Geophysical Research Letters

Lunar backscatter and neutralization of the solar wind: First observations of neutral atoms from the Moon



[1] The solar wind continuously flows out from the Sun, filling interplanetary space and impinging directly on the lunar regolith. While most solar wind ions are implanted into the lunar dust, a significant fraction is expected to scatter back and be emitted as energetic neutral atoms (ENAs). However, this population has never been observed, let alone characterized. Here we show the first observations of backscattered neutral atoms from the Moon and determine that the efficiency for this process, the lunar ENA albedo, is ∼10%. This indicates that the Moon emits ∼150 metric tons of hydrogen per year. Our observations are important for understanding the universal processes of backscattering and neutralization from complex surfaces, which occur wherever space plasmas interact with dust and other small bodies throughout our solar system as well as in exoplanetary systems throughout the galaxy and beyond.

1. Introduction

[2] The Moon emits a broad array of scattered or secondary radiation and particles. The geometric albedo (GA) of the Moon depends on the surface composition and ranges from ∼11.3% for visible light to ∼2–8% in the interval 60–170 nm. (The geometric albedo is the ratio of the surface brightness at zero phase angle (i.e., reflected back toward the incident radiation source) to an idealized fully reflective (and diffusely scattering) surface of the same geometrical cross-section.) The latter contribution is produced by the fluorescence of surface minerals [Flynn et al., 1998]. The total lunar GA drops to ∼0.014% at 5 nm in the soft X-ray region [Flynn et al., 1998]. Images of the Moon by ROSAT and Chandra showed that the illuminated side glows in X-rays from 0.5–7 keV, while the dark side spectra show charge exchange products from solar wind interactions with the geocorona [Schmitt et al., 1991; Wargelin et al., 2004]. The Moon emits gamma rays and neutrons when galactic GeV cosmic ray protons and alphas interact with the surface [Thompson et al., 1997; Feldman et al., 1998]. The Moon is also brighter than the Sun at gamma-ray energies of 100–300 MeV, with fluxes ranging from ∼1–5 × 10−7 cm−2s−1 [Orlando and Strong, 2008].

[3] Solar wind ions, which are 96% H and 4% He with only traces of heavy species, have a typical speed of ∼400 km s−1 (∼1 keV/nucleon) and impinge continuously on the sunward side of the Moon. These ions weather the lunar surface, which is also “gardened” by micrometeorites. In the geomagnetic tail, plasma sheet ions are incident on the Moon isotropically at mean energies of ∼4keV. The solar wind sputters neutral atoms out of the regolith with energies of several eV. This sputtering erodes surface grains at a rate of 0.01–0.04 nm/yr [e.g., Wehner et al., 1963; Johnson and Baragiola, 1991], however, the porous nature of the regolith may trap >90% of forward-directed sputtered species [Hapke, 1986]. Ion sensors onboard the SELENE spacecraft in lunar orbit recently detected backscattered solar wind particles and found that ∼0.1–1% of the solar wind protons bombarding the lunar surface backscatter from the Moon as ions [Saito et al., 2008]. During the backscatter process, however, incident solar wind ions are mostly neutralized, and the vast majority should exit as ENAs with a broad distribution of energies up to the solar wind energy. These ENAs then travel along ballistic trajectories, unimpeded by magnetic fields from the lunar surface, Earth's magnetotail, or interplanetary space, making them observable at significant distances from the Moon. Finally, solar wind passing downstream along the terminators of the Moon can be neutralized by passing through the lunar exosphere [Futaana et al., 2008] or very small levitated dust grains [Wimmer-Schweingruber and Bochsler, 2003]. Pickup ions from this lunar exosphere were observed with the AMPTE SULEICA sensor in the mass ranges around O and Si [Hilchenbach et al., 1993]. However, to date, no neutral atoms from any of these lunar sources have been directly observed.

2. Observations

[4] NASA's Interstellar Boundary Explorer (IBEX) mission was launched on October 19, 2008 and is currently in a very high altitude (∼3 × 105 km apogee) Earth orbit, making the first all-sky maps of ENAs produced by the interaction of the heliosphere with the local interstellar medium at heliocentric distances of ∼100 AU [McComas et al., 2004, 2009]. On December 3, 2008, early in the first orbit after IBEX's higher energy sensor (IBEX-Hi) [Funsten et al., 2009] was turned on, the Moon crossed through its field of view (FOV), as shown in Figure 1. The pointing of IBEX's spin axis results in observations of solar wind ions backscattered as ENAs into a narrow angular range around ∼98° relative to the solar wind velocity vector. Even though the Moon was ∼2 × 105 km away, IBEX's exceptional sensitivity allowed this first detection of lunar ENAs.

Figure 1.

(top) First detection of lunar ENAs and (bottom) geometry for the IBEX observations (Earth, Moon, and spacecraft not to scale). IBEX spins at roughly 4 rpm with its 7° FWHM field of view sweeping over the Moon each spin for ∼10 hours on December 3, 2008. The ENAs are summed in 6° bins with the lunar direction indicated by the white arrow. The light blue horizontal line shows the lunar ENAs. The wider range of pixels at ∼1600 is an instrumental effect that we did not remove from this raw data plot. The color bar indicates triple coincidence counts from the lowest two ESA steps (0.38–0.95 keV FWHM) summed over 96 spins. IBEX detects ENAs produced by backscattering and neutralization of the incident solar wind protons at an angle of ∼98° from the initial solar wind velocity vector.

[5] Figure 2 shows count rates of lunar ENAs from IBEX-Hi; the requirement of “triple coincidence” reliably suppresses noise [Funsten et al., 2009]. The color coded curves represent the four lowest energy pass bands of the instrument, spanning ∼0.38–2.5 keV FWHM. For the first third of the ∼10 hours of lunar viewing, the solar wind speed was low and the flux decreased significantly. During this interval most of the ENAs were detected in the lowest, and least sensitive, energy pass band (0.38–0.59 keV). Over the remainder of the lunar viewing, the solar wind speed (energy of incident protons) increased, producing enhanced fluxes in the second (0.52–0.95 keV) and then third (0.84–1.4 keV) energy channels as the solar wind protons reached energies of ∼1 keV. Thus, the lunar interaction produces backscattered neutrals with a broad energy spectrum from the incident solar wind energy downward throughout the range of energies measurable by IBEX.

Figure 2.

ENAs detected by IBEX-Hi and associated solar wind conditions during the passage of the Moon across the IBEX-Hi field of view. (top) The ENAs are triple coincidence measurements with a linearly interpolated background subtraction and a running average smoothing over each eight consecutive data points. (middle) The solar wind proton flux and (bottom) speed (equivalently proton energy on right scale) are from the ACE spacecraft ∼1.2 × 106 km upstream from the Moon. Using the measured solar wind speed, the ACE-Moon distance, the Moon-IBEX distance and a typical ENA energy of ∼500 eV gives the ∼84 minute and ∼63 minute delays for the start and end of the event, respectively, as shown.

[6] We simulated the interaction of solar wind protons with the lunar regolith utilizing the Stopping and Range of Ions in Matter (SRIM) code [Ziegler et al., 2008], which has been used extensively for modeling ion-solid interactions and is ideally suited to simulate amorphous-rimmed grains characteristic of the regolith [Keller and McKay, 1993]. We assumed a regolith composition 80% by mass, typical for highlands, and 20% for maria based on the IBEX viewing geometry. Similar calculations assuming significant compositional variations, including the addition of 10% (by number) implanted H, provided similar results. The SRIM simulations indicate that, with the exception of near grazing collisions, the backscatter probability for ENAs within the IBEX-Hi energy range decreases nearly linearly with increasing energy up to the solar wind energy.

[7] Using this result, we convolved the IBEX-Hi angular and energy response functions with a modeled ENA energy spectrum, assuming for each measurement an ENA spectrum had a cut off at the incident solar wind energy and increased linearly with decreasing energy. We also convolved the detailed shapes of the IBEX-Hi energy pass bands, which included, for example, integrating the lowest energy channel down to its 1% response level at 238 eV. In this process, the slopes of the backscattered distributions as a function of energy derived from SRIM dropped out of the calculation making these results highly model independent.

[8] The measured ENA flux observed at IBEX (scaled back to the lunar surface with the distance squared) is shown in Figure 3 as a function of incident solar wind flux. Clearly, the measured flux is well correlated with the incident solar wind, as expected for the solar wind being the source population for the ENAs observed by IBEX. The slope of this curve is the ENA albedo as viewed by IBEX at its unique backscatter angle of ∼98°. This albedo accounts for the probabilities of proton backscatter, neutralization, and escape from the rough lunar surface, integrated over the approximately quarter of the Moon that is simultaneously hit by the solar wind and viewed by IBEX. We find a best fit slope of ∼0.005 sr−1. Assuming that this value is similar at other viewing angles, which seems reasonable given the lunar surface roughness on various scales, the ENA albedo within the IBEX-Hi energy range emitted into 2π sr is ∼3.8%. SRIM simulations indicate that ENA fluxes below IBEX-Hi's energy range may be one to two times the number observed, leading to an estimate of the full ENA albedo of ∼10%.

Figure 3.

Observed ENA flux as a function of solar wind flux. The measurements show ENA observations from Figure 2 where the solar wind was propagated individually for each sample. Error bars represent statistical counting uncertainty only; additional errors associated with IBEX-Hi's angular and energy responses are largest for the lowest solar wind energies and times when the Moon was at the edge of the FOV, which correspond to the right side of this figure where the solar wind flux >3.5 × 108 cm−2 s−1. The solar wind ions produce lunar ENAs through backscatter and neutralization on the lunar surface. Multiplying the measured slope of ∼0.005 sr−1 by 2π and accounting for ENAs below IBEX-Hi's energy threshold leads to a total lunar ENA albedo of ∼10%.

[9] This ENA albedo is consistent with the recent measurements of 0.1–1% backscatter efficiency of solar wind protons [Saito et al., 2008], since the fraction that comes off ionized is only expected to be a few percent of the backscattered neutral fraction. Our estimate is also consistent with experimentally determined trapping efficiencies for light solar wind ions on various solids [e.g., Bühler et al., 1966], which are typically 80 to 90% [e.g., Grimberg et al., 2009]. We stress that these experimental results represent efficiencies for He at solar wind energies interacting with solids. Hydrogen is lighter, and presumably, has a somewhat lower trapping efficiency because it interacts more strongly with heavier ions in the regolith. Thus, an ENA albedo of ∼10% or even higher should not be surprising. Furthermore, our SRIM simulations over all the surface orientations relative to the solar wind indicate a total backscatter yield of ∼30%. Comparing this with the total ENA albedo of ∼10% yields the escape probability of a backscattered ENA from the lunar surface of ∼30%, which is consistent with simulation results of the escape of sputtered particles from a regolith [Cassidy and Johnson, 2005].

3. Conclusion

[10] In conclusion, we report here the first observations of ENAs from the Moon. We show that their energy spectrum and total flux follow the variations in the incident solar wind energy and flux. We also derive an ENA neutral hydrogen albedo for the Moon of ∼10%, for the first time. This measured ENA albedo results in a continuous emission of hydrogen atoms from the Moon at a rate of ∼3 × 1024 s−1 (∼1.5 × 105 kg yr−1). The implications of these observations are far reaching. In addition to quantifying the production of neutrals from the interaction of the solar wind with a solid body such as the Moon, these results can be applied to the solar wind interaction with other objects throughout the solar system from tiny dust grains to asteroids, Kuiper belt objects, the moons of Mars, and when the solar wind pressure is high enough, even Mercury, which appears to have a surface similar to the lunar highlands [Taylor et al., 2006]. These neutrals augment the neutral hydrogen geocorona and, eventually, will again become ionized and swept out of the heliosphere with the solar wind. We have thus identified another source of heliospheric pickup ions. Plasma-surface interactions are also important in a broad range of astrophysical contexts. About 16% of cool stars have debris disks that signal the presence of exoplanet systems [Trilling et al., 2008], while up to 28% of known giant gas exoplanets may have moons, with some of them located in regions of intense stellar wind [Scharf, 2006]. The formation of lunar ENAs is directly relevant to plasma-surface interactions on these solid surfaces as well as in interactions with the gaseous/dusty proto-stellar nebulae that surround accreting stars during their T-Tauri phases. The results from this study thus provide important clues to the evolution of dust and rocky moons in exoplanet systems throughout and beyond our Galaxy.


[11] We gratefully acknowledge all of the contributions made by the outstanding men and women of the IBEX team who have been and are making this mission a tremendous success. This work was funded by NASA as a part of the Explorer Program under contract NNG05EC85C; parts of IBEX were also funded by the Swiss Prodex program.