50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 2325.pdf

CRYPTOMARE MINEROLOGY AND LAVA LAKE PROCESSES WITHIN THE GASSENDI CRATER. T.A. Giguere1, J.J. Gillis-Davis1, and M. Lemelin2. 1Hawaii Institute of Geophysics and Planetology, Univ. of Ha- waii, Honolulu, HI 96822, 2Department of Earth and Space Science and Engineering, York University, Toronto, CAN M3J 1P3 (giguere@hawaii.edu).

Introduction: The Gassendi region, on the west- was used [9]. The high resolution (0.5 m/pixel) NAC ern nearside of the Moon just north of Mare Humorum, images were critical for the study of the smallest vol- exhibits a variety of volcanic features. Recent studies canic features. Topographic data were provided by the of these features reveal the region’s complex volcanic LROC GLD100 [10] DTM and NAC DTMs. history. For instance, discovery of new and confirma- tion of pre-existing cryptomare [1], pyroclastic depos- its [2,3], mare deposits [1], lava lakes [1,4], irregular depressions, moat-like wall boundaries, possible vol- canic constructs on the floor of Gassendi [5,6] and a cryptomare west of Gassendi [7].

Figure 2. Multiple scarps on the southeast side of the NE lava lake in Gassendi crater (NAC DTM, 3 m/pix, stereo pairs M1213319041, M1213326073. Scarp elevations (black arrows), mean floor elevation (gray arrow), and inset NAC image (M1123846148R, Inc 72.4, Res 1.8 m/pix). Image data from the SELENE “Kaguya” mono- chromatic Terrain Camera (TC) [11] and the Multi- Figure 1. Kaguya MI FeO image of the Gassendi crater Band Imager (MI) [12] visible and near-infrared multi- floor (62 m/pix). Clementine spectral and Kaguya minerol- spectral camera were used for detailed surface and ge- ogy data were taken at locations A - E. SW, NW, and NE ochemical analyses. Maps of mineral and glass distri- lava lakes (blue arrows). White box is the area for Figure 2. bution were produced with MI spectral data at a resolu- This study focuses on constructs and compositions tion of 62 m/pixel. The abundance of olivine, orthopy- of volcanic features on the floor of Gassendi crater roxene, clinopyroxene, plagioclase, and pyroclastic (110 km, 17.55°S 39.96°W). Gassendi is a floor- glass is determined for each map pixel. These miner- fractured crater [6] that exhibits evidence of crypto- al/glass maps were produced by applying Hapke’s ra- mare and lava lakes. The goals of this study include: diative transfer equations to the MI data [13]. The 1) Determine the minerology of the cryptomare depos- physical and compositional parameters selected for the its within Gassendi, 2) Identify whether the cryptomare modeling are as follows: regolith grain size of 17um, was created by a mare basalt flow or pyroclastic erup- glass grain size of 60um, and Mg number of 65. Justi- tion, 3) Identify multiple scarp locations, 4) Determine fication of these regolith parameters is given in [13]. the processes that created multiple scarps in each lava Cryptomare Minerology: The cryptomare was lake. Studying the array of volcanic features within mapped [1] and basaltic composition was confirmed Gassendi region/crater can reveal information about with Clementine five-point spectra [1]. We revisit this styles of volcanic emplacement that may have occurred result using Kaguya MI imager data (Figure 1), which elsewhere on the Moon but are now buried by exten- has higher spatial resolution and the reflectance has sive volcanic activity. been corrected for the shading effects of topography Methods: Lunar Reconnaissance Orbiter Camera which results in lower mineral estimation errors for (LROC) Wide Angle Camera (WAC) and Narrow An- sloped terrain [11, 12]. gle Camera (NAC) images were used in this investiga- tion [8]. Both high and low incidence WAC imagery 50th Lunar and Planetary Science Conference 2019 (LPI Contrib. No. 2132) 2325.pdf

Clementine – Legacy Results. Two cryptomare de- within the conduit and, potentially, exsolution of bub- posits on the southwest and northwest floor of Gassen- bles within the magma reservoir [18], which relates to di were identified by five dark halo craters (craters A-E the state of the mantle undergoing partial melting. in Figure 1) with FeO-rich (14.3 to 14.8 wt.%) ejecta. The FeO abundances of this cryptomare are in the range (14-18 wt.%) of the surface mare deposits mapped [14] in the S and SE portion of the crater. We confirmed the clinopyroxene minerology of the deposit with Clementine five-point spectra [1], which exhibit strong “1μm” bands centered at or longward of 0.95μm [1]. Clinopyroxene, elevated FeO contents, and low albedo are diagnostic features of mare basalt [15]. Kaguya – New Results. We determined the abun- dance of clinopyroxene, orthopyroxene, plagioclase, and pyroclastic glass for each of the locations shown in Figure 1. These points are the same locations used for the Clementine spectral and geochemical determina- tions. Our Kaguya mineral maps show the mean abun- Figure 3. Kilauea Iki lava lake levels, Hawaii. Multiple lev- dances for the five locations are: clinopyroxene els of the lava lake (a–d) are preserved in this view looking west. Wall is on the N side of the lava lake, opposite the vent. (~20%), orthopyroxene (~3%), olivine (<1%) and pla- gioclase (~28%). The large percentage of clinopyrox- Multi-layer Scarps in Gassendi. All three lava ene may confirm the Clementine result that the buried lakes within Gassendi crater have bounding lava scarps material is mare basalt. However, the model indicated as part of the identifying criteria [1]. In some locations a high percentage of glass was present. Future model we see multi-layered scarps [4, Figure 2]. On the runs are required to distinguish agglutinitic and pyro- boundary of the Northeast lava lake, we see three lev- clastic glass. There is a measurable amount of high- els of scarps above the floor (Table 1). The NE lake lands contamination in the cryptomare, as is to be ex- had a sustained active period, which created the wide pected, evidenced by the increased amounts of plagio- main scarp [19]. Followed by two cycles where the clase and orthopyroxene relative to typical mare basalt. lava drained, created scarp 2, then drained again to Lava Lake Processes: Lava lake morphology var- create the smaller scarp 1. Next the lava completely ies between the SW, NW, and NE lava lakes (Figure drained leaving the floor exposed. The other plausible 1), which suggests that the eruption process also var- scenario involves the active filling of the lake, which ied. We examine the morphology to determine the created scarp 1, followed by a second fill episode to process that created each lava lake. Lava lakes are create scarp 2, then the final sustained fill episode cre- classified as two types: active and inactive [16,17]. ates the main scarp. Scarps 1 and 2 would be thermally Table 1. Multi-layered scarp on the Gassendi NE lava lake. eroded as the main scarp is created. Kilauea Iki pro- Scarp Elevation (m) Height (m) Width (m) vides a terrestrial example of multiple layers of lava Floor -2232 0 NA lake scarps (Figure 3). In contrast to the wide Gassen- di scarps, these scarps are narrow in width, due to the 1 -2194 38 70 short duration of the overall eruption (37 days) [19]. 2 -2175 19 108 References: [1] Hawke B. et al. (2015) LPSC 46th, #1310. [2] Main -2158 17 100’s m th Active lava lakes represent the exposed upper surface Giguere T. et al. (2007) LPSC 38 , #1132. [3] Giguere T. et al. th th of an active magma column. Inactive lakes are rootless (2017) LPSC 48 , #2760. [4] Giguere T. et al. (2016) LPSC 47 , and stagnant and do not form directly on top of the #1884. [5] Schultz P. (1976) Moon Morphology, 626. [6] Schultz P. magma column. Our focus is on active lava lakes, (1976) Moon, 15, 241. [7] Hawke B. et al. (1993) GRL, 20, 419. [8] which are classified as sustained or cyclical. Identify- Robinson, M. et al., (2010) Spac Sci. Rvw 150, 81. [9] Speyerer, E. ing lava like type is significant because it ascribes et al. (2011) LPSC XLII, #2387. [10] Scholten F. et al. (2012) JGR, whether the lake was in steady-state circulation with 117, 12 pp. [11] Haruyama et al. (2008) Earth Plan. Space, 60, magma column, or periodically filling and then empty- 243–256. [12] Ohtake, M. et al. (2008) Earth Plan. Space, 60, 257- ing as the magma column rose and fell. The difference 264. [13] Lemelin, M. et al. (2015) JGR, 120, 869-887. [14] Titley in behavior between these two lava lake types is influ- S. (1967) U.S.G.S. Map I-495. [15] Antonenko, I. et al. (1995) enced by the interactions of pressure within the reser- Earth, Moon Planets 69, 141–172. [16] Swanson, D. et al. (1979). voir, exsolution and decompression of gas bubbles [17] Harris, A. et al. (1999). [18] Witham and Llewellin, (2006). [19] Stovall, W.K., et al. (2009) Bul. of Vol., 71, 767.