The deep space galactic cosmic ray lineal energy spectrum at solar minimum


Corresponding author: A. W. Case, Harvard-Smithsonian Center for Astrophysics, 60 Garden St. MS-58, Cambridge, MA 02138, USA. (


[1] The Cosmic Ray Telescope for the Effects of Radiation (CRaTER) instrument is an energetic particle telescope on board the Lunar Reconnaissance Orbiter spacecraft. CRaTER measures energetic charged particles that have sufficient energy to penetrate the outer shielding of the instrument (about 12 MeV/nucleon). Galactic cosmic rays (GCR) with these energies are the primary radiation concern for spacecraft and astronauts outside of the Earth's magnetosphere during times of minimal solar activity. These particles can easily penetrate typical shielding and damage electronics, causing increased electronics failure rates and single event upsets. When this radiation impacts biological cells, it causes an increased risk of cancer. The CRaTER instrument was built to characterize the radiation dose and lineal energy with unprecedented time and energy resolution and was fortuitously flown during a period of time that coincided with the highest GCR fluxes in the modern space age. We report here this worst-case GCR lineal energy spectrum. Observations are made behind a thin aluminum window and different thicknesses of tissue-equivalent plastic. These measurements provide important observational data points to compare with current model predictions of the dose deposited by energetic particles within a tissue-like material.

1 Introduction

[2] Galactic cosmic rays (GCR) are highly energetic particles that are pervasive throughout the galaxy. These particles are accelerated in the explosions of supernovae [Baade and Zwicky, 1934; Koyama et al., 1995], and a small percentage of them are able to propagate via drifts and diffusion into the inner heliosphere [Parker, 1958; Florinski et al., 2003]. At 1 AU, outside of the Earth's magnetosphere, the GCR flux is nearly isotropic and spans energies from 106 to ~1020 eV. Particles with these energies are moving a significant fraction of the speed of light and are able to easily penetrate the thin shielding of most spacecraft. GCR provide an ever present radiation hazard to biologic materials and electronics in space. Solar processes such as flares and coronal mass ejections can also produce particles with energies equal to lower energy GCR, but in this paper, we will limit our discussion to a period surrounding solar minimum in which solar particles contributed a negligible flux to this particle population. Our focus is the minimum of solar cycle 23 (2009–2011) when low solar activity contributed to the space-age high flux of galactic cosmic rays [Mewaldt et al., 2010].

[3] As GCR pass through a target, a multitude of electromagnetic and nuclear interactions cause the incident particles to deposit some of their kinetic energy into the target material. The energy is deposited primarily in the form of ionization of atoms in the target. The rate at which the incident particle deposits its energy in the target is termed linear energy transfer (LET=−dE/dx, energy deposited per unit path length). The LET spectrum and its evolution through the human body are essential ingredients in understanding and mitigating the potential radiation risk posed by energetic particles [Cucinotta and Durante, 2006].

[4] Measurements of fundamental physics quantities in space are typically focused on the composition and spectra of GCR and solar energetic particles (SEP) (e.g., the ACE/Cosmic Ray Isotope Spectrometer experiment). Also, measurements of dose and dose rate are routinely performed on crewed space missions [Zhou et al., 2007]. However, neither of these types of measurements gives a comprehensive picture of LET spectra in unshielded space or at depth in tissue. This is an important distinction because the damage incurred by a material depends not only on the total dose deposited but also on the relative portion of the dose that was due to high LET events [Cucinotta et al., 2000].

[5] Typically, energetic particle transport models [e.g., Wilson et al., 1995; Agostinelli et al., 2003] in conjunction with measurements of the incident particle spectrum have been used to calculate the expected LET spectrum at various depths within a target material of interest (typically either silicon or water). The Cosmic Ray Telescope for the Effects of Radiation (CRaTER) instrument is designed to directly measure a very close approximation of LET, the lineal energy spectrum, behind different areal densities of a material. In this paper, we refer to lineal energy, “y,” rather than “LET” to indicate that we are measuring the average energy deposit over the particle's path length through our detector (ΔEx), rather than its instantaneous energy loss rate (dE/dx). See Appendix B for a description of our calculation of lineal energy. The lineal energy measurements made by CRaTER can then be used to assess the accuracy of the various LET models currently available to the community.

[6] The CRaTER instrument [Spence et al., 2010] is an energetic particle telescope on board the Lunar Reconnaissance Orbiter (LRO) [Tooley et al., 2010] in lunar orbit. During the period discussed in this paper, LRO orbited the Moon in a 50 km (average) polar orbit with a period of about 100 min. The spacecraft is three axis stabilized and spends the majority of the time with one side of the spacecraft pointed directly toward the surface of the Moon at the subspacecraft point (i.e., “nadir pointing”). The CRaTER instrument is mounted on the spacecraft in a way that allows an unimpeded view in the nadir (toward the Moon) and zenith (away from the Moon) directions. CRaTER measures the lineal energy spectrum in three pairs of detectors situated behind different areal densities of tissue-equivalent plastic (TEP). Each pair of detectors has one thin and one thick detector to maximize the dynamic range of lineal energy that can be measured. From zenith to nadir, the detectors are numbered 1 through 6 and are referred to by a capital “D” followed by their number, with the odd-numbered detectors being thin (and optimized for high lineal energies) and the even-numbered detectors being thick (and optimized for lower lineal energies). See Appendix A for a deeper discussion of the instrument design.

[7] As in the case of many cosmic ray telescopes, CRaTER does not directly measure a particle's incident energy or species, although some species can be separated statistically with sufficient observations. CRaTER has three unique capabilities: (1) CRaTER's location in deep space allows it to observe the lineal energy spectrum without influence from the Earth's magnetosphere. (2) The TEP allows CRaTER to simulate the influence of human tissue on the spectrum of particles, allowing direct measurement of particles behind biologically important areal densities of matter. (3) The extremely high data rate allows CRaTER to telemeter the full-resolution energy deposit information for every event in every detector during the time period discussed in this paper.

[8] The remainder of this paper is structured as follows. Section 2 presents the observed lineal energy spectra for each detector pair. Section 3 makes some comparisons to other fragmentation observations and discusses the implications and uses of the observations presented in this paper. In Appendix A, we discuss some relevant information regarding the instrument, and in Appendix B, we describe the methods used to prepare the higher-level data products used in this paper from the standard level 2 CRaTER data that are available from the Planetary Data System ( website.

2 Observations

[9] Here we present the lineal energy spectra due to particles that had energy deposits above threshold in the D2, D4, and D6 detectors (referred to as “triples”). These events are caused by particles that transit through the entire telescope and so must have incident energies of at least 114.5 MeV/nucleon for protons (or higher for other ions; see section 3). This is the narrowest constraint that we can put on the entrance angle and hence the most well-defined geometric factor and path length. D1, D3, and D5 (the thin detectors) are not explicitly required in this coincidence, but at high lineal energies, they respond to all of the same events as the thick detectors and can be used for lineal energies above the saturation point of the thick detectors.

[10] Figure 1 shows the spectra for all three pairs of detectors. See Appendix B for an in-depth discussion of the creation of these spectra. The data from each detector pair are shown in different colors, and the vertical dotted line shows the dividing line between the regions in which the spectrum is calculated from the thin or thick detector in that pair. Bins are combined as described in section B4 to ensure that all bins have a Poissonian uncertainty of less than 3.5%. Some additional uncertainty is contributed by the geometric factor correction (shown in Figure B3) which is negligible at low y and may be a factor of 2 at high y.

Figure 1.

Spectra for all three pairs of detectors. The left side of the vertical dotted line shows data from the thick detector in each pair, and the right side shows data from the thin detectors.

[11] The local maxima in these spectra roughly correspond to the minimum-ionizing energy of different elements. The lowest peak, at ~0.3 keV/μm, is from protons, and the highest peak, at ~300 keV/μm, is from iron. In many cases, agreement between the spectra is fairly good (e.g., around the proton peak), though at other y values, some differences can be observed from one detector pair to the next. We can take a closer look at this phenomenon by plotting the ratio of the fluxes between each pair of detectors.

[12] Figure 2 shows the ratio of the spectra in the three pairs of detectors. In all cases, the detectors that are deeper in the telescope stack are in the numerator of the ratio. Therefore, a ratio of greater than 1 indicates a higher flux of particles at that y deeper in the telescope.

Figure 2.

Ratios of the spectra shown in Figure 1. The horizontal lines denote the mean flux for each detector ratio over the range of lineal energy denoted by their width. The vertical bars indicate the standard deviation of the points in that lineal energy range.

[13] At around 1 and 7 keV/μm, there are peaks seen in the D3D4/D1D2 and D5D6/D1D2 ratios. These are regions of y space in which more events were seen deeper in the telescope. Above 10 keV/μm, the D5D6/D1D2 and D5D6/D3D4 ratios are significantly depressed due to nuclear fragmentation of the incident heavy ions.

[14] We calculated the mean and standard deviation of these points for each ratio within the region of 20 and 200 keV/μm. The mean (standard deviation) is indicated by a horizontal (vertical) bar in the region of interest.

[15] It is clear from these plots that some particles with high y at the entrance to the telescope undergo fragmentation in one of the TEP layers, resulting in lower y deeper in the telescope. We can gain some more insight into the processes of fragmentation by simultaneously plotting lineal energy in multiple detectors.

[16] Figure 3 shows a two-dimensional histogram of the CRaTER data from the same time period as the previous figures. Again, only events that caused coincident energy deposits in D2, D4, and D6 are included in this plot. In this figure, color represents the flux of particles measured in that energy bin and is plotted as a function of y in D1 and D5. The majority of events at low y fall onto a diagonal line running from the lower left to the upper right. These counts are due to events in which the incoming particle had a high enough energy to penetrate both pieces of TEP without slowing down considerably, and hence, they deposit the same amount of energy into both D1 and D5. The separate local maxima are analogous to the peaks in Figure 1 and correspond to species with different nuclear charges. Only the thin detector spectra are shown here, and hence, protons are not visible on this plot.

Figure 3.

Two-dimensional spectrum of lineal energy in D1 (top of the telescope) and D5 (bottom of the telescope). Color (log scale) indicates the flux of particles in that energy bin. Deposits falling on the diagonal line from the lower left to the upper right indicate minimum-ionizing particles that do not slow down significantly as they transit through the telescope and deposit the same amount of energy in both D1 and D5.

[17] The collection of points at y of about 250 keV/μm in D1 and nearly zero deposit in D5 corresponds to events caused by iron nuclei. The absence of many counts at this energy along the 1-to-1 diagonal indicates that the majority of the incident iron particles fragment in the TEP between D1 and D5 or are otherwise triggering a detection in D5 via scattered particles.

[18] This plot also clearly shows a competing process that tends to increase the lineal energy as the particles transit through the telescope. The lineal energy of particles depends on the particle species as well as its kinetic energy, and so at lower energies, different species separate into separate bands on the D1/D5 plot due to their different slowing-down rates as they deposit energy into the detectors and TEP. These tracks can be seen for two separate species in the lower left-hand corner of the plot. The tracks curve upward away from the 1-to-1 diagonal at lineal energies of about 20 and 40 keV/μm in D1.

3 Conclusions and Discussion

[19] We develop here the first detailed observational analysis that enables determination of lineal energy spectra from the CRaTER instrument in orbit on LRO around the Moon. Due to the restriction of our data to only those events with well-known path lengths through the detectors, we can assure that the lineal energy measured here is a very good approximation of the true LET of the particles. Along with some knowledge of the CRaTER instrument response, our observations provide the basis for detailed comparison with models of LET spectra [e.g., Looper et al., 2013; Porter et al., ; this issue], thereby elucidating the underlying physical interactions that give rise to the detailed shape of the LET spectrum. These observations represent the fulfillment of a portion of the top level LRO objectives regarding the radiation environment around the Moon. We have observed the lineal energy spectrum of the quiet time GCR population with unprecedented energy resolution over the range of energies most relevant to future space exploration.

[20] We observe significant evolution of the lineal energy spectrum as particles travel deeper into the instrument and transit through the large portions of tissue-equivalent plastic (TEP) that are contained within the telescope stack. In particular, we calculate significant reductions in the flux of particles with y greater than 20 keV/μm in the middle and end of the telescope. In the spectra shown in this study, the flux of particles with high y (above 20 keV/μm) decreases by ~20% after traversing 80.9 mm of TEP.

[21] Our observations of the fragmentation of iron nuclei within the CRaTER instrument can be compared to similar observations made during ground testing of the CRaTER engineering model (EM), a high-fidelity prototype that is nearly identical to the flight model. The EM was tested at Brookhaven National Laboratory's NASA Space Radiation Laboratory in a beam of 600 MeV/nucleon iron nuclei. Zhou et al. [2010] used CR-39 detectors placed in front of and behind the CRaTER EM in order to measure the LET and charge spectrum both upstream and downstream of the CRaTER EM. The CR-39 measured a reduction in iron events of approximately 60% from the front of the telescope to the back. Qualitatively, the data shown in Figure 3 agree with the significant reduction in iron flux after the many nuclear interactions within the TEP. Differences in these observations may be investigated in future studies through the use of other energetic ion beam tests.

[22] The requirement that all events deposit significant energy in all three thick detectors (a so-called “triple” event) implicitly imposes an energy selection that depends on the range of the incident particle. Specifically, ions that enter the telescope but do not interact must have sufficient range to reach D6. Ions that do not have sufficient range may, with some small probability, undergo a nuclear interaction and produce one or more secondary ions that do reach D6. The energy threshold for noninteracting ions to meet the range requirement is a function of the charge Z and mass A of the ion. Figure 4 shows the energy per nucleon required to reach D6 for noninteracting ions, calculated for the most abundant naturally occurring isotope of each species from hydrogen to iron. For 1H and 4He, the energy requirement is about 115 MeV/nucleon for normal incidence, and the Badhwar-ONeill model predicts that about 3%of the 1H flux and 6% of the 4He flux fall below this energy. At the high end, 56Fe ions with energies below about 525 MeV/nucleon do not reach D6, and the Badhwar-ONeill model predicts that nearly 40% of all incident GCR Fe ions will fail this criterion (though this precise number is dependent on the modulation parameter used). Thus, the requirement of deposited energy in D6 biases the reported y distribution for the zenith and middle pairs of detectors. The threshold requirement is slightly mitigated by nuclear interactions. Depending on the particle species, these nuclear interactions may change the fraction of particles that meet the threshold criterion by 1–5%.

Figure 4.

The kinetic energy needed for the most common isotope of each element (from hydrogen to iron) to transit through the entire CRaTER telescope and hence to be included in the spectra shown in this paper. The simulation does not include nuclear interactions.

[23] Some predictions of the lineal energy spectrum that would be observed by CRaTER were published before observations from the CRaTER instrument were available [Townsend et al., 2010]. We leave detailed comparisons between the observations discussed in this paper and current state-of-the-art models for other studies [e.g., Porter et al., 2013; this issue].

[24] The lineal energy spectra calculated in this paper are extremely useful in calculating and understanding the dose received in lunar orbit and by proxy in deep space. Smaller instruments that require fewer spacecraft resources can also accurately calculate the dose (see Mazur et al. [2011] for an example of one such sensor that is embedded within the CRaTER instrument) but cannot determine which parts of the lineal energy spectrum are contributing to the dose. To accurately convert from dose to dose equivalent (by multiplying by the appropriate quality factor), one must know the lineal energy spectrum of particles that contributed to the dose. Recent studies have used the CRaTER measurements to derive the dose rate over time [Schwadron et al., 2012] and have compared these measurements to dose rate modeling predictions [Joyce et al., 2013; this issue].

[25] The analysis performed here opens the door to the generation of lineal energy spectra as a legacy data set from the CRaTER instrument. While the appropriate period for generation of statistically valid lineal energy spectra is under study, the goal of periodically generated lineal energy spectra detailing the temporal evolution of all lineal energy components including protons and heavy ions from GCR and SEP is well within reach. With well-calibrated lineal energy spectra, we are able to quantify the biological effects of SEP and GCR broken down in terms of dose and dose equivalents, as well as the effects due to light versus heavy ions. Thus, we develop the critically needed observational basis for a significantly improved specification of the biological effects of the radiation environment in deep space.

Appendix A: Instrument Detail

[26] Figure A1 shows a schematic cross section of the telescope portion of the CRaTER instrument. The associated power supply and measurement electronics are not shown. The thickness of the thin detectors is not shown to scale, but the rest of the schematic is. As shown, each detector is referred to by a capital “D” followed by its numerical order in the stack, going from zenith to nadir. CRaTER consists of three pairs of silicon detectors each made up of a thin (148 μm) and a thick (1000 μm) detector. The gains of the signal processing systems were designed so that the thin detectors are optimized for high lineal energy and the thick detectors for low lineal energy particles, thereby maximizing the dynamic range. The detectors in each pair combine to respond to lineal energy values between 0.1 keV/μm and 2.2 MeV/μm. There are pieces of TEP in between adjacent detector pairs. Charged particles that pass through the instrument lose energy in the TEP so that the LET spectrum is measured behind different amounts of simulated tissue. The TEP is simply a passive material that serves to absorb energy from and fragment the particles as they transit through the instrument, allowing a measurement of the same particle after it has passed through the material representing a significant amount of tissue.

Figure A1.

Schematic of the telescope portion of the CRaTER instrument. The thick detectors (D2, D4, and D6) are 1000 μm thick, and the thin detectors (D1, D3, and D5) are 148 μm thick. Detector thicknesses are not shown to scale. The diagram shows the thick shield support structures which support the 0.8 mm thick shields.

[27] The 0.8 mm thick aluminum shields on the zenith and nadir ends of the telescope block ions with energies less than about 12 MeV/nucleon from entering the telescope. Shielding around the rest of the telescope varies in thickness and blocks out particles with energies less than about 30 MeV/nucleon. Since each solid-state detector is only a single pixel, directionality of the incoming particles is inferred from detection “coincidences” in which more than one detector is triggered by the same particle (or its daughter particles) as it travels through the telescope. For this study, we investigate only particles that have produced a signal above threshold in all three of the thick detectors (D2, D4, and D6). This restricts the nominal, geometrically derived, field of view to 31.4° (half angle), resulting in a geometric factor of 0.605 cm2sr. In a thin detector system with no TEP, this geometric factor would be accurate enough to convert from count rates to proper fluxes. However, the significant scattering and fragmentation of particles within CRaTER force us to consider more complicated LET-dependent correction factors. The requirement that all particles impact all three thick detectors implicitly imposes a species-dependent energy threshold that is discussed in section 3.

[28] Each detector is equipped with a configurable low-level energy threshold which is used as an event trigger. An event is triggered whenever a particle deposits an amount of energy above that threshold. Thus, each event consists of a measurement of the energy deposit in every detector. This energy deposit produces a current of electron-hole pairs in the detector electronics that is amplified and converted into a shaped voltage pulse, the amplitude of which is proportional to the amount of energy deposited in the detector as electron-hole pairs. This amplitude is then digitized by a 12 bit analog to digital converter (ADC). The digital value for every detector is telemetered back to the ground whether or not the energy deposit in that specific detector was above threshold. Ground processing software converts each digital value back into an energy deposit and compares each detector's energy deposit to its threshold to determine which detectors the particle passed through in that event. In this way, we identify the subset of our events for which the geometric factor is well known and convert event rates into absolute differential fluxes.

Appendix B: Data Analysis

[29] The lineal energy spectra shown in section 2 are designed to be as close to the raw CRaTER data as possible. This section will describe the minor modifications necessary to remove intricacies of the instrument that would otherwise obscure the physical quantity being measured.

B1 Time Period for Spectrum Accumulation

[30] In its nominal observing mode, the LRO spacecraft is pointed toward nadir, but for calibration and some special observation purposes, the spacecraft spends a portion of its time (less than 1%) pointing in a nonnadir direction. The CRaTER instrument is statically mounted to the spacecraft and is subject to the same pointing as the spacecraft. For the purposes of this paper, we exclude any periods of time when the spacecraft is pointed more than 1° away from the nadir direction.

[31] We accumulate data from the beginning of the nominal LRO mission (15 September 2009) up to the end of 2010. This work, which calculates the mean spectra over the entire period, implicitly assumes that changes in the incident spectrum of GCR particles are small over this time period. A further assumption in this method is that the instrument response is not changing over this time period. Frequent onboard calibrations, in which stable pulses are injected into the signal processing electronics, ensure that this is true to within better than 1%. Figure B1 shows the results of these calibrations. The deviation of the gain of each detector's measurement electronics from the mean value for that detector is shown for the time period analyzed in this paper. No single measurement of the gain deviated from the mean by more than 0.4%.

Figure B1.

Stability of the gain of the measurement electronics as a function of time.

[32] Over the time period covered in this paper, LRO/CRaTER spent >93% of its time in nominal observation mode. The bulk of the time spent in nonnominal configurations was due to calibration activities, lunar eclipses, and other activities on board the spacecraft that required CRaTER to be powered off. The total accumulation time over this period was over 517 days.

B2 Calculating Lineal Energy

[33] In this paper, we refer to lineal energy, “y,” rather than “LET” to indicate that we are measuring the average energy deposit over the particle's path length through our detector (ΔEx), rather than its instantaneous energy loss rate (dE/dx). The CRaTER instrument inherently measures the total ionizing energy that was deposited into each detector (ΔE) during an event. To calculate y for each detector in each event, one must precisely know the path taken by the particle through each detector (Δx). In the case of the CRaTER instrument, this value is not well known due to path length straggling (small deviation from a linear path due to Coulomb scattering) and deviations from normal incidence that still fall within the acceptance cone for a particular detection coincidence. All values of y reported in this paper will refer to lineal energy in silicon, the material in which the measurements are made. We also note that the lineal energies reported from CRaTER measurements differ from what many refer to as “LET” due to secondary electrons that are able to escape from the surfaces of the detector without depositing their energy into the measurement electronics.

[34] We use a Monte Carlo simulation to estimate the mean path length of particles through each detector (as specified in Sullivan [1971]). We consider an isotropic distribution of straight line particle paths and calculate the mean path length for those paths that intersect both D2 and D6. The average of these path lengths is then used to calculate y for particles satisfying those coincidence conditions. The mean path length is 1% longer than the thickness of each detector.

B3 Energy Calibration

[35] The most precise published calibration of the CRaTER detection electronics took place on the ground before launch using a combination of energetic particles from a proton beam and alpha particles from a radioactive Americium-241 source [Spence et al., 2010]. The calibration process calculates the needed coefficients to convert from analog-digital units (ADU) to the physical units of lineal energy. The relationship is linear and is described by a gain (keV/ADU) and an offset (keV). With the methods used at the time of launch, we were able to calculate the gain of the thin detectors to within about 10%, and the thick detector gains were calibrated to within about 1%. This difference in accuracy is because low-energy protons and alpha particles were used in the calibration, which produce a much smaller signal in the thin detectors than in the thick detectors.

[36] Since launch, CRaTER has accumulated over 3 years worth of data, and these data can be used to refine our calculations of the interrelationships of the calibration coefficients of the detectors. The calibration and recalibration of the CRaTER detectors will be discussed at length in an upcoming paper. For the current study, we use the previously calculated ground calibrations.

B4 ADC Nonlinearity

[37] During the process of measuring a given particle's energy deposit, it is necessary to digitize a voltage within the measurement electronics using an analog to digital converter (ADC). All ADCs suffer from some amount of differential and integral nonlinearity, and the CRaTER ADCs are no exception. The result is that some energy bins within the CRaTER energy spectra are oversampled or undersampled. This effect is particularly noticeable at higher energy deposits where the flux of particles is much lower. This region of the spectrum is also affected by statistical noise due to the lack of counts at the highest energy deposit bins. Both of these effects are counteracted by combining adjacent bins.

[38] In this analysis, the combining of bins takes place by enforcing a minimum allowable count number in each bin. For thick detectors, we use a value of 5000, and for thin detectors, we use 1000. These values were chosen by visually inspecting the spectra and balancing the decrease in statistical uncertainty with the decrease in energy resolution. For example, this process results in combined bin widths for the D2 detector that are 1 bin for y < 3 keV/μm, 10 bins at 6.5 keV/μm, and 100 bins at 17.8 keV/μm.

B5 Secondary Particles Within Instrument

[39] A correction to the spectra is needed due to secondary particle processes that occur within the telescope. Some particles that are incident from outside our nominal field of view are able to produce secondary particles at large angles and trigger a deeper detector that would not have been triggered by the primary particle. The correction can be thought of as a modification to the geometric factor for a given field of view (as determined by the choice of detectors included in the coincidence requirement). The number (and range) of secondaries is determined by the species and energy of the incident particle. For relatively lower LET ions (protons and helium), the effect of secondary particles is minimal, but for heavier ions, the needed correction can be significant.

[40] Because the instrument makes no explicit measurement of an incoming particle's direction, it is quite possible that false coincidences can occur. For example, consider an incoming particle incident on the D2 detector from outside the D2-D6 field of view that would nominally impact only D2 and D4. The interactions of this primary particle with the detectors and TEP create secondary particles (both electrons and ions) at large angles and cause an indirect signal in D6, becoming what we term a “false triple.” This paper refers to a D2+D4+D6 coincidence as a “triple” and a sixfold coincidence as a “sextuple.”

[41] One way to see this effect is by investigating the triples spectrum compared to the sextuples spectrum at high y. Nominally, any particle that makes it through all three thick detectors with a sufficiently high y should also deposit significant energy in the three thin detectors. This means that at high y values, the triples spectrum should match that of the sextuples (since false sextuples are rejected by the high thin detector thresholds). However, what we see (shown in Figure B2) is more than a factor of 3 difference between these two spectra. This figure also shows the relative agreement of the thin and thick detectors in the region of overlap.

Figure B2.

Triples and sextuples (labeled “all”) for the zenith detector pair (D1, D2).

[42] In effect, the false triples are causing the instrument's geometric factor to be larger than that if only particles within the nominal field of view are considered. A correction factor is obtained by assuming the following: (1) Below 2 keV/μm, false coincidences are negligible. (2) Above 20 keV/μm, the adjustment ratio is the ratio of the sextuples spectrum to the triples spectrum. (3) The points at 2 and 20 keV/μm are connected by a linear function. Qualitatively, these assumptions are supported by initial results from testing at the Heavy Ion Medical Accelerator in Chiba (Japan) [Murakami et al., 1997] using the CRaTER engineering model and a variety of ion species and energies. Figure B3 shows the resulting adjustment factor for each pair of detectors. The triples spectra (e.g., those shown in Figure B2) are combined into a single spectrum for each thin/thick pair, rebinned, and then multiplied by this adjustment factor.

Figure B3.

Adjustment factor for lineal energy spectra.


[43] This work is supported by NASA CRaTER contract NNG11PA03C. We wish to thank the LRO spacecraft team for their excellence in designing, building, and operating the spacecraft. We thank the team at the NASA Space Radiation Laboratory (NSRL) at Brookhaven National Laboratory for providing ion beams for instrument characterization. We also thank the staff at the HIMAC facility for providing ion beams for instrument characterization.