Glaciation in Mid-Latitude Terra Sirenum Craters of Mars

GLACIATION IN MID-LATITUDE TERRA SIRENUM CRATERS OF MARS

John G. Apostolou May 06, 2020

I. INTRODUCTION II. BACKGROUND

The evolution of the Martian surface, According to global circulation models, especially as pertaining to the presence of large scale transportation of glacial ice has water, is a fundamental consideration for occurred from polar regions equatorially understanding the history and current as driven by oscillatory behavior in the state of the red planet. Modern efforts obliquity of Mars (Mischna et al., 2003). have helped to develop a tentative past of This relationship is exists such that periods hydrologic activity and possibly even hab- of high obliquity correspond to extensions itability on Mars. In the present era of in- of ice-rich mantle from the poles, repre- creasing interest in the deeper exploration senting a Martian ice age, whereas peri- and potential colonization of the planet, ods of low obliquity align with a recession knowledge of the varied surface proves vi- of this ice to the poles (Head et al., 2003). tal. Based on existing observations of The orbital trends of Mars beyond ~20 Myr Mars, it is known that the presence of wa- into the past are relatively uncertain. How- ter on the surface has been influential in ever, recent estimates place the present the past as both liquid and ice through a obliquity of Mars at ~25.2°, which is con- variety of geologic processes. This has sidered low in relation to the trends of the yielded regions of considerable complexity last 20 Myr (Laskar et al., 2004). It is then displaying diverse features such as chan- reasonable to assume that the current neling and carving intermingled with ejecta state of the Martian surface is interglacial and impact cratering, such as within the with smaller scale flow features and rem- southern mid-latitudes. In deeper context, nants. evidence of glacial activity has been rec- Of these features, a considerable vari- ognized as especially prominent in the ety have been observed at the mid-latitude mid-latitude bands where it may be recog- bands as features of the craters present nized as a significant contributor to crater there. Lobate debris aprons (LDA) are de- degradation (Berman et al., 2008). In gen- scribed as protruding flows displaying pat- eral, a survey of craters within a given lati- terns of down-gradient flow, bending tude band, featuring examination of lobate around topographic obstacles, and broad- flow fronts, gullies, and carving features, ening when unrestricted by the present to- may provide insight into the latitudinal and pography (Baker et al., 2009). The for- regional distribution of particular glacial mation of these flows is not certain, lead- features as well as their extent. ing to an array of proposals regarding the

methods of development, such as frost stemming from an alcoved head and end- creep, rock glacier flow, and debris-cov- ing with a fanned apron at the base of the ered glacier flow. While many of these fea- slope. A related classification of flow features are analogous to terrestrial glacial ture is the moraine-like ridge (MLR), which features, there is disagreement over clas- are accumulations of glacial debris of up to sifications for such terrestrial counterparts, 100 m in height. Systems of these ridges providing a further obstacle to identifying are sometimes seen near the fan-like out- Martian flows (Baker et al., 2009). flow of the gulley bottoms. The majority of Often associated with lobate flows is these ridges are found within craters of the evidence of sublimation, primarily textur- southern mid-latitudes. Beyond this, the ing. Sublimation, in general, is the phase most prominent, developed of these MLR transformation from solid directly to vapor features are considered to exist at the without transitioning to the liquid state. Es- base of pole-facing crater walls (Arfstrom pecially in regard to glacial ice, the subli- and Hartmann, 2004). mation is generally rate-controlled by va- There exist a variety of other features por transport from the outer surface of interest when examining Martian cra- (Douglas and Mellon, 2019). This process ters, although not necessarily tied to gla- can result in heavily textured glacial sur- cial activity. Channeling can be found in faces with patterns of varying complexity. many regions of the Martian surface but Larger scale trends in the pattern of subli- likewise can have a presence within a mation texturing may also provide insight crater or on its rim, generally indicating a into the direction of an associated lobate previous or current fluid flow. Similarly, flow. Alternatively, dissected mantle ter- craters may contain evidence of alluvial rains may be the result of subsurface sub- fanning as a result of the presence of such limation. Regardless, the combination of fluid, namely liquid water. Beyond hydro- these processes, ice flow, and differential logical and glacial activity remains the pos- sublimation can be responsible for more sibility of dunes and dune-like features complex and potentially curvilinear valley along a crater floor. It is suspected that terrains (Douglas and Mellon, 2019). It is such dunes would be more prevalent in therefore reasonable that such texturing larger craters where wind flow may be less can serve as evidence supporting a pres- obstructed. ence or history of glacial ice on the Martian surface. III. METHODOLOGY Gullies are features that tend to be common in the mid and upper latitudes of In attempting to examine and poten- Mars characterized as small, narrow inci- tially classify craters exhibiting glacial ac- sions resembling terrestrial features of the tivity or history, a region of study was first same name. These Martian features are selected for use with ArcGIS, ArcMap. typically found on steeper slopes, such as Given the significance of preserved glacial the inner gradients of sharp crater rims, evidence at the Martian mid-latitudes, the

band of interest was chosen to be the by the northern -40.000° and southern range from -40.000° to -50.000°, corre- -50.000° boundaries. sponding to a portion of the southern mid- Beyond identifying information, the latitudes. Starting from the -180.000° lon- characteristics observed in the craters gitudinal line, the intent was to examine were broken into two additional primary the selected latitude band from the starting categories, titled ‘floor’ and ‘rim’ in refer- point to as far east as possible. Given cir- ence to the corresponding component of cumstantial constraints, the range of ex- each crater. The first recorded character- amination extended only to approximately istic of the ‘floor’ data was the type, classi- -126.700° longitude. This corresponds to fied as either flat or complex. Although ra- the Terra Sirenum region of Mars. ther intuitive, a flat rating corresponded to The data taken was organized into a relatively smooth inner surface of the three primary categories and further bro- crater, disregarding the presence of a cen- ken into several subcategories. Regarding tral peak. In contrast, the complex classifi- the selection of craters for examination, it cation was used to indicate significant was determined to consider craters with a complexity, such as a concentric crater or minimum diameter of 10.00 km, thus incor- substantial ridging and peak remnants. porating the larger, more prominent cra- Following this, the presence of a defined ters of the region. Under this criterion, for central peak was recorded along with an every crater examined, identifying infor- alternative distinction to indicate whether mation was first recorded. This consisted only remnants were visible. of the name, corresponding geographic The next floor subcategory, titled ‘Gla- coordinates, and diameter as provided by cial Carving / Flow Front,’ was the first di- the Robbins Crater Database, making up rectly related to glacial flow. The data un- the ‘identification’ primary category of the der this column was meant to indicate the project data. presence and relative location of a lobate The actual structural and characteristic flow (LDA), linear flow front, significant gla- examination of the craters was conducted cial carving, or a related feature. Figure 1 by first identifying a small group of craters provides examples of each of these fea- at the 1:1,250,000 scale and examining ture classifications as observed in the re- each crater within the group through pri- spective craters. Each entry provided a di- marily Context (CTX) imaging. Once all rection with respect to the crater center of craters in a group had been examined, fo- where such a feature was found, if appli- cus was shifted either north or south until cable. The following subcategory was then a limit of the latitude band was reached. used similarly to recognize the presence Once this occurred, the process was re- and locations of apparent sublimation tex- peated for the Martian surface to the east. turing. Although this often corresponded to Essentially, the immediate area of exami- the data entries for glacial carving and flow nation passed eastward using something fronts, there were many cases where tex- of a boxed serpentine course, constrained turing was visible without a defined flow

a b

c d

Figure 1: (a) Annotated map displaying converging lobate flows (LDA) within crater 24-001016. (b) An- notated map displaying a more linear flow front to the north and northwest of the central peak in crater 24-000159. (c) and (d) Annotated maps exhibiting areas of glacial carving at different points (southwest and northeast, respectively) on the rim of crater 24-000413. front or carving was apparent without ac- subcategories regarding glacial flow and companying sublimation. Figure 2 pro- sublimation texturing, in cases where the vides a reasonable example of the former. feature was present over nearly all of the Furthermore, attempts were made to dis- crater floor, the term ‘areal’ was used. One regard texturing that was clearly the result such case is provided in Figure 3. In con- of ejecta from a nearby, younger crater. In trast, if the feature was absent, the label cases where it was difficult to make this ‘none’ was assigned. Lastly, ‘circumferen- distinction or there was the possibility of tial’ was used to describe cases in which various different texturing sources pre- the feature was present across the crater sent, a note was made suggesting the tex- floor, excluding the center. turing was possibly or likely ejecta, de- This directional recording scheme was pending on the approximate level of obser- likewise used for the remaining floor sub- vational confidence. For both of these categories. The first of these was dedicat-

Figure 2: Alternatively scaled, contrast-enhanced images of crater 24-000405 displaying apparent sublimation texturing without a defined flow front verifying glacial activity or nearby cratering to indicate ejecta.

ed to the presence of channeling on the the approximate consistency of the crater crater floor. Subsequent was a column for rim. Factors contributing to a ‘varied’ clas- the identification of any sort of alluvial fan- sification rather than ‘uniform’ are noticea- ning features, if present. Additionally, if bly asymmetric weathering, disruption of dunes were located on the crater floor, this the rim by another crater, or drastically var- was recorded in a separate subcategory ying rim width. using relative directions. Figure 4 shows an example of dunes observed. The last component of the ‘floor’ primary category was a section for noting any abnormal or interesting features of the floor or further explaining prior classifications. Following the ‘identification’ and ‘floor’ primary categories was a similar grouping for features defining the crater rim. The first ‘rim’ subcategory was a general classification of the rim erosion, based on ob- servation. The rim was given a ‘rounded’ rating to describe substantial weathering Figure 3: Example of areal glacial flow on the floor of its structure. Alternatively, a ‘sharp’ clas- of crater 24-001333. sification was used to represent craters with stronger, more defined rims suggest- Further ‘rim’ subcategories were orga- ing a younger age. This data was recorded nized similarly to those of the crater floor. in conjunction with a ‘relative uniformity’ The only difference was that rather than entry which was meant to communicate considering the presence of dunes, an

However, in some cases, regions of a crater were obscured by surface shadows or the Context (CTX) mosaic was incom- plete. In these scenarios, the identification of surface features relied on the lower res- olution THEMIS IR imaging. In these scenarios, a note was made with the corresponding data indicating presumed entries based on the THEMIS displays. In some cases, as with the images given in this paper, the crater maps were examined with an altered contrast in order to try and given Figure 4: Dunes located on the eastern floor of crater 24-000182. surface features more definition relative to the rest of the image. examination into the presence of gullies was included. An exhibit of IV. RESULTS these gullies is shown in Figure 5a along with a defined set of MLR in Figure 5b. The observed data, although originally In order to attempt to discern which fea- recorded in the Excel spreadsheet format, tures were being observed in a crater, the was reformatted to CSV for use in a Py- scale was adjusted as necessary. In gen- thon environment. Using Python, a simple eral, the CTX imaging provided the degree code was written in order to interpret the of detail needed to discern surface fea- CSV data, primarily by counting the num- tures for the purposes of this examination. ber of craters exhibiting a particular trait.

a b

Figure 5: (a) Series of gullies on the southwestern rim of crater 24-001138, complete with fanned aprons at the base. (b) Set of MLR (moraine-like ridges) along the northeastern rim of crater 24- 001053.

In total, 400 craters were examined. Of Relative Location Number of Craters these, 333 craters possessed rounded or Areal 108 generally worn rims whereas the other 67 Center 35 were considered to have generally sharp North 124 rims. Only 21 of the craters studied exhib- South 93 ited a defined central peak or remnants of East 127 one. West 135 Of the 400 craters examined in the Northeast 148 scope of this survey, 324 exhibited evi- Northwest 150 Southeast 131 dence of glacial flows in some form, spe- Southwest 123 cifically lobate flows, glacial carving, and flow fronts. The relative directional distri- Table 2: Table where the first column provides the bution of these flows is given in Table 1 directional location of observed texturing, assumed to be due to sublimation, whereas the second col- and alternatively in Figure 6a. umn gives the number of craters exhibiting the feature in the corresponding location.

Relative Location Number of Craters ing corresponded to the presence of a de- Areal 88 fined flow front, lobe, or carving. The max- Center 37 imum difference between the two sets of North 123 values is 20 craters whereas the mean dif- South 92 ference in count is only 6.22 craters. East 135 West 146 Regarding other recorded features of Northeast 156 the craters, 50 possessed a significant Northwest 154 channel or channeling on the crater floor. Southeast 133 However, 107 craters displayed similar Southwest 130 features at some point on the rim. Interest-

Table 1: Table where the first column provides the ingly, alluvial fans were not found in any of directional location of observed lobate flows, glacial the 400 craters, whether on the floor or carving, or flow fronts whereas the second column along the rim. It is possible that this is the gives the number of craters exhibiting such features in the corresponding location. result of an oversight. Otherwise, this indi- cates a relative rarity of this feature. Dunes Beyond evidence of explicit lobate flows, were found on the floor of a mere 9 craters. glacial carving, or other flow fronts, the Gullies were distributed somewhat presence of potential sublimation texturing more extensively, however. Of the 400 was also examined. Of the 400 craters craters examined, 96 craters, approxi- studied, 313 exhibited texturing that was mately one fourth, held gullies at some interpreted to be a result of sublimation. point on the crater rim. The directional dis- The relative directional distribution of this tribution of these gullies is given in Table 3 texturing is given in Table 2 and displayed and alternatively in Figure 7. alternatively in Figure 6b. In general, the presence of observed sublimation textur-

a b

Figure 6: (a) Bar representation of the number of craters exhibiting a lobate flow, flow front, or glacial carving in each location with respect to the crater center. (b) Bar representation of the number of craters exhibiting sublimation texturing in each location with respect to the crater center.

production of distributions was done for Relative Location Number of Craters sublimation texturing and is shown in Fig- North 5 ure 8b. South 19 East 14 West 26 Northeast 20 Northwest 12 Southeast 9 Southwest 59

Table 3: Table where the first column provides the directional location of observed gullies whereas the second column gives the number of craters exhibiting the feature in the corresponding location.

An additional relationship considered how Figure 7: Bar representation of the number of cra- the frequency of glacial activity changed ters with observed gullies in each location with re- with latitude as correlated to the observa- spect to the crater center. tions made. In order to do this, the latitudi- Interestingly, the latitudinal distribution of nal distribution of craters exhibiting lobate craters with the respective feature in the flows, flow fronts, or glacial carving was northern half of the crater appears to exist examined as well as that of craters exhib- more northward whereas the distribution iting these features in the northern half of of craters exhibiting the feature in the the crater and those exhibiting the behav- southern half exists more southward. In or- ior in the southern half. These distributions der to verify that the differences displayed are shown in Figure 8a. A similar in Figure 8 are statistically significant, a

Figure 8: (a) Box plot describing the latitude-based distribution of craters with glacial activity and again when broken into northern activity and southern activity. (b) Box plot describing the latitude- based distribution of craters with sublimation texturing and again when broken into northern activity and southern activity.

two-sample t-test was conducted for both V. DISCUSSION glacial activity and sublimation texturing. The former yielded a p-value of approxi- To follow the previously introduced sta- mately 4.942 ´ 10-5 and the latter a p-value tistical analysis, both resulting p-values of approximately 6.456 ´ 10-7. were well below the commonly accepted critical value of 0.05, it can be concluded that the difference found in the distribution

of craters with glacial and texturing fea- equator-facing orientations rather than tures in the northern half from those with pole-facing. While recognizing a slightly the same features in the southern hemi- different range of examined latitudes, the sphere is statistically significant. This sug- study conducted in this paper yielded data gests that there is indeed a relationship that supports the findings of Berman et al. between the orientation of glacial flows (2008). To specify, of the 400 craters ex- and sublimation texturing and latitude. amined, a combined total of 281 displayed With this possibility considered, it is im- lobate flows, flow fronts, or carving in the portant to recognize liabilities to this find- east-west direction as opposed to 215 with ing. One potential source of error involves a north-south orientation. This likewise the differentiation of the relative direction. holds true for sublimation texturing and For example, what orientation constitutes gully distribution. A combined 262 craters an eastern data entry versus a northeast- displayed the texturing in equatorial direc- ern data entry may vary between observ- tions versus a combined 217 in the polar. ers. This is significant given that a north- A combined 40 craters had gullies on the eastern data entry would be considered in eastern or western rim in contrast to a the analysis of Figure 8 whereas an east- combined 24 craters displaying gullies on ern indication would not. the northern or southern rim. According to the work done by Berman As described in Pearce et al. (2010), lo- et al. (2008), craters were examined in two bate flows examined in Utopia Planitia mid-latitude study regions, one of which possess core characteristic similarities to was located in the southern hemisphere of terrestrial flows. These similarities are Mars. Lobate flows, channels, and addi- summarized as a common, lobate form, tional glacial features were found to be curvilinear ridges on the surface, and common in craters greater than 20 km in frontal ridges resembling terminal and re- size. The validity of this finding can be con- cessionary moraines seen on Earth. As sidered in terms of the examination con- these features can be recognized in the ducted in this paper. Of the 400 craters ex- craters displayed in Figures 1a and 3, the amined, 174 possessed diameters greater study conducted in this paper generally than 20.00 km. Of these, 145 displayed ev- agrees with this similarity. idence of glacial activity in the form of a lobate flow, flow front, or carving, represent- VI. CONCLUSION ing a frequency of 83.33%. Based on this value, there is reason to believe that the The conclusions reached from this finding of Berman et al. (2008) is sup- study can be summarized as follows. It is ported. found that the distribution of craters with An additional finding by the aforemen- northern glacial activity and the distribution tioned authors suggests that within the lat- of craters with southern glacial activity itude band ranging from ~45° to 60°, lo- have some latitudinal dependence. The bate flows and gullies tend to have characteristic high frequency of lobate

flows and additional glacial features in cra- would be certain and consistent across ters larger than 20 km in diameter is sup- craters, and there would be no dispute re- ported. In addition, the relative orientation garding whether a feature should be listed of lobate flows, sublimation texturing, and as a cardinal direction or intercardinal. gullies in the eastern and western direc- One other evolution of this study could tions is more common than in the northern involve introducing greater distinction to and southern directions, agreeing with the features examined in each crater. For other studies. Lastly, the observations example, rather than recognizing lobate made agree with a structural similarity be- flows, more linear flow fronts, and glacial tween Martian and terrestrial lobate flows. carving as a single subcategory of the floor Moving forward, the methodology driv- and rim, individual data entries could be ing this study could be advanced by first made for each feature. Furthermore, addi- and foremost examining an overall larger tional subcategories could be included for sample of craters falling within the afore- moraine-like ridges, lineated valley fill, and mentioned -40.000° to -50.000° latitude several other distinct characteristics. This band. This would most notably improve the would allow a more in-depth consideration representation of craters in the southern of the craters and add the potential for the mid-latitudes of Mars. Likewise, there is discovery of further relationships. the supplemental possibility that relationships with latitude would become more de- VII. ACKNOWLEDGEMENTS fined or more significant with a larger sample. As previously mentioned, a larger I gratefully acknowledge Alan D. How- sample was aligned with the original intent ard, who served as an advisor and contrib- of this study, but circumstantial constraints utor to this study, for his guidance, instruc- resulted in the limitation to 400 craters of tion in the use of ArcGIS, and introduction the Terra Sirenum region. to the glacial trends present on the Martian Additionally, a more systematic ap- surface. I also extend gratitude to the Uni- proach to defining the relative orientations versity of Virginia for providing access to with respect to the crater center could be the ArcGIS software used in the examina- implemented to provide a more definite tion of Mars. distribution of glacial features. One possibility involves dividing the circumference of a crater in order to establish angular ranges that would properly categorize features by direction. This range would then act as a degree of tolerance to the 45° counterclockwise from the equator corresponding to northeast, for example. In this manner, the definition of the directions

VIII. BIBLIOGRAPHY

Arfstrom, J., Hartmann, W.K., 2004. Martian flow features, moraine-like ridges, and gullies: Terrestrial analogs and interrelationships

Baker, D.M.H., Head, J.W., Marchant, D.R., 2009. Flow patterns of lobate debris aprons and lineated valley fill north of Ismeniae Fossae, Mars: Evidence for extensive mid-latitude glaciation in the Late Amazonian

Berman, D.C., Crown, D.A., Bleamaster III, L.F., 2008. Degradation of mid-latitude craters on Mars

Dickson, J.L., Head, J.W., Fassett, C.I., 2012. Patterns of accumulation and flow of ice in the mid-latitudes of Mars during the Amazonian

Douglas, T.A., Mellon, M.T., 2019. Sublimation of terrestrial permafrost and the implica- tions for ice-loss processes on Mars

Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003. Re- cent ice ages on Mars

Laskar, J., Correia, A.C.M., Gastineau, M., Joutel, F., Levrard, B., Robutel, P., 2004. Longterm evolution and chaotic diffusion of the insolation quantities of Mars

Mischna, M.A., Richardson, M.I., Wilson, R.J., McCleese, D.J., 2003. On the orbital forcing of Martian water and CO2cycles: A general circulation model study with simpli- fied volatile schemes

Pearce, G., Osinski, G.R., Soare, R.J., 2010. Intra-crater glacial processes in central Utopia Planitia, Mars