Potential Rock Glaciers on Mars: Comparison with Terrestrial Analogs
Total Page：16
File Type：pdf, Size：1020Kb
Seventh International Conference on Mars 3353.pdf Potential Rock Glaciers on Mars: Comparison with Terrestrial Analogs. J.L. Piatek1 C. Hardgrove1and J.E. Moersch1, 1Dept. of Earth and Planetary Sciences, University of Tennessee, Knoxville, TN 3799 ([email protected]) Introduction: A number of potential periglacial of chaotic terrain. Dao, Niger, and Harmarkhis all show features have been identified on Mars, including con- evidence for both subsurface and surface flow [10], centric crater fill, lineated valley fill, potential glacial while Reull Vallis appears to have a more complex landforms such as pingos, and lobate debris aprons. history that has erased much of the evidence for sub- [e.g. 1,2]. Concentrations of these debris aprons have surface flow [9]. The most recent deposits (as late as been noted in the southern highlands [3-6]. These fea- Middle Amazonian) include crater fill, gullies along tures extend from local topographic highs (typically channel walls, and debris aprons. massifs), have a distinct lobate form, and have rela- Rock Glaciers. The term ‘rock glacier’ has been tively steep margins. In some cases, these features applied in the terrestrial literature to a variety of geo- display surface lineations approximately parallel to the morphologic features, but these definitions can gener- downslope direction. Previous studies have identified ally be summarized as lobate, steep-margined features a population of these features in the Reull Vallis region. consisting of angular debris, and are often associated [5] examined 54 features in this area using Viking and with mountainous terrain and exposed ice. Surface Mars Global Surveyor datasets, and concluded that the morphologies typically consist of flow features like morphologies appeared similar to terrestrial rock gla- ridges and furrows [11-15]. The debris layers of sev- ciers. [7,8] extended this earlier work using nighttime eral terrestrial rock glaciers can be broken up into two infrared images taken from the Thermal Emission horizons: an upper coarse-grained layer, and a lower Mapping Spectrometer (THEMIS) aboard Mars Odys- finer grained level. The upper debris layer consists of sey. These studies noted variations in the thermophysi- blocks a few decimeters in size, while the lower layer cal properties of these features that could be related to consists of sand and silt [16]. the formation mechanism. These varations are typi- Multiple formation mechanisms for these features cally lineations parallel to the downslope direction, have been proposed, and are directly related to the pro- similar to features observed in visible albedo images, portion of ice and rock. These are summarized by [17] but also include curvilinear features that occur either in as a permafrost model, consisting of an intimate mix- the lobate toe of the apron, or where the apron material ture of rock and ice; a glacier ice core model (or debris- appears to flow around obstacles. These sets of linea- covered glacier), which is a layer of ice covered by a tions are suggestive of formation by fluid or plastic thick layer of debris; and a catastrophic landslide flow, and the steep nature of the toes of these aprons is model, which consists only of rock debris with water or similar to the morphology of terrestrial rock glaciers, ice possibly acting to trigger the mass movement. debris-covered glaciers, and protalus lobes. Method: THEMIS. Thermophysical properties of The study presented here is a continuation of the these features are examined through estimation of the work of [7,8]. The results of this earlier work are com- thermal inertia. This estimation process requires in- pared to apparent thermal inertia (ATI) images of ter- formation about the surface kinetic temperature, eleva- restrial rock glaciers, as well as to morphologic char- tion, and albedo: in this case, the temperature is de- acteristics of these features. rived from THEMIS nighttime infrared images, eleva- Background: Regional Geology. The Reull Vallis tion from the Mars Orbiter Laser Altimeter (MOLA) region is located east of the Hellas basin, from gridded dataset, and albedo from Thermal Emission 27.5°-47.5°S and 90°-115°E, and the geologic history Spectrometer (TES) global bolometric albedo map. has been previously discussed in detail [e.g. 9]. The THEMIS images are identified via JMARS oldest materials are Noachian-aged highland terrains (http://jmars.asu.edu), and are initially processed via which have been extensively modified by later events, the THMPROC website (http://thmproc.mars.asu.edu) including impact cratering, mass wasting, fluvial, and using the undrift/dewobble, unrectify, deplaid, and aeolian processes. Hesperian-aged volcanic units rectify options. The resulting radiance images are likely sourced from nearby Hadriaca and Tyrrhena processed to separate temperature and emissivity via Paterae are interbedded with sedimentary units. There the normalized emissivity method [18,19]. Orbital are two primary pieces of evidence for the presence of information from the THEMIS nighttime image is used subsurface volatiles during the Hesperian: pitted to determine local solar time and solar longitude at the plains, which possibly formed due to creep of subsur- time of image acquisition. These values are given as face ice, and the formation of outflow channels. Reull, inputs to a lookup table routine that uses results from Dao, Niger, and Harmarkhis Valles all appear to have the model of [20] to estimate values of surface thermal formed during the Hesperian, and all source from areas inertia. Seventh International Conference on Mars 3353.pdf One difficulty in the determination of surface ther- in temperature (!T) of the pixel from the coldest to the mal inertia is the estimation of atmospheric dust opac- warmest times of the day. Pairs of day-night ASTER ity at the time the image was acquired. Typically, the images with similar collection dates to minimize sea- infrared dust opacity is assumed to be 0.2 for thermal sonal variations in temperature. Data was also filtered inertia calculations. However, it is often the case that for minimal cloud cover. Albedo was determined by overlapping THEMIS images have significant differ- calculating a weighted average of the reflectance in the ences in surface temperature that are larger than ex- each of the three ASTER visible wavelength bands pected based on seasonal variations. These variations multiplied by the amount of incident solar radiation at affect the entire image, and are not limited to specific those wavelengths. ASTER derived surface kinetic geologic units within the scene. In this case, it is pos- temperatures were used for the day-night temperature sible that atmospheric dust opacity has altered the maxima and minima respectively, however, tempera- cooling rate of the surface, as the dust essentially acts ture data is limited to times when the EOS-Terra satel- like a blanket to retard surface cooling. In order to lite passes over the area of interest, which for the account for some of these variations, values of dust Alaska Range is approximately 11:30 local time for opacity were estimated using surface emissivity values day images and 21:00 UTC for night images. It should calculated during temperature separation. Because the also be noted that ATI images, as well as Differential surface temperatures are quite low, the majority of this ATI images [25], are sensitive to variations in slope derived emissivity ought to be due to absorption in the angle and orientation. [26] have shown that for sur- atmosphere. Use of these derived opacity values sig- faces of homogeneous composition, particle size, and nificantly reduced apparent differences in thermal iner- induration, changes in slope angle and orientation will tia in overlapping images, but did not completely alle- affect the duration of diurnal heating that surface expe- viate this issue. riences. These changes will appear as variations in the Thermal inertia is a physical property of the surface ATI value for surfaces of homogeneous particle size. A that is related to the material composition, particle size, correction to the ATI model using ASTER derived and packing state/induration. In this particular study, it digital elevation maps is being developed [26], similar is reasonable to assume that the debris apron and the to what is shown here for Mars. Correlating the non- source massif are likely similar in composition. If the corrected ATI to the ASTER-derived DEM should help rock glacier hypothesis is correct, then it is also likely determine if the variation in ATI is related to changes that that debris at the surface is not cemented or indu- in particle size or simply changes in slope. rated, and any variation in packing is likely due to Additional Datasets. In order to determine if varia- variations in particle size. For these reasons, variations tions seen in thermophysical images of Martian debris in thermal inertia will be directly correlated with po- aprons are correlated with topography. Individual Mars tential variations in particle size. Orbiter Laser Altimeter (MOLA) data records were Terrestrial Analogs. Comparison of Martian fea- used to create small scale digital elevation models by tures with Earth analogs is performed using an analog interpolation between data points. These datasets were for THEMIS: the Advanced Spaceborne Thermal used to look for small scale elevation changes within Emission and Reflectance Radiometer (ASTER) aprons, and to identify slope changes possibly related aboard the Terra satellite [21]. ASTER collects data in to emplacement mechanisms. Visible images such as both the visible/near-infrared (VNIR) and thermal in- those from the High Resolution Stereo Camera (HRSC) frared (TIR) wavelengths, with spatial resolutions that are used to examine apron surface morphology are similar to THEMIS (15 m ASTER VNIR vs. 18 m Results: THEMIS thermal inertia images for ex- THEMIS VNIR; 90 m ASTER TIR vs. 100 m THEMIS ample debris aprons in the Reull Vallis region are TIR) [21,22].