Lunar weather measurements at three Apollo sites 1969–1976



[1] The first lunar weather stations, matchbox-sized, 270 g Apollo Dust Detector Experiments about 100 cm above the surface of the Moon near Apollo 12, 14, and 15 landing sites, measured dust accretion, charged particle radiation, and temperature changes—three environmental factors proved during Apollo to affect technical systems deployed on the Moon. Degradation of seven horizontal solar cells was measured every lunar daytime from 1969 to 1976. The anomalously intense August 1972 solar particle event (SPE) degraded three covered cells by less than 1%, while two cells desensitized by intense preirradiation showed no measurable effects. Although independent studies estimated the long-term fluence bombarding the cells was less than half that of the August SPE, long-term gradual degradation of five covered cells (normalized to 2000 days) was an order of magnitude greater, between 4% and 10%. If the long-term effects were totally caused by dust, with articulated caveats including simulated (maria) Minnesota Lunar Simulant-1 dust particles with diameters 20 to 38 µm, this provides the first direct measured long-term net accretion of dust with an upper limit of order 100 µg cm−2 yr−1, equivalent to a layer 1 mm thick in 1000 years, but it may be significantly less. Two bare cells were abruptly degraded by 7% during the August SPE, however long-term they measured additional damage of 29% and 24%, indicating a long-neglected suite of low-energy radiation, posing risks for bare materials exposed on the surface of the Moon.

1 Introduction

[2] The Apollo missions, recognized as among the most magnificent and daring human expeditions to date, were prepared in full recognition of radiation hazards in space with four individual radiation dosimeters [Bailey, 1975]. However, there was little preparation for the risks caused by sticky, abrasive fine-powdered lunar dust [Bean et al., 1970; Beattie, 2001; Gaier, 2005; Wagner, 2006; Hapke, 2006] which led to inescapable difficulties best summarized by the last astronaut on the Moon, Gene Cernan, “We can overcome other physiological or physical or mechanical problems except dust” [Cernan, 1973]. NASA required dust covers for susceptible experiments of the Apollo lunar surface experiments package [Beattie, 2001], however the only Apollo risk-management experiment to measure possible dust effects was the Dust Detector Experiment (DDE) invented on 12 January 1966 [O'Brien, 1966, 2009]. The purpose of this paper is to report the first simultaneous measurements of radiation damage and dust accretion on seven horizontal solar cells of the Apollo 12, 14, and 15 DDEs located 100 cm above the lunar surface from November 1969 to 29 February 1976 and to distinguish between the two effects.

[3] The particular purposes of the DDEs were to assist risk management for (i) dust causing overheating and mechanical damage of experiments and hardware [Gaier, 2005; Gaier and Jaworske, 2007; Wagner, 2006], (ii) radiation damage and failures of solar cell power supplies, and (iii) toxicity of very fine dust to humans [Liu et al., 2008].

[4] Greatly increased knowledge of radiation in proximity of the Moon has since been provided by such instruments as CRaTER on the Lunar Reconnaissance Orbiter [Kozarev et al., 2009; Schwadron et al., 2012]. To this, we add here the first measurements and comparisons of damage from both radiation and dust to bare and covered solar cells on the Moon. New knowledge of the long-term net accretion of dust is topical and necessary for many reasons, including Ground Truth support of the Lunar Atmosphere and Dust Environment Explorer (LADEE) which will orbit the Moon in late 2013 [Elphic et al., 2012]. It is understood that the Chinese lunar rover of Chang'e 3 will also be operational in late 2013. Furthermore, dust accretion is suspected of causing degradation and limiting ongoing operations of Apollo Lunar Laser Ranging RetroReflectors (LRRRs) [Currie and LLRRA-21 Teams, 2010; Murphy et al., 2010].

2 Apollo Dust Detector Experiments (DDEs)

[5] The minimalist matchbox-sized Apollo 12 DDE has three orthogonal similar 2 cm × 2 cm n-on-p 10 ohm cm silicon solar cells with blue filters and heavily shielded against radiation by 1.5 mm silicon covers so as to focus on dust detection. The cell used here is horizontal with an individual thermistor in thermal contact underneath (Figure 1) [O'Brien, 1966]. Apollo 14 and 15 modified DDEs, also known as the Dust, Thermal, Radiation and Thermal Engineering Measurements (DTREM) package, have three cells 2 cm × 1 cm, all horizontal, one bare, and two with glass covers 0.15 mm thick, one of which was irradiated before launch by 1015 1 MeV electrons cm−2 [NASA, 1974; Sullivan, 1994; Bates et al., 1969]. Solar cells had not been used previously for dust detection in space and all solar cells are several orders of magnitude smaller than conventional impact ionization dust detectors in space such as the one on the Lunar Atmosphere and Dust Environment Explorer (LADEE). Each modified DDE has only one “Cell Thermometer” with two other nickel resistance thermometers measuring lunar surface temperature [Hickson, 1969]. DDE modifications and rationale are outside the scope here and are discussed by B. J. O'Brien (manuscript in preparation, 2013).

Figure 1.

(left) Modified Apollo 14 and 15 DDEs (also known as DTREM) with three horizontal and different solar cells and (right) the original orthogonal configuration of the Apollo 12 DDE.

Figure 2.

Apollo 12 DDE 100 cm above lunar surface on Central Processing Unit (Source: NASA AS12-47-6927).

[6] The output of each solar cell was the short-circuit current measured across a 1 Ω resistor [Bates and Fang, 1992]. Voltages from three solar cells and three thermometers were converted by a shared analog-to-digital converter into a serial stream of 8 bit digital words as “counts” in about 54 s cycles, with 6 DDE channels of data transmitted to Earth and recorded [O'Brien et al., 1970]. NASA announced in 2006 that its DDE data and tapes had been misplaced before archiving. These long-term DDE data were not analyzed rigorously for over 3 decades until after O'Brien [2009] revisited partial personal files and then in collaboration with Professor Yosio Nakamura at the University of Texas retrieved Apollo 12, 14, and 15 DDE data from his Apollo Passive Seismometer data, which had been copied into accessible modern formats in 1990–1991 [Nakamura, 1992; UWA media release 26 Nov, 2009; McBride et al., 2013].

3 Methods and Measurements

[7] Previous reports on the DDEs studied the first 100 h soon after human activities began on each Apollo landing [O'Brien, 2009]. All measurements used here begin several hundred hours after direct human contamination ended. Conditions of cells at the beginning of operation can be gauged from O'Brien [2009, 2011].

[8] The rigorous methodologies presented below aim to distinguish between causes of degradation of sunlit solar cells on the Moon by (i) direct obscuration of sunlight by dust on their surface with a nonlinear response dependent on particle size and shape [Katzan et al., 1991]; (ii) penetrating charged-particle radiation displacing silicon atoms in the crystal lattice of the cell thereby degrading logarithmically the optical efficiency of the cell response to sunlight [Tada et al., 1982]; and (iii) other effects such as discoloration and/or darkening reducing cell cover transmittance.

[9] Mass of dust accreted is estimated here by translating percentage changes in counts using results of laboratory tests by Katzan et al. [1991]. A layer of about 0.5 mg cm−2 of 20 to 38 µm particles of simulated dust Minnesota Lunar Simulant (MLS)-1, a basaltic-based simulant for dust in lunar lowlands (maria) [Colwell et al., 2007; Taylor and Liu, 2010], causes about a 10% decrease in short-circuit current for spherical dust particles and about 20% for cubic particles [Katzan et al., 1991]. We assume with Katzan that accretion of dust is uniform over each cell, but have no reason to make the assumption for the lunar surface. Marshall et al. [2011] examined in the laboratory whether dust accumulated as a uniform spread of fine grains or as clumps. Astronaut photos of Apollo 12 DDE (NASA AS12-47-6927) and Apollo 14 DDE (NASA AS14-67-9381) give examples of each respectively, while Apollo 15 DDE (NASA AS15-86-11592) gives examples of both [Apollo Image Atlas, 2013].

[10] Particle radiation on the Moon varies over a wide range of phenomena and suites of electrons, protons, and heavy ions with energies and intensities widely varying in space and time, including high energy, penetrating particle radiation which are hazardous to human health and semiconductors, sporadic solar particle events (SPEs) and ubiquitous galactic cosmic radiation (GCR) [National Research Council, 2008; Farrell et al., 2012; Cucinotta et al., 2013; Schwadron et al., 2012]. The most important cause of permanent radiation damage in solar cells is displacement of silicon atoms in the crystal lattice [Tada et al., 1982; Messenger et al., 2002], also known as Nonionizing Energy Loss (NIEL), as distinct from the thousandfold greater energy loss by ionization that causes the better known biological damage measured in units of Sievert. Displacement damage is caused by electrons and protons above 150 keV and 100 eV, respectively, in bare silicon cells [Srour and McGarrity, 1988], peaking for protons around 1 to 3 MeV [Tada et al., 1982; Messenger et al., 2002] and diminishing with increased energy over 3 to 3000 MeV [Burke, 1986; de Lafond, 1969]. GCR particles, richer in higher energies, are therefore less effective than SPE particles in damaging solar cells.

[11] Solar cells are insensitive and primitive radiation instruments. A 100-fold increase in radiation fluence of 10 MeV protons from 1011 cm−2, or about 3000 times that fluence of 1 MeV electrons, reduces the short-circuit current of n-on-p 10 ohm cm solar cells by only one third [Tada et al., 1982, Figure 4]. For isotropic fluxes, DDEs only crudely differentiate between energies of protons by using covers of 0.15 mm or 1.5 mm to block vertically incident protons with less than 4 MeV or 20 MeV, respectively (Table 1) [after Bates, 1974 in NASA, 1974; Tada et al., 1982]. Covers also slow down higher-energy particles, some of which stop in the cell, inflicting greater damage [Messenger et al., 1997]. As a rule of thumb, for isotropic fluxes, 90% of the damage is caused by particles with energies ranging from just exceeding the threshold imposed by the cover to 4 times this energy.

Table 1. Shielding Covers and Sensitivity to Particle Radiation of Apollo 12, 14, and 15 DDE Solar Cells
ApolloCell CoveringThresholds (MeV)Sources of Radiation Damagea
  1. aSPE is defined above; GCR is Galactic Cosmic Radiation; SW is Solar Wind.
121.5 mm200.2SPE, GCR
14Bare0.060.2SPE, GCR, SW, Others
0.15 mm40.2SPE >4 MeV, GCR
0.15 mm preirradiated40.2Relatively inert to radiation
15Bare0.060.2GCR, SPE, SW, Others
0.15 mm40.2GCR, SPE > 4 MeV
0.15 mm preirradiated40.2Relatively inert to radiation

[12] Disaggregation of radiation effects from all other causes including dust is treated in discussions below, optimized by comparison between astronomically normalized counts using solar positional information through ephemerides generated by the Jet Propulsion Laboratory Horizons System [Hollick, 2011], thus eliminating the variations in counts in response to cyclic brightness of sunlight dominated by varying Sun-Moon distances, shown as inset to Figure 3. Corrections are also made for deployment tilts of each DDE in the East–West plane, derived here by three independent methods for the Apollo 12, 14, and 15 DDEs to be 4–5°, 4–6°, and 1.5–2.5°, respectively [Hollick, 2011], in order to extract relevant solar positional information relative to the cells' surfaces not parallel to the local lunar horizontal.

Figure 3.

Astronomically normalized long-term counts at lunar noon of Apollo 12 (black), 14 (blue), and 15 (red) covered horizontal solar cells with respect to Earth time—from each day of deployment, 19 November 1969, 5 February 1971, and 31 July 1971 (respectively), until 29 February 1976. Normalization removes the first-order 7% seasonal variation with Sun-Moon distance in the raw data (shown in Inset). Note: The Apollo 12 cell is double the size of the Apollo 14 and 15 covered cells.

[13] Counts from the analog-to-digital converter reading of instantaneous raw voltages at noon of each Lunar Day (LD) are the parameter of choice for this analysis for five important reasons—rigor, transparency, inclusiveness, minimization of uncertainties, and slow changes of output with time. However, NASA archives of DDE digital data will not contain raw counts but processed voltages made available in the Lunar Data Node of the NASA Planetary Data System [Williams et al., 2013; McBride et al., 2013] after initial processing based on historic Sanborn charts of analog translations of telemetered digital data and thereafter two steps of processing, first to millivolts-uncorrected and then to millivolts-corrected for temperature dependence of cells (see supporting information for more detailed discussion).

4 Analyses

[14] The first application of the precision and reliability of the measurements, shown in Figures 3-6, was the detailed tracking of gradual degradation of all seven horizontal cells of the Apollo 12, 14, and 15 DDEs over their 5 to 6 years of continuous operation. A geometric 10% difference between Apollo 14 and Apollo 15 counts is caused by the Apollo 15 DDE location being at higher latitude (26.1°N) than the near-equatorial Apollo 12 and 14 locations of about 3°S. The Apollo 12 horizontal cell (Figure 3) and the Apollo 14 bare cell (Figures 4 and 5) show a quicker degradation over the first few lunations, whose cause(s) is not yet understood.

Figure 4.

Comparison of astronomically normalized counts of Apollo 14 and 15 bare cells. The correlation in seasonal behavior is evident and counts (noons) closest to perihelion are marked.

Figure 5.

Astronomically normalized counts at lunar noon of three solar cells of Apollo 14 DDE. One cell is bare (B, green), one covered by 0.15 mm glass and blue filter (C, red), and one with a similar cover was desensitized by preirradiation by 1015 electrons cm−2 of 1 MeV (P, blue). The abrupt damage to the bare cell was caused by the August 1972 Solar Particle Event (SPE).

Figure 6.

Astronomically normalized counts at lunar noon of three solar cells of Apollo 15 DDE. One cell is bare (B green), one covered by 0.15 mm glass and blue filter (C, red), and one with a similar cover was desensitized by preirradiation by 1015 electrons cm−2 of 1 MeV (P, blue). The abrupt damage to the bare cell was caused by the August 1972 Solar Particle Event (SPE).

[15] In Figures 3–6, five small cyclic fluctuations remain in the data after astronomical normalization. Plotting against Earth time in Figures 3 and 4 reveals the location of the peaks to be close to annual perihelion. The pairs of contemporaneous counts of the Apollo 14 and 15 bare cells, 1100 km apart, are also very strongly correlated (see Figure S2). Such a small but clear and highly correlated effect in 5 cycles confirms both (i) the reliability and accuracy of the astronomical normalization process removing the first-order (annual) effect of solar brightness and (ii) that the small magnitude second-order cyclic effects are caused by temperature sensitivities of each cell.

[16] No temperature corrections to counts are made in this analysis for long-term DDE temperature increases of 6 to 14°C between the cells' first and last lunar days of operation. This corresponds to further degradation of 1 to 2 counts for the most temperature sensitive cells, the preirradiated cells, whose long-term degradation without temperature correction was between 2 and 7 counts (Table 2). The result is an acceptable underestimation of the long-term degradation and thence of upper limits of dust accretion for the cells. (See supporting information for detailed discussion of uncertainties in corrections for temperature effects.)

Table 2. Long-Term Degradation of Seven Horizontal DDE Solar Cells, Including Effects of the Event (Row 5) and “non-Event” Degradation (Rows 6 to 8)
Cell1.5 mm Cover A12 HSCBare0.15 mm CoverPreirradation
  1. aThere is a ± 0.5 count inherent digitization uncertainty for each individual datum count. Therefore, the effects of the Event on preirradiated cells (see Figure 8) were too small to be measured with confidence, and in fact one gave an apparent increase and the other an apparent decrease in counts. Accordingly, in deriving the non-Event degradation for these cells we take the effect of the Event as zero. Normalized degradation to 2000 Earth days (Row 8) assumes linear degradation.
First Noon (Counts)221.6113.796.7105.890.390.868.8
Last Noon (Counts)207.373.367.596.382.484.366.5
Lifetime(Earth Days)2294185116761851167618511676
Event Fluence (Particles cm−2)1 × 1010 >20 MeV [King, 1974]~1011 >1 MeV [Simnett, 1976]2 × 1010 >10 MeV [King, 1974]
Effect of EventCounts−0.4−7.3−6.1−1.1−0.7+0.1−0.2
Non-Event DegradationCounts−13.9−33−23−8.4−7.2−6.5a−2.3a
Normalized Non-Event Degradation to 2000 Earth Days5.6%31%29%8.6%9.5%7.7%a4.0%a

[17] The second major application of the data was to measure the individual damage caused by the August 1972 SPE (indicated by arrows in Figures 5 and 6)—the Event—the most intense SPE by an order of magnitude during Solar Cycle 20 [King, 1974].

[18] The effects of the Event are clearly shown by the two bare cells (see Figures 4–6), however much smaller effects in the five covered cells required two measurement methods: (i) the difference between counts at noon on the Lunar Day (LD) before and noon after the Event; (ii) “best fit” estimates from 10 LDs before and 10 after the Event. Figures 7 and 8 show the second methodology for the Apollo 14 covered and preirradiated cells, with the abrupt change measureable for the covered but not reliably measureable for the preirradiated cell due to the digitization of each count. In Table 2, the latter method is employed for all cells except the two preirradiated cells, for which we show the former methodology, however, we adopt zero effect for this pair of cells in our analyses.

Figure 7.

Plot of best fit to astronomically normalized counts from Apollo 14 covered cell over 10 Lunar Days (almost a year) before and after the Event, showing it coincided with a small but definite degradation of 1.1 counts (Table 2).

Figure 8.

Plot of astronomically normalized counts from Apollo 14 preirradiated cell over 10 Lunar Days (almost a year) before and after the Event, showing no clear measurable degradation, which we assign a value of an increase of 0.1 from the noon before the Event to the noon after (Table 2), but for analyses select zero effect.

[19] Table 2 separates degradation in counts for each cell into two groups: (i) effects of the Event and (ii) all other effects (non-Event degradation) which possibly include dust accretion or removal, discoloration, darkening, and non-Event particle radiation discussed below. We state a caveat here that a cell count reduction caused by radiation and a cell count caused by dust are treated equally in Table 2 but were caused logarithmically for radiation [Tada et al., 1982; Curtis and Swartz, 1987] and nonlinearly with increasing surface area to volume ratio for dust particles [Katzan et al., 1991].

5 Discussions

[20] The serendipitous occurrence of the Event, the most intense SPE by an order of magnitude in Solar Cycle 20 [Hakura, 1976; Smart and Shea, 1990; King, 1974], provided invaluable benchmark information about radiation responses of each cell because of the abruptness and anomalous intensity of the Event which cannot be matched by any naturally occurring dust or other source of damage.

[21] The fluence of ubiquitous particle radiation from GCR is typically at least 2 orders of magnitude lower than that of SPE particles [National Research Council, 2008; Girish and Aranya, 2012]. Since the operational period of the DDEs occurred in the closing stages of Solar Cycle 20, we apply the finding of King [1974] that “the August 1972 fluxes of protons above 10, 30, 60, and 100 MeV constitute respectively, 69%, 84%, 84% and 83% of the fluxes … obtained by integrating over the entire solar cycle”. This statistic applied only to 25 periods of about a week for which the time integrated flux (fluence) of protons above 10 MeV exceeded 25 × 107 cm−2. With this caveat, it is therefore simplistic but not unreasonable to deduce from King [1974] that, absent the most dominating Event, the long-term fluence and therefore the long-term damage to the DDEs was about half that of the Event. This assumption is fragile, for example, because of the restriction to 25 significant events, and because the proportion of heavy ions in SPEs varies greatly [Farrell et al., 2012; Mewaldt et al., 2005; Simnett, 1976]. For these and other reasons, here we conservatively assume the total damage caused by long-term radiation is comparable to 100% of that caused by the Event.

[22] For Apollo 12, from Messenger et al.'s [1997] rule of thumb, principal cell damage is caused by protons between 20 MeV and 80 MeV, falling in the range of energies analyzed by King [1974] and significantly higher than the 1–3 MeV range where solar cells are most sensitive [Tada et al., 1982]. Therefore, we can be most confident that long-term radiation damage to the Apollo 12 cell does not exceed that caused by the Event, and consequently is most indicative of long-term dust accretion. In spite of its anomalous intensity [King, 1974], the Event damaged the Apollo 12 cell by only 0.4 counts or 0.2% of its initial output and similarly caused small or no measurable effects to all covered cells (Table 2). By contrast, all five covered cells showed significant long-term non-Event degradation (Table 2), with the Apollo 12 cell degrading by 6.3%. In fact, and despite their diversity, all five covered cells showed similar long-term non-Event degradation—6%, 9%, 10%, 8%, and 4% of their initial outputs (Table 2)—a result consistent with dust.

[23] Despite significant damage caused by the Event early in their operating life (Table 2), the Apollo 14 and Apollo 15 bare cells exhibited long-term non-Event damage of roughly 4 times greater than that caused by the Event, respectively, and several times greater than the long-term damage to the five covered cells. Unlike radiation, a particular quantity of dust accumulation affects all solar cells roughly equally; therefore, this enhanced long-term degradation of the bare cells is not realistically solely attributable to dust. Accordingly, we conclude that the largest component of long-term damage of the two bare cells was caused by low-energy radiation, say protons and heavy ions with less that about 4 MeV/nucleon. We do not attempt here a more sophisticated argument invoking the logarithmic dependence of cell damage on radiation fluences. However, the rates of long-term degradation of the seven cells, as shown in Figures 3–6, are consistent with (i) all covered cell damage being dominated by nonlinear degradation from dust, with a much smaller effect from all long-term radiation from protons > 4 MeV and (ii) bare cell damage dominated by logarithmic damage from low-energy protons of 1 to 4 MeV and comparable heavy ions, with relatively smaller effects from dust.

[24] Based on the effects of simulated MLS-1 dust particles between 20 and 38 µm and the average long-term damage to the five covered cells of roughly 7%, we present an upper limit for long-term net dust accumulation of order 100 µg cm−2 yr−1, equivalent to a layer 1 mm thick over 1000 years assuming a porosity of 33% [see Hapke and Van Horn, 1963]. This result is based on analyses to separate particle radiation effects from dust and other nonradiation effects. Contributors to long-term degradation other than radiation and dust appear less significant, e.g., darkening effects due to long-term UV radiation do not account for more than about 1% of total degradation over 600 equivalent solar days (ESD) [Matcham et al., 1998], but information is limited, as it is for discoloration. However, our key finding, being an upper limit of dust accretion, still holds.

[25] We have articulated possible weaknesses of this approach and considered alternatives and supplementary methodologies. One alternative considered was to collate measurements of the actual space radiation environment in the manner of the Apollo 17 Cosmic Ray Experiment [Walker et al., 1973] and the spacecraft IMP-G, IMP-H, IMP-I, Pioneer 9, and Vela. This method and comparisons with solar cells on other spacecraft, were rejected as impracticable for many reasons including temporary immersion of such spacecraft in the magnetosphere and therefore exposure to both geomagnetically trapped radiation and modified SPE radiation. Another alternative might have been more modern measurements near the Moon such as CRaTER, the Cosmic Ray Telescope for the Effects of Radiation instrument on the Lunar Reconnaissance Orbiter [Looper et al., 2013; Schwadron et al., 2012], but it addresses only protons with energies >10 MeV or equivalent.

[26] Our estimates of net dust accretion have a time resolution of 1 year, which puts them between timescales of geophysical active transport and geological processes. Relevant geophysical processes may include electrically charged particles actively transported by electrical fields, as suggested by various theories and astronaut observations [Criswell, 1973; Colwell et al., 2007; Stubbs et al., 2006], which would influence our net dust accretion estimate at 3 Apollo landing sites, and may not continue indefinitely on geological time-scales. Gold et al. [1970] suggested that the common absence of dust on lunar rocks revealed by astronaut observations and photographs [Schmitt, 1973; NASA Mission Reports, 1969–1973] may be associated with “a dust-transportation process over the lunar surface that has a strong tendency for downhill flow in which particles are not lifted as high (i.e., more than 5 or 10 cm) as the surface of the rocks that exhibit the clean areas”. It is important to note in this context that the Apollo DDEs studied here are 100 cm above the lunar surface (Figure 2). However, astronaut-scientist Harrison Schmitt comments that “although a brownish patina covered most exposed, in situ rock surfaces, I could identify major minerals by their color, shape, and cleavage without difficulty”. (H. Schmitt, Pers. Comm. 2013). Relevant geological processes include (i) bombardment of the Moon annually by about 109 g of interplanetary micrometeoroids of cometary and asteroidal origin, with an estimated incoming flux at the lunar surface of dust with mass > 10−10 g of order 0.01 particles cm−2 yr−1 [Grun et al., 2011], (ii) lunar ejecta from meteoroid impacts [Berg et al., 1973], returning to the Moon as one physical form of dust transport over the lunar surface. Geological processes considered by Horz and Cintala [1997] suggest a typical growth rate for a 5 m thick lunar soil layer of <10−7 cm yr−1 at present. More recently, measurements of lunar maria regolith thickness of about 6 m by Kobayashi et al. [2010], require an accretion rate of order 0.1 µg cm−2 yr−1 (B. Hapke, personal communication, 2013).

6 Conclusions

[27] Analyses of astronomically normalized measurements of seven sensors on three Apollo DDEs present with full transparency and articulated caveats the first measurements of natural causes and effects of solar cell degradation from both charged particle bombardment and dust accretion at heights of 100 cm above the lunar surface at three low and midlatitude locations. This report includes, to our knowledge, the first rigorous analysis of the effects of the anomalously intense August 1972 SPE [King, 1974] on a set of bare and covered solar cells on the Moon, revealing severe damage to two bare cells and small to immeasurable damage to five differing covered horizontal solar cells. This analysis enabled disaggregation of the effects of radiation from other lunar weather, particularly dust, leading to the first direct, active measurement of an upper limit of long-term net dust accretion (in terms of MLS-1 dust simulant particles) of 100 µg cm−2 yr−1, equivalent to 1 mm per millennium, near three Apollo landing sites. Apollo-era solar cells are available for new laboratory experiments to refine this limit for simulated dust particles other than MLS-1, such as lunar ejecta.

[28] If our upper limit is the actual accretion of dust, then radiation damage on covered, unhardened Apollo-era solar cells from both the most intense SPE of Solar Cycle 20 and over the long-term was significantly smaller than the damage caused by net dust accretion over 5 to 6 years. Future lunar expeditions relying on power from solar cells should take note that further hardening of modern solar cells against radiation damage may be less effective in safeguarding long-term efficiency than a concerted bid to manage and mitigate dust accretion, which caused a wide range of operational problems such as overheating and subsequent failure of experiments from Apollo 11 to Apollo 17 [Gaier and Jaworske, 2007].

[29] This is also the first report of the total cumulative damage caused to bare solar cells by low-energy ionizing particle radiation 100 cm above the surface of the Moon at two sites over about 5 years. Indications are that the long-term damage was (i) caused by long-neglected low-energy (e.g., protons < 4 MeV and equally penetrating heavy ions) radiation and (ii) roughly 3 times the damage recorded over one lunar day by the anomalously intense August 1972 SPE. Applications from this result include that the use and management of bare hardware and materials exposed on the lunar surface should consider such previously unrecognized long-term low-energy radiation effects.

[30] One indicative potential benefit from these studies of the August 1972 SPE is in modeling the 1 AU radiation environment relevant to determining whether long space missions outside of low-Earth orbit can be carried out with acceptable risk, such as the Earth-Moon-Mars Radiation Environment Module (EMMREM) [Schwadron et al., 2012]. Examination of sophisticated dust modeling or a submodule of EMMREM for the Moon or rocky asteroids is beyond the scope of this report.

[31] Six important scientific and operational issues are identified which could gain immediate benefits from this first measurement of an upper limit of net dust accretion: (i) to provide Ground Truth support to the 2013 lunar orbit operations of the U.S. LADEE satellite and the lunar surface operations of the rover deployed by the Chinese Chang'e 3, (ii) for assessment of whether dust is a major cause of observed degradation of Apollo LRRR; (iii) to help optimize future lunar optical experiments and second-generation LRRR units; (iv) to help optimize operational plans for thermal controls of long duration lunar missions; (v) to help inform the aerospace community that solar cell arrays for lengthy lunar expeditions now have targets for mitigation of dust and its influence on thermal controls, (vi) to add a significant new knowledge-based benchmark of naturally occurring dust accretion to give perspective to, and update, existing NASA Recommendations to Space-Faring Entities about Go-No Go areas on the Moon, with a Guidelines agreement with Google Lunar X [NASA, 2011].


[32] We acknowledge with thanks Ron Burman for contributing greatly to astronomical and other issues early in this project. We appreciate broad and academic support from the University of Western Australia, particularly from the Farrant family for the support provided by the John and Patricia Farrant Scholarship which assisted one of us (M.H.) to begin this project. Many stimulating discussions with US researchers have assisted willingly, with particular thanks to Bruce Hapke and Jack Schmitt for discussion of dust transport. Thanks also to Jim Gaier, Hal Goldwire, Doug Currie, Lou Lazerotti, Joe King, and the referees. Yosio Nakamura and Chris Holloway were of great assistance in obtaining 1990 copies of digital data in 2009, and Chris and Guy Holmes of SpectrumData are thanked for storing Apollo tapes. The unwavering support of James Boden is greatly appreciated. Facilities for data analyses are supported by Brian J. O'Brien and Associates Pty Ltd, and our continued gratitude is expressed to Avril S. O'Brien for her long support and willing help in prolonged self-funding of Apollo dust research.