The Subsurface Geology of Mars: Remote Sensing of Impact Craters Using Themis, Tes, Moc and Mola

Lunar and Planetary Science XXXVII (2006) 1739.pdf

THE SUBSURFACE GEOLOGY OF MARS: REMOTE SENSING OF IMPACT CRATERS USING THEMIS, TES, MOC AND MOLA. L. L. Tornabene1, J. E. Moersch1, H. Y. McSween Jr.1, V. E. Hamilton2, J. L. Piatek1, K. A. Milam and P. R. Christensen3; 1Department of Earth and Planetary Sciences, University of Tennessee, Knoxville, Tennessee 37996-1410, 2Hawai’i Institute of Geophysics and Planetology, University of Hawai’i, Honolulu, Hawai’i, 96822, USA, 3 Department of Geological Sciences, Arizona State University, Tempe, Arizona 85287–6305, USA.

Introduction: Impact craters provide natural ex- features (e.g. Fig. 1). THEMIS daytime images cover- posures of subsurface composition, stratigraphy and, in ing the crater in question taken prior to 17:00 Local some cases, geologic features that are not otherwise Solar Time (LST) (images within the 15:00-hour are exposed at the surface (e.g., uplift, rifting, tectonics, best) and with average temperatures of 240K [5] were etc.). Remote sensing of craters on the Moon [1], Earth sought as they are best for making good spectral dis- [2] and Mars [3-8] have been used to effectively cap- tinctions. Then, a Decorrelation Stretch (DCS) [15] ture a natural glimpse of the subsurface geology and and Minimum Noise Fraction (MNF) transform [16] provide some additional insight into the differentiation was applied to the THEMIS daytime TIR data to ob- and petrogenetic history of these terrestrial bodies. serve if any spectrally distinct units were present. In These studies indicate that stratigraphic relationships addition, MOLA Digital Elevation Models (DEMs) and compositional changes with depth, as exposed in can be used to determine the spectral unit’s relation- crater morphologic features (i.e., central uplifts, crater ship, if any, to elevation. Finally, high-resolution visi- walls/terraces, and ejecta), may be ascertained from ble imagery (THEMIS, or MOC), if available, and remote sensing data with at least moderate (~100- thermal inertia data were sought to distinguish if these m/pixel) spatial resolutions. spectral units possessed good evidence for being bed- The purpose of this study is two-fold: 1) to use in- rock exposures vs. post-impact infill deposits. frared data from the Thermal Emission Imaging Sys- tem (THEMIS) and the Thermal Emission System

(TES) to map and spectrally identify mineral or lithologic units on Mars that are exposed by craters, and 2) use all available visible-wavelength imagery (e.g., THEMIS visible, Mars Orbiter Camera (MOC), etc.), TES-albedo, thermal inertia data, topographic data (i.e., Mars Orbiter Laser Altimeter [MOLA]) and crater-scaling observations to make interpretations of the geologic structures that are exposed within crater morphologic features. Because craters are also natural traps for transported sediments [9-11], it is vital to differentiate between crater spectral units that specifi- cally arise from crater-exposed lithologies from those that are derived from aeolian deposits, or by other means of post-impact infilling (e.g., lava flows, etc.). Methods: Our first step toward finding craters that expose spectrally distinct materials was to create a Fig. 1. A THEMIS 579-band emissivity image (up is NW) draped on list of craters that are relatively well-preserved (i.e. a THEMIS band-9 brightness temperature image (crater D ~ 18 km). intact ejecta blankets, sharp rims, walls, terraces and Olivine-rich materials correlate 1:1 with both the ejecta and the wall central uplifts). We first used the Martian crater data- rock as a magenta-colored unit [see 6]. base, maintained in Arcinfo-format by [12-14], to con- strain the number of craters in our list to only those Preliminary results: As a result from this survey, we that possessed a central peak feature and an intact have now come to focus on craters in four specific ejecta blanket. Due to temperature constraints and to regions. These regions span the presently recognized avoid dust-covered craters, we then focused on these geologically diverse Martian settings: Isidis Planitia relatively pristine craters from 45ºN to 45ºS within (large impact basin), Thaumasia (tectonic and vol- intermediate- to low-albedo regions. Next, these cra- canic), Meridiani Planum (shallow sea?) and Acidalia ters were then assessed in both visible and thermal Planitia (outflow deposits/degraded lava plains?). Two infrared (TIR) images and using the TES-derived dust or more craters within these regions have been mapped index map. THEMIS nighttime TIR images can be bearing spectrally distinct crater-units using THEMIS used as a proxy for thermal inertia and were used to TIR. These units in turn have 1:1 correlations with ascertain if the craters possessed distinct thermophysi- crater morphologic units. Layered-bedrock can often cal units that correlate 1:1 with crater morphological be observed in high-resolution MOC images of these Lunar and Planetary Science XXXVII (2006) 1739.pdf

craters, which in some cases were subjected to rota- tinct units as mapped by THEMIS TIR. We are cur- tion, folding and faulting [e.g., Fig. 2]. rently examining the THEMIS and TES spectra of these features in order to better understand the geologic nature of this region and these subsurface exposures. Meridiani: Nine craters have been reported within Meridiani with apparent diameters ranging from 2 km to 20 km that have excavated through and depos- ited hematite-poor ejecta on top of the hematite-rich unit Ph [19]. The ~20-km crater has a reduced the TES-derived hematite-signature by superimposing hematite-poor crater ejecta up to a 60 km radius, while the smallest crater in this group has reduced hematite concentrations up to 25 km away (i.e., hematite signature decreased to below TES detectability <10%). These observations and the widespread geo- graphic distribution of these craters suggest that the hematite lag forms only a thin (<0.2 km) surface ve- neer throughout Meridiani, consistent with observations at the surface by the Opportunity rover [3, 20]. This stratigraphic observation also suggests that these Fig. 2. MOC image R0900880 showing the finely-layered materials impacts post-date the onset of hematite deposi- exposed within the central uplift of an unnamed Da=50 km crater in tion/formation. These crater-related observations, Thaumasia Planum. The width of this image is ~3km. along with their sharp morphological features, also suggest that they are amongst the youngest craters of Isidis: Well documented olivine-rich surface de- their size-class within the Meridiani region. posits have been mapped by TES in the southern por- Deconvolution analyses of the ejecta of these cra- tion of the Isidis Basin [17, 18]. However, a close ex- ters may yield important insights into the composition amination of an 18-km diameter impact crater, situated and stratigraphic sequence of the subsurface of Merid- approximately 50 km north of the TES-documented iani and may have important implications for the for- deposits, reveals that this unit is present just below the mation of this region. surface. This strongly suggests that the olivine-rich Acidalia: Eight craters ranging from ~5-30 km in unit underlies Isidis plains materials. Olivine-rich ma- southeast Acidalia have been mapped with spectrally terials are exposed within the ejecta blanket, crater distinct ejecta blankets and/or crater walls/terraces. wall, and slumped terrace blocks adjacent on the crater Preliminary THEMIS results indicate that “fresh” ba- floor. THEMIS VIS (V09409006) and MOC (S06- saltic materials (surface type 1) lie in less then ~ 0.5 01873p) images reveal that the crater walls are rocky, km beneath the surface of this otherwise surface type 2 layered and possess a relatively low albedo (suggest- (ST2) regions. If confirmed by TES deconvolutions, ing little to no dust-cover). Small craters (D <5km) a these observations will have substantial implications appear to excavate to this olivine-rich unit as well as a for the interpretation of ST2 as debated by [21, 22]. 10-km diameter crater ~80 km to the NW of the 18-km References: [1] Tompkins S. and Pieters C. M. (1999), MAPS, crater. The extent of this subsurface olivine-rich unit 34, 25-41. [2] Tornabene L.L., et al. (2005), MAPS, 40, (in press). within the Isidis basin is not observable as TES- [3] Squyres S. W., et al. (2004), Science, 306, 1698-1703. [4] Grant derived dust index (DCI) rapidly increases to the N J.A., et al. (2004), Science, 306, [5] Bandfield J. L., et al. (2004), JGR.-Planets, 109. [6] Rodgers A. D., et al. (2005), JGR-Planets, and NE obscuring even the most pristine appearing 110. [8] Bibring J. -P., et al. (2005), Science, 307, 1576-1581. [9] craters. However, we can conclude from our observa- Arvidson R.E. (1974), Icarus, 21, 12-27. [10] Christensen P. R. tions that the unit exists as a subsurface layer within a (1983) Icarus, 56, 496-518. [11] Thomas P. (1984), Icarus, 57, 205- few 100 meters beneath the outer portion of the basin 227. [12] Roddy D. J., et al. (1998a), LPSC XXIX, #1874. [13] Roddy D. J., et al. (1998b), LPSC XXIX, #1879. [14] Barlow N. G., (i.e., within 60 km of Libya Montes – the Isidis rim). et al. (2000), JGR -Planets, 105, 26,733-26,738. [15]Gillespie A. R., Thaumasia: Oudemans (Da=125 km) and two et al. (1986), Remote Sens. Environ. 20, 209–235. [16] Green A. A., unnamed craters (Da=50 and 60 km) are of consider- et al. (1988), IEEE Trans. Geosc. and Remote Sen. 26, 65–74. [17] able interest within the bounds of the Thaumasia Hamilton V.E., et al. (2003) MAPS, 38, 871-885. [18] Hoefen T. M., et al. (2003), Science, 302, 627-630. [19] Christensen P. R. and Ruff Montes region that bear interesting crater morphologi- S. W. (2004), JGR, 109, doi: 10.1029/2003JE002233. [20] Christen- cal features (i.e., multi-layered strata exposures; see sen P. R. (2004), Science, 306, 1733-1739. [21] Bandfield J. L., et al. example in Fig. 2.) that correlate with spectrally dis- (2000), Science, 287, 1626-1630. [22] Wyatt M. B. and McSween H. Y. (2002), Nature, 417, 263-266.