Elsevier

Icarus

Volume 214, Issue 1, July 2011, Pages 21-29
Icarus

A ground-based observation of the LCROSS impact events using the Subaru Telescope

https://doi.org/10.1016/j.icarus.2011.05.008Get rights and content

Abstract

The Lunar Crater Observation and Sensing Satellite (LCROSS) mission was an impact exploration searching for a volatile deposit in a permanently shadowed region (PSR) by excavating near-surface material. We conducted infrared spectral and imaging observations of the LCROSS impacts from 15 min before the first collision through 2 min after the second collision using the Subaru Telescope in order to measure ejecta dust and water. Such a ground-based observation is important because the viewing geometry and wavelength coverage are very different from the LCROSS spacecraft. We used the Echelle spectrograph with spectral resolution λλ  10,000 to observe the non-resonant H2O rotational emission lines near 2.9 μm and the slit viewer with a K′ filter for imaging observation of ejecta plumes. Pre-impact calculations using a homogeneous projectile predicted that 2000 kg of ejecta and 10 kg of H2O were excavated and thrown into the analyzed area immediately above the slit within the field of view (FOV) of the K′ imager and the FOV of spectrometer slit, respectively. However, no unambiguous emission line of H2O or dust was detected. The estimated upper limits of the amount of dust and H2O from the main Centaur impact were 800 kg and 40 kg for the 3σ of noise in the analyzed area within the imager FOV and in the slit FOV, respectively. If we take 1σ as detection limit, the upper limits are 300 kg and 14 kg, respectively. Although the upper limit for water mass is comparable to a prediction by a standard theoretical prediction, that for dust mass is significantly smaller than that predicted by a standard impact theory. This discrepancy in ejecta dust mass between a theoretical prediction and our observation result suggests that the cratering process induced by the LCROSS impacts may have been substantially different from the standard cratering theory, possibly because of its hollow projectile structure.

Highlights

► We conducted infrared spectral and imaging observations of the LCROSS impacts using the Subaru Telescope. ► No unambiguous emission line of H2O or dust was detected. ► The upper limit for dust mass was significantly smaller than that predicted by a standard impact theory. ► The cratering process may have been greatly different from the standard cratering theory possibly because of its hollow projectile structure.

Introduction

The permanently shadowed regions (PSRs) near the lunar poles may have a significant mass of volatile deposits, including H2O ice, presumably supplied by cometary and asteroidal impacts, because the temperature of the PSRs have remained extremely cold over geologic time scales (Watson et al., 1961, Arnold, 1979). If such volatile deposits exist, its chemical and isotopic compositions are of great scientific interest and will also be an important resource for the future manned missions. The first data suggesting the existence of ice were obtained by radar observations by the Clementine mission (Nozette et al., 1996). However, the presence of water ice was not necessarily supported by a subsequent reanalysis (Simpson and Tyler, 1999) or ground-based radar observations (Stacy et al., 1997, Campbell et al., 2003, Campbell et al., 2006). More recently, however, the neutron spectrometer of the Lunar Prospector (LP) discovered high hydrogen abundances in the lunar poles, strongly supporting the presence of volatile deposits (Feldman et al., 1998, Feldman et al., 2001, Lawrence et al., 2006). The data have raised several important questions. First, what is the molecular state (H2O, H2, CnHm, etc.) of the hydrogen at the lunar poles? This question is connected strongly with the origin of the hydrogen, and there are many proposed hypotheses. For example, impacts of comets and meteorites, implantation of solar wind, supply from the lunar interior or terrestrial magnetosphere or interstellar molecular cloud (e.g., Stern, 1999, Lucey, 2009). Second, what is the vertical profile of hydrogen distribution? The radar and neutron observations can probe the average concentration of the top ten centimeters from the surface. LP’s neutron observation strongly supports the high hydrogen abundance within the PSR’s including that inside Shackleton crater, but SELENE’s imaging observation of the PSR inside Shackleton indicates no exposed ice on the surface (Haruyama et al., 2008). Thus, the vertical profile of ice concentration may be very complicated. Third, the horizontal distribution of the hydrogen is also very unclear. Third, what is the horizontal distribution of the hydrogen? The neutron data suggest the possibility of the presence of hydrogen in the lunar sunlit regions outside the PSRs. Furthermore, a few weeks before the LCROSS impacts, the discoveries of the absorption spectra near 3 μm consistent with hydrates in the lunar high latitudes were published (Clark, 2009, Sunshine et al., 2009, Pieters et al., 2009). These results suggest that hydrogen in the PSR’s in the high latitudes may not be H2O but may be hydrated minerals. Thus, the presence of the hydrogen within the PSRs is not decisive evidence for water ice.

The Lunar Crater Observation and Sensing Satellite (LCROSS) mission was designed to directly investigate the presence of water on the Moon by excavating surface materials in a PSR in a lunar pole (Colaprete et al., 2010a). The target PSR was selected inside Cabeus crater. Two projectiles were planned to impact at velocity 2.5 km s−1 and incident angle 85° ± 5°. The empty Centaur upper stage of the Atlas V (mass ∼2000 kg, bulk density ∼28 kg m−3) was used as the primary excavating projectile. Although the Shepherding Spacecraft (S-S/C) with 700 kg of mass was not intended as a projectile but contained monitoring instruments, it also served as a secondary projectile for instruments on the Earth. When Centaur impacted, the S-S/C was observing the ejecta from above the impact site. Approximately four minutes later, the S-S/C also impacted the same PSR. Cabeus crater was chosen as the impact site due to its high hydrogen abundance confirmed by the high-spatial-resolution neutron observation by the Lunar Exploration Neutron Detector (LEND) on the Lunar Reconnaissance Orbiter (LRO) (Mitrofanov et al., 2010, Colaprete et al., 2010a), although the results of the radar observation by Chandrayaan-1 were negative (Spudis et al., 2010).

If a large amount of ejecta was thrown over the Cabeus rim (2 km), it could be observed by ground-based telescopes. A pre-impact estimate by SPH calculations (Korycansky et al., 2009) indicates that the mass of dust thrown over the Cabeus rim is 2 × 104 kg, and the mass of H2O is 200 kg, assuming 1 wt.% (i.e., a typical water content estimated by LP observation). Assuming ballistic trajectories above the ground, we can estimate that based on a SPH simulation by Korycansky et al. (2009) for a homogeneous cylindrical projectile shows that the maximum mass of ejecta between 3.0 km and 3.7 km altitude (i.e., the analyzed area within the imager field of view (FOV)) is about 2000 kg at ∼10 s after the Centaur impact. Similarly, the maximum H2O mass observed with 40 s of exposure between 2.5 km and 3.0 km of altitude (i.e., the FOV of the slit of Subaru IRCS) was predicted to be about 10 kg assuming 1 wt.% water content.

It is noted that the estimated dust mass depends on two factors. Korycansky et al. (2009) indicate that the hollow structure of the projectile affects the dust mass. In their calculations, a two-plate projectile separated by 10 m with no rotation (i.e., impact of the Centaur upper stage standing vertical to the surface) yields the maximum dust mass of about 1100 kg in our analyzed area at ∼15 s after the impact. This mass is almost half of that for a homogeneous projectile, but a SPH simulation estimated that a 90°-rotated two-plate projectile (i.e., impact of the Centaur upper stage lying parallel to the surface) gives a maximum dust mass (1800 kg at ∼55 s after) comparable to that for a homogeneous projectile. Thus, it was not very clear the degree of the effect of hollow nature of a projectile. In contrast, a significant reduction in ejecta mass due to hollow nature of projectiles was argued based on laboratory experiments (Schultz et al., 2009, Hermalyn et al., 2009a, Hermalyn et al., 2009b). Another factor that can affect the dust mass is porosity of the target material. Based on the SPH simulations including the effect of early-time compression due to porous target material, Korycansky et al. (2009) found that the impact to porous material yields a only slightly smaller high-velocity ejecta mass than the impact to nonporous material, and the total mass reaching an altitude of 2 km is almost same between porous target and nonporous target. However, based on the laboratory experiments, Schultz (2006) argued for a much lower total ejected mass (2800 kg higher than an altitude of 2 km about 50 s after the impact; ∼1/7 of prediction for no rotated projectile by Korycansky et al. (2009)) due to early-time compression effects in the regolith. Thus, comparison between the ejecta amount derived from pre-impact estimates using cratering theories and laboratory experiments and that from actual observations from both the S-S/C and the ground-based telescopes is very valuable. In fact, LCROSS mission has the significance that it is the first near-vertical impact experiment on the lunar surface with close coordination with many Earth-based large telescopes (e.g., Heldmann et al., 2011). Although, Japanese spacecraft HITEN impacted the Moon at a near vertical angle, there were not many large telescopes observing it (e.g., Uesugi et al., 1994).

The main objectives of our observation were to measure water independently from the LCROSS spacecraft and to obtain constraints on the vertical distribution of column density and H2O concentration of ejecta curtains from LCROSS impacts. The vertical distribution of H2O would constrain the depth distribution of underground lunar ice in the PSR. The vertical distribution of column density of an impact ejecta curtain is controlled by the mass-velocity relation and ejection angles. Such data will allow us to test the validity of cratering theories and scaling laws and it could also serve as a test for the above different predictions on ejection mechanisms. Although the vertical distribution of ejecta can also be inferred based on top-view from the S-S/C (e.g., Schultz et al., 2010), the side-view observation from ground-based telescope has the advantage that it can obtain the constraints on vertical ejecta mass distribution very directly. The same technique was used for the Deep Impact collision on comet 9P/Tempel 1; the ground-based observation with the Subaru Telescope revealed the stratified nature of the surface layer on the comet (Kadono et al., 2007). Furthermore, wavelength resolution and range unavailable for the S-S/C can be used by ground-based telescopes. This is also very similar to the situation for the Deep Impact collision on a comet (e.g., Sugita et al., 2005).

The LCROSS impact was carried out on schedule. The data obtained by the S-S/C indicate the generation of an impact crater 25–30 m in diameter and an impact plume which had about 8 km in diameter 40 s after the impact (Schultz et al., 2010). Based on near-infrared (NIR) and ultraviolet/visible (UV/VIS) spectroscopic observations by the S-S/C, Colaprete et al. (2010b) estimate that more than 150 kg of water was contained in the impact plume. Also, LCROSS-impact-released 1.5 ± 1 kg of Na has been detected by a ground-based observation (Killen et al., 2010). The weather condition at the top of Mauna Kea, however, was not very good during the night of the observation. The time variation of the relative humidity was large; more than 80% approximately two hours before the observation. Nevertheless, the weather condition was improving very rapidly toward the beginning of the LCROSS impacts; the relative humidity reduced to about 30% in the beginning of the observation. This enabled us to conduct the observation. There were also cirrus clouds in the high altitudes, causing significant light scattering.

Section snippets

Imaging observation and analysis

We carried out imaging observation of the LCROSS impact ejecta with the Infrared Camera and Spectrograph (IRCS) at the Subaru Telescope on Mauna Kea, Hawaii (Kobayashi et al., 2000). The IRCS can obtain 1024 × 1024 pixel images simultaneously with spectroscopic observation. The infrared K′ band filter (central wavelength 2.12 μm, wavelength width 0.35 μm) was used, because atmospheric absorption in this wavelength range is small. We obtained images with 0.5 s exposure time every 4 s throughout the

Imaging observation results

Fig. 3a and b are the images taken 51 s before the Centaur impact and 51 s after, respectively. The impact site is located underneath the center of the slit not directly seen from the Earth, because it is blocked by the high crater rim of Cabeus crater. If the Centaur impact ejected a large amount of lunar surface material as predicted by SPH simulations (e.g., Korycansky et al., 2009), an inverted cone-shaped ejecta would be observed around the center of the slit. Fig. 3c and d shows the image

Implications for the cratering mechanism

We discuss why the ejecta mass from the LCROSS impact would be less than predicted from standard theoretical models. Comparison with our upper limit estimate (800 kg) for dust mass in the analyzed area for imaging observation (3.0 km to 3.7 km of altitude) indicates that the actual Centaur impact generated less than a half of the predicted dust mass (2000 kg) by the SPH calculation for a homogeneous projectile. Instead of a homogeneous projectile, Korycansky et al. (2009) also calculate the dust

Conclusions

We observed infrared spectra of water emission lines in the ejecta of the LCROSS impact using the Subaru Telescope to measure water independently of the LCROSS spacecraft. We also observed infrared images to constrain the vertical distribution of the ejecta mass. Although pre-impact simulations using a homogeneous projectile predicted that 2000 kg mass of dust could be excavated into the analyzed area (3.0–3.7 km from the crater floor) immediately above the slit FOV within the imager FOV, no

Acknowledgments

The authors would thank David A. Paige for letting us use his lunar topography projection software and Mike Kelly from University of Maryland for providing a script for telescope pointing of the impact site at the Moon. This observation was conducted under Subaru Open Use Program (S09A-154) and supported partially by Grant in Aid from Japan Society for the Promotion of Science.

References (47)

  • A. Dollfus et al.

    Polarimetric properties of the lunar surface and its interpretation

    Astron. Astrophys.

    (1970)
  • W.C. Feldman et al.

    Fluxes of fast and epithermal neutrons from Lunar Prospector: Evidence for water ice at the lunar poles

    Science

    (1998)
  • W.C. Feldman

    Evidence for water ice near the lunar poles

    J. Geophys. Res.

    (2001)
  • D.B. Goldstein et al.

    Modeling the vapor and dust dynamics due to impact of the LCROSS spacecraft on the Moon

    AIP Conf. Proc.

    (2008)
  • J. Haruyama

    Lack of exposed ice inside lunar south pole Shackleton crater

    Science

    (2008)
  • Hayano, Y. et al., 2008. Current status of the laser guide star adaptive optics system for Subaru Telescope. SPIE Proc....
  • Hayano, Y. et al., 2010. Commissioning status of Subaru laser guide star adaptive optics system. SPIE Proc. 7736,...
  • Heldmann, J.L. et al., 2011. LCROSS (Lunar Crater Observation and Sensing Satellite) Observation campaign: Strategies,...
  • Hermalyn, B., Schultz, P.H., Colaprete, A., 2009a. LCROSS impact conditions and ejecta evolution: Insight from...
  • Hermalyn, B., Schultz, P.H., Heineck, J.T., 2009b. Early-stage ejecta velocity distribution. Proc. Lunar Sci. Conf. XL,...
  • Hermalyn, B., Schultz, P.H., Colaprete, A., Shirley, M., Ennico, K., 2010. LCROSS ejecta dynamics: Insight from...
  • K.R. Housen et al.

    Crater ejecta scaling laws: Fundamental forms based on dimensional analysis

    J. Geophys. Res.

    (1983)
  • T. Kadono et al.

    The thickness and formation age of the surface layer on Comet 9P/Tempel 1

    Astrophys. J.

    (2007)
  • View full text