eruption (e.g., mud volcano) features (Brož et al., 2015;Dapremont & Wray, 2021;Farrand et al., 2005;Harrison, 2012;Hauber et al., 2015;Komatsu et al., 2016;Meresse et al., 2008;Oehler & Allen, 2010;Okubo, 2016;Tanaka, 1997Tanaka, , 1999. While we cannot be certain without further data or analysis, we favor a sedimentary volcanic interpretation for these features. Cones and mounds appear to be the youngest geologic units in the region. Three observations support this interpretation that the formation of cones and mounds is the most recent observable local activity, post-dating the Hypanis light toned layered fan deposit. (a) These cones and mounds superpose or overlap impact craters in unit lNh (Figure 8a). (b) They appear to emerge through the Hypanis deposit (unit Cd or Fd) without diverting paleoflow lineations (Figure 8b). (c) They appear to have a "fresh" surface morphology in HiRISE images in both types of mounds observed (Figures 5f and 5g). These two morphologies are: small pitted texture which appears crisp not degraded or filled; and a dark rough texture appearing crisp not degraded. Cones and mounds appear younger than the late Noachian due to stratigraphic relations with unit lNh and wrinkle ridges.
A recent publication (McNeil et al., 2021) mapped many of the same "isolated landform" features we have also identified and at a much wider scale in three zones around Chryse Planitia. McNeil et al. (2021) classified landforms into four classes based on morphology and came to two conclusions which are not supported by our work further examining mounds and cones in the Hypanis deposit region (their Zone 3). One conclusion was that the mounds were layered. Evidence of such layering was not apparent in HiRISE or CTX images of cones and mounds in the Hypanis region. Another conclusion put forth in McNeil et al. (2021) was that an extensive contiguous layer (or layers) at least 500 m thick covered the circum-Chryse region and has been eroded down to create these isolated landforms. While this is possible for plateaus, which are irregularly shaped, have some layering, and likely extended beyond their current truncated extent (Figure 5a), cones and mounds make up a majority of the mapped features and do not contain this layering. Furthermore, cones are round and have central craters. Their setting and morphology appear volcanic and thus the conclusion that these are eroded rounded remnants of a thick plateau unit does not seem to apply, at least in this region. McNeil et al. (2021) suggests that mounds are embayed by the Chryse Planitia plains and are therefore Noachian in age, but the observed stratigraphic relationships in our map area suggest the mounds are younger (Figure 8).
The broader area mapping of McNeil et al. is nonetheless very useful and timely, as there appears to be a connection between isolated landforms near Hypanis and those near the ExoMars Rosalind Franklin rover landing site in Oxia Planum. Thus, our mapping and this prior work suggests a wider depositional and erosional history than at each individual site. This warrants future investigation as similarities between the ExoMars and our mapped region are striking (Früh et al., 2021).
The coexistence of faults, wrinkle ridges, buried crater rims, and volcanic features in the deposit map region suggests that the erupted materials might have used existing faults and fractures as conduits to the surface. Some cones and mounds appear to have experienced further faulting or some degree of internal collapse after formation. For the faulted mounds that formed along wrinkle ridges, this may indicate reactivation of wrinkle ridge tectonism or co-formation around 3.2 Ga. In most cases, however, cones and mounds along wrinkle ridges are not faulted, implying formation after wrinkle ridge activity. Cases like the one shown in Figure 8 superimpose craters in the plains material, demonstrating that the plains do not embay these mounds, but that the mounds erupt through the plains.
We hypothesize that cones and mounds in the deposit region may have been formed by mud volcanism or sedimentary diapirism. This would be consistent with the interpretations of other nearby studies (Farrand et al., 2005;Komatsu et al., 2016;Oehler & Allen, 2010;Skinner & Tanaka, 2007;Tanaka, 1999) and is consistent with expectation for a tectonically compressed sedimentary basin (Skinner & Mazzini, 2009). The Chryse basin experienced repeated burial of water-rich sediments by fluvial activity and regional plains style and explosive volcanism, thus several reasonable mechanisms for eruption of buried material could be proposed. While one could also investigate cones and mounds separately, as though mounds may have formed by a separate non-eruptive process, the only apparent difference noted between the two features is an identifiable central pit. All other aspects including setting, size, and surface morphology are identical. Thus, a common origin is the only explanation supported by our analysis. More work is necessary to properly investigate the sedimentary eruption formation mechanism of these cone and mound features, including remote analysis, numeric modeling, and lab experiments.

Catchment Region
The materials in the Hypanis Valles catchment are diverse in age and origin because the Hypanis Valles and Nanedi Valles systems are extensive. The mapped Noachian units contain reworked impact, volcanic, and fluvial materials as interpreted by many works including the Tanaka et al. (2014) map with which our work generally agrees.
Because the composition of the Hypanis deposit is unclear (due to dusty spectral observations), we broadly searched for compositional clues within the catchment and deposit region. Unfortunately, most of the catchment appears to have significant dust cover, and the largest relatively dust-free areas today occur in the southern distal catchment, a region of surfaces suspected to be younger than the age of Hypanis sedimentation. To search for the best time to examine older dust free surfaces that were more likely to have contributed to Hypanis, we analyzed MARCI albedo videos for Mars years 28-33 (Wellington & Bell, 2020). The video for the Lunae Palus quad did not show a darkening pattern to constrain our search. Thus, we caution that the mineralogy identified in Figure 9 is biased toward locations where CRISM high-resolution images exist, and locations that are less dusty in the distal catchment.
Despite these limitations, several components have been identified including basaltic endmembers and widespread alteration minerals. Scattered throughout the proximal and middle catchment areas, phyllosilicate and plagioclase observations have been reported by multiple sources (Figure 9). This has been shown to be ubiquitous for Noachian-aged highland terrain, and is thus expected (Ehlmann & Edwards, 2014). No clear clumping or spatially consistent patterns are identified at our scale in the proximal and mid-catchment. A unique locality mid-catchment was found to contain Felsic rocks near hydrated silica and kaolinite Pan et al., 2021;Wray et al., 2013). In the distal catchment a basaltic component was inferred from the combination of pyroxene and plagioclase in quantities well above the detection threshold (pink squares in Figure 9). These surfaces were mapped as Ri, Sp, Vms, and possibly Rl (Figure 7). In lower albedo locations, this pyroxene and plagioclase signature appears to correspond to exposures of eNh in the distal catchment. This would suggest, as expected, a basaltic component to the early Noachian highlands materials, through which Hypanis Valles formed. The presence of olivine as well in the distal catchment supports the volcanic origin of a significant component of these materials.
Several types of crater morphologies in the catchment and deposit region were observed, which may support the role of widespread ice in the Hypanis Valles channel network formation process. We noted slurry ejecta and floor filled craters with medium to large diameters, suggesting agreement with a late Noachian to early Hesperian age (Tanaka et al., 2014). These craters with slurry ejecta and and/or with central pits have morphologies indicative of volatiles either on the surface or at depth (e.g., Figure 4, Reiss et al., 2006). Their stratigraphic relations imply that buried ice was closer to the surface at 3.8 Ga, likely when catchment channel systems were forming, and ice depth increased with time (Reiss et al., 2006). These craters were found throughout the catchment (e.g., Figure  2 map in Reiss et al., 2006). The most recent impacts we observed did not contain ice-influenced morphologies perhaps because they were too small to penetrate to a subsurface ice layer. It is unknown if subsurface ice still exists in the catchment.

No Evidence of a Land Dam Ponding a Southern Chryse Lake
The current topography of the Chryse basin is unconfined to the north at the elevations of the Hypanis deposit. Thus, if (a) Hypanis was truly formed subaqueously and if (b) the regional topography was not much different than today, then there must have been a northern ocean. Our mapping partly addressed the latter condition, and found no evidence of an ancient land dam that could have ponded water locally in the late Noachian. Others have suggested the rounded buttes and mesas are a sign of a previously larger unit to the north, but timing for formation of these features is post-Hypanis, and their eruptive origins do not support this claim. Further work could examine the role of ice or glaciers in the formation of Hypanis and determine if an ice dam would be plausible.

Conclusions
We created geomorphic maps of the Hypanis Valles watershed and of the Hypanis deposit region. We defined map units both quantitatively and qualitatively from orbital images, and data sets. Units were designated according to their morphology, albedo, TI, and spectral signature. Hypanis Valles carved through several unique Noachian units older than most terrain surrounding the Chryse basin. Evidence of recent and ancient volcanic activity is present in the catchment and basin. Hypanis Valles was fluvially active throughout the Noachian and formed the terminal deposit in the late Noachian. As proposed in prior works, Hypanis may have formed subaqueously as a delta, and may record a water level drop of about 500 m as a shoreline retreated to the northeast. We identified kilometer-sized cones and mounds which appear to have erupted onto the surface. Characteristics of these features more closely resemble those of outgassing, sedimentary diapirism, and mud volcanism rather than of igneous volcanism. Future work should determine if Mars erupted sediments and volatiles in this region, and by what mechanism. Further study of the deposit map region may be relevant for understanding the current extent of buried reservoirs of volatiles in sedimentary basins on Mars, and study of the Hypanis system may further contribute to our understanding of the total inventory of water on Mars in the Noachian.

Data Availability Statement
Data used in this study including CRISM and TES products are publicly available via the Geoscience Node of the Planetary Data System (https://pds-geosciences.wustl.edu/). Thermal Emission Imaging System data are available at https://pds-imaging.jpl.nasa.gov/volumes/ody.html and https://themis.mars.asu.edu/landingsites/ mslsite_22. Four CTX mosaic tiles of the deposit region can be downloaded from http://murray-lab.caltech.edu/ CTX/ by navigating to the lat/lon of relevance. High Resolution Stereo Camera MOLA blended elevation data is available from the USGS Astrogeology website: https://astrogeology.usgs.gov/search/map/Mars. Topography/ HRSC_MOLA_Blend/Mars_HRSC_MOLA_BlendDEM_Global_200mp_v2. Map shapefiles from this work are available for download in Adler et al. (2021).