Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 Current Knowledge of the Radiation Environment Astronauts and spacecraft participating in missions to the Moon and Mars will be exposed to a hazardous radiation environment made up of background galactic cosmic radiation (GCR) and punctuated by brief but intense solar particle events (SPEs) (Box 2-1). Accurate and timely information about this environment is required in order to plan, design, and execute human exploration missions. This information consists of estimates or measurements of the time of occurrence, temporal evolution, and spatial distribution of the radiation, as well as the type, maxi- mum intensity, and energy spectrum of the constituent particles. Unfortunately, the prediction and forecasting of solar activity and space weather are severely hampered by incomplete understanding of how the Sun affects interÂ planetary space and the local environments of Earth, the Moon, and Mars. Scientific progress in this field, leading to accurate long-term and short-term predictions of the space radiation environment, will contribute to the role that solar and space physics scientists can play in human exploration missions. STATE OF RADIATION ENVIRONMENT KNOWLEDGE Since the Moon has no atmosphere, deep space measurements are close approximations to the lunar surface environment. Instruments on spacecraft such as the Advanced Composition Explorer (ACE), Solar and Heliospheric Observatory (SOHO), Wind, and Solar Terrestrial Relations Observatory (STEREO) provide data on the energetic- particle environment of deep space. However, these measurements do not provide insight into the surface radiation component from secondary particles created by incident energetic galactic cosmic rays and SPEs. Measurements taken from lunar orbit are closer approximations to the lunar surface environment. Limited orbital measurements have been taken by Explorer 35, Clementine, and Lunar Prospector. Some surface measurements of the lunar ra- diation environment were made during the Apollo program, but they were generally of short duration with limited resolution. Instruments will be carried on the Lunar Reconnaissance Orbiter mission in late 2008 that will provide additional data on the neutron and charged-particle spectra at the Moon. See Box 2-2 for additional details. There have been no direct measurements of the radiation environment made on the surface of Mars. Unlike the airless Moon, the thin atmosphere of Mars will stop all protons and alphas with energy less than ~120 MeV per nucleon, and high-Z GCR particles with energy less than about ~0.2-0.6 GeV per nucleon, depending on the species. However, some secondary particles, especially neutrons, will make it to the martian surface. Very energetic protons and galactic cosmic rays will also reach the surface and create secondary particles from collisions with the martian regolith. To model the martian surface environment, one must use transport codes, as has been done by NASA and other researchers. 19
20 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION BOX 2-1 Main Characteristics of Space Radiation Solar Particle Events â¢ Composed largely of protons, generally with low to medium energies (tens to a few hundred MeV per nucleon). â¢ More likely at solar maximum. Onset, duration, dose rate, and dose are at present unpredictable. â¢ With adequate warning and access to shelter (>10 g/cm2 aluminum-equivalent), radiation hazard can be reduced to acceptable levels. â¢ Biological effects are similar to those from x-rays or gamma rays. â¢ Major research questions involve the prediction of onset and relevant characteristics and the health risks due to the residual low dose rates when under shielding. Galactic Cosmic Rays â¢ Composed of protons, alpha particles, and heavy ions, up to very high energies exceeding tens of GeV per nucleon. â¢ Steady background varying over the 11-year cycle roughly by a factor of 2. â¢ Shielding is ineffective because ions penetrate hundreds of centimeters of material and produce secondary radiation. â¢ Biological effects are poorly understood, with large uncertainties in projections because there are no human data on which to base estimates. â¢ Major research questions involve solar cycle variations and the need for understanding of mechanisms linking r Â adiation exposure to health risk. There have been several solar system measurements of the GCR at distances at and beyond Mars, providing confidence in extrapolations of more detailed GCR measures taken near Earth. There have also been limited radia- tion measurements taken from satellites in orbit around Mars (relevant instruments on Mars Odyssey include the Mars Radiation Environment Experiment [MARIE], the High Energy Neutron Detector, and neutron information from the Gamma Ray Spectrometer; there have also been indirect measurements using instruments on Mars Global Surveyor). Use of these data for estimates of the surface radiation environment requires transport code estimates of particles created in or scattered from the martian atmosphere and calculations of particle transport from the surface through the atmosphere. The Radiation Assessment Detector (RAD) on the 2009 Mars Science Laboratory will characterize the broad spectrum of radiation at the martian surface. RAD will measure high-energy charged particles coming through the martian atmosphere. In addition to identifying neutrons, gamma rays, protons, and alpha particles, RAD will measure galactic cosmic ray heavy ions up to iron on the periodic table. Finding 2-1. Current knowledge of the radiation environment on the Moon. Data from many satellites have enabled the characterization of GCR and SPEs near Earth, and these results serve to characterize the radiation inci- dent on the surface of the Moon. Knowledge of the secondary radiation, which is produced by galactic Âcosmic rays and SPEs interacting with material on the lunar surface, is currently based on data from Apollo, Lunar ÂProspector, and Clementine and on calculations. Finding 2-2. Current knowledge of the radiation environment on Mars. The radial extrapolation of the GCR environment from Earth to Mars is well understood, based on measurements made by numerous scientific satellites as they traveled outward through the solar system. To within a few percent or so, the GCR environment at the top of the martian atmosphere is expected to be the same as that near Earth. There are very few simultaneous measure- ments of SPEs at Earth and at Mars, and current models are inadequate to extrapolate near-Earth measurements of SPEs to Mars. Knowledge of the secondary radiation environment on the surface of Mars is currently based on calculations and measurements taken by spacecraft in Mars orbit.
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 21 BOX 2-2 Radiation Measurements on or near the Lunar Surface â¢ Explorer 35 (1967-1973) had a combination of an ionization chamber and Geiger counters. The ionization chamber responded to electrons above 0.7 MeV and protons above 12 MeV. One of the Geiger counters was used for low- energy electrons. The second responded to electrons and protons above 22 keV and 300 keV, respectively. â¢ Clementine was launched in January 1994 and orbited the Moon between February and April 1994. Among its instruments were a charged particle telescope and solid-state dosimeters. The charged particle telescope on Cle- mentine measured the flux and spectra of energetic protons (3 MeV to 80 MeV) and electrons (25 keV to 500 keV). The dosimeters were proton-sensitive static random access memory chips sensitive to protons with energies from a few to more than 20 MeV. â¢ Lunar Prospector collected data in lunar orbit from January 1998 until July 1999. It contained a gamma ray spec- trometer (with a fast neutron spectrometer with sensitivity to albedo neutrons with energy up to 8 MeV), a neutron spectrometer (with sensitivity to neutrons with energy less than 1 keV), and an alpha particle spectrometer (detect- ing alpha particle decay products with energy of a few MeV). Each instrument was optimized to provide information about the lunar surface composition, not the lunar surface radiation environment. â¢ Apollo Measurements The most relevant of the Apollo measurements were made with the Cosmic Ray Detector Experiment, a set of passive glass detectors with sensitivity from 100 keV to 150 MeV per nucleon. The exposure time was limited. Other indirect surface measurements during the Apollo program included the following: â The filter glass of Surveyor III (brought back by Apollo 12) was analyzed; â The window of the Apollo 12 spacecraft was analyzed for cosmic ray tracks; â One helmet on Apollo 8 and three worn on Apollo 12 were used as heavy-particle dosimeters; â A control helmet was also exposed to cosmic rays at a balloon altitude of 41 km; and â Lunar regolith was analyzed for upper limits on high-energy exposure over the half-lives of long-lived isotopes (thousands to millions of years). â¢ The Lunar Reconnaissance Orbiter, scheduled for launch in late 2008, will have two radiation monitoring instru- ments: Cosmic Ray Telescope for the Effects of Radiation (CRaTER) and Low Energy Neutron Detector (LEND). The CRaTER telescope consists of three ion-implanted silicon detectors separated by two pieces of tissue-equivalent plastic. It will be sensitive to protons with energy above 20 MeV and energetic high-Z particles (iron nuclei with en- ergy greater than 90 MeV per nucleon, for example). The LEND instrument will provide data similar to the Neutron Spectrometer on Lunar Prospector. Galactic Cosmic Radiation Interplanetary space is bathed by a low flux (particles per square centimeter per second or particles per square centimeter per steradian per second) of essentially uniformly distributed, highly energetic, and extremely penetrat- ing ions that are believed to be accelerated by supernova shocks in our Galaxy. These ions make up the GCR. The highest-intensity GCR is found between a few tenths and a few tens of GeV per nucleon, where the particles can penetrate tens to hundreds of centimeters of shielding. Every naturally occurring element in the periodic table is present in the GCR: nearly 90 percent are protons (hydrogen), close to 10 percent are helium, and the remaining percentage are elements heavier than helium, with a relative abundance roughly similar to that found in our solar system (Figure 2-1). These subtle compositional differences were a key factor in understanding the origin of GCR and provided the original impetus for the development of heavy-ion transport codes and heavy-ion cross-section libraries that we take for granted today. Galactic cosmic rays also include electrons and positrons, but their intensi- ties are too low to be of practical concern. The GCR flux outside the solar system is presumed to be constant, at least on timescales of tens of millions of years related to the solar systemâs motion through the Galaxy (Lieberman and Melott, 2007) and barring a nearby âFluxâ is an older, but not preferred term, which has been replaced by âfluence rate.â However, since âfluxâ is more common within the â space weather and radiation protection communities, it is used in this report.
22 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION H 1.00E+06 He Galactic Cosmic Rays Solar System 1.00E+04 C O Ne Mg Si Fe 1.00E+02 S N Ca Ti Cr Be Ar Ni Na Al 1.00E+00 P Cl K Mn Co F 1.00E-02 V B Sc 1.00E-04 Li 1.00E-06 FIGURE 2-1â Relative abundance of elements in galactic cosmic rays and in the solar system. The very large GCR excesses at Li-Be-B and in the sub-Fe elements (Sc-Mn) are the products of spallation interactions during the cosmic raysâ journey through the interstellar medium. SOURCE: Based on data from Astrophysics Science Division, NASA Goddard Space Flight Center Web site, available at http://helios.gsfc.nasa.gov/ace/abund_plot.html; http://edmall.gsfc.nasa.gov/99invest.Site/ACE/plot.html. R01155, Figure 2-1 supernova (Pohl and Esposito, 1998). However, to reach Earth, GCR must penetrate the heliosphere, the magnetic plasma that surrounds the Sun, which suppresses the entry of charged particles from the interplanetary space. The strength of the interplanetary magnetic field increases with proximity to the Sun. As a result, the intensity of the GCR flux is lower around the inner planets than it is in the outer heliosphere. The interplanetary magnetic field v Â aries with the solar activity cycle; the GCR flux near Earth is at a peak near solar minimum (when the interÂ planetary magnetic field is weak) and at its low point at solar maximum (when the interplanetary magnetic field is strongest) (Cane et al., 1999). This solar-cycle variation in the strength of the interplanetary magnetic field is most likely due to the changing rate of coronal mass ejections (CMEs) (Cliver and Ling, 2001; Owens and Crooker, 2006). The higher rate of CMEs at solar maximum may also impede cosmic-ray access to the inner heliosphere by increasing the level of magnetic turbulence (Newkirk et al., 1981; Bieber et al., 1993) and through coalescence into large magnetic structures in the outer heliosphere (Burlaga et al., 1993; McDonald, 1998; Lara et al., 2005). The temporal changes in the GCR intensity near Earth varies with the energy of the GCR particles and with the solar maximum in an understood and approximately predictable way, given accurate forecasts of the solar cycle. In the energy range of less than a few GeV per nucleon, the flux decreases from solar minimum to solar maximum by 30 to 50 percent (Figure 2-2). The solar-cycle variation in the GCR intensity near Earth is illustrated in FigureÂ 2-3, which shows the observed count-rate at the Climax Neutron Monitor (NM) for 1951-2006. The Climax NM rate primarily reflects the flux of protons with energies above ~2 GeV. It is the empirical proxy employed to specify the level of solar modulation in Badhwar and OâNeillâs GCR model (OâNeill, 2007); alternatively, Nymmik uses
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 23 1.0E+01 1.0E+00 1.0E-01 Flux, ions/m2-sr-sec-MeV/n 1.0E-02 1.0E-03 1.0E-04 1.0E-05 H 1.0E-06 Differential energy spectra for major ions He 1.0E-07 Solar minimum ( =450 MV, solid line) O Solar maximum ( =1800 MV, solid line) Fe 1.0E-08 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 1.0E+06 Energy, MeV/n FIGURE 2-2â Differential solar energy spectra at solar minimum and solar maximum. SOURCE: Provided by P. OâNeill, NASA. Generated by data from the Badhwar-OâNeill model (OâNeill, 2007), which incorporates data from the Advanced Composition Explorer satellite, as well as older satellite and balloon data. R01155, Figure 2-2 27-day-averaged counts per hour Year FIGURE 2-3â Solar cycle variation in the count-rate at the Climax Neutron Monitor, 1951-2006. The count-rates have been averaged over the 27-day solar-rotation period. Dates with significant contributions from solar particle events have been e Â xcluded from the averages. SOURCE: Based on data provided by the Neutron Monitor Datasets Web site of the University R01155, Figure 2-3 of New Hampshire, National Science Foundation Grant ATM-0339527, available at http://ulysses.sr.unh.edu/NeutronMonitor/ neutron_mon.html. Fixed image, not changeable
24 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION the smoothed sunspot numbers over the preceding 12 months (Nymmik et al., 1992, 1996). The advantage of the former method is that the NM count-rate is the direct product of GCR interactions on the atmosphere (unless there is a large, high-energy solar particle event in progress). The latter method may have some particular attractions, in that sunspot numbers are a readily available descriptor of solar activity that the solar community attempts to forecast, albeit with limited success. However, McCracken and McDonald (2004) indicate that the long-term varia- tion in GCR intensity, as reflected by ice core data, is poorly correlated to sunspot numbers during some periods of low solar activity prior to the space age. Figure 2-4 shows the ratio of GCR dose rates from these two models versus depth of shielding. For solar minimum, the Badhwar-OâNeill results are systematically higher by about 10 percent; at solar maximum, they are systematically lower by about 20 percent. The energy dependence in the ratios may be due, at least in part, to differences in the transport codes. The two models were developed using many of the same GCR datasets, except that the Badhwar-OâNeill model also incorporated recent data from ACE. It is therefore not surprising that the results agree relatively well. These remaining discrepancies serve to quantify the level of systematic uncertainty associated with the details of each modelâs choices in parameterizing the GCR data. For particular year-dates (rather than just the extremes of the solar cycle, as compared in Figure 2-4), the discrepancies between the two models may be somewhat larger. The discrepancies illustrated in Figure 2-4 are small compared with those that arise from uncertainty in estima- tions of the biological effects of GCR. One may therefore conclude that the GCR component of the interplanetary radiation environment is known sufficiently well to support the needs of the Exploration missions, at least within the time frame of lunar missions. However, two caveats should be noted. FIGURE 2-4â Ratio of dose rates calculated with the Badhwar-OâNeill model (OâNeill, 2007) and with the Nymmik et al. (1992) GCR model, versus depth of shielding. Ratios are shown for both solar minimum (top) and solar maximum. NOTE: Both of these codes output the dose in silicon, which is related to the radiation dose in electronics. The ratio of doses in water (a b Â etter approximation for biology) would be similar. SOURCE: Badhwar-OâNeill dose rates for this comparison were provided for this report by P.M. OâNeill ofR01155,Johnson Space Center; the Nymmik et al. (1992) rates were calculated with the the NASA Figure 2-4 CREME96 model, https://creme96.nrl.navy.mil/ and Tylka et al., 1997. fixed image, not changeable
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 25 First, the GCR modulation actually follows a 22-year cycle, related not only to the 11-year sunspot cycle but also to changes in the polarity of the solar magnetic field. ACE, which can, in principle, continue to report GCR measurements until 2025 (Withbroe et al., 2006), is only now beginning its observations of the second half of this 22-year cycle. ACE measures GCR with unprecedented precision. It will therefore be beneficial to see if the reported accuracy of the Badhwar-OâNeill and other GCR models continues over the coming decade. Second, all of these statements about the reliability of the GCR models are predicated on our experience during the space age. But, as discussed later in this chapter, there are indications from polar ice cores that solar modulation levels were lower and GCR fluxes concomitantly higher for extended periods of time during the past 1,150 years (McCracken and McDonald, 2004). It is impossible to say how reliable the GCR models might be if it were to become necessary to extrapolate them into this unknown regime. Finding 2-3. Lunar GCR environment. Given the far larger uncertainties in biological effects, the committee finds that knowledge of the composition, energy spectrum, and temporal variation of the âfree spaceâ GCR component of the interplanetary radiation environment is sufficient to support the needs of the Constellation lunar missions. Nevertheless, it will be useful to benchmark GCR models against measurements reported by ACE in the upcoming second half of the 22-year GCR modulation cycle. Because the GCR flux inversely follows the solar cycle (peaking at solar minimum, lowest at solar maximum) and because the rate of solar events is reduced at solar minimum, a significant gap in long-term forecasts of the radiation environment is in the area of forecasting the solar cycle. The current challenge in this area is illustrated by a recent effort by a National Oceanic and Atmospheric Administration (NOAA)/NASA panel to forecast the next solar cycle, Cycle 24 (Figure 2-5). The expert panel assembled to make the forecast noted that the end of Cycle 23 is later than anticipated by up to a year, and it could not reach consensus on whether the next solar cycle would be larger or smaller than an âaverageâ solar cycle. This further translated into uncertainty over the timing of the next solar maximum, with the panel agreeing that the next solar maximum would occur between October 2011 (for a cycle larger than average) and August 2012 (for a cycle smaller than average). SOLAR PARTICLE EVENTS Energetic particles, occasionally with energies exceeding several GeV, are accelerated in sporadic events at the Sun associated with solar activity. These energetic particles are produced by processes whose details are still being studied. SPEs occur intermittently throughout the solar cycle, although much less frequently near sunspot minimum. At the present time, the occurrence and intensity of SPEs cannot be predicted. In addition to the particles themselves, signatures of SPEs also include significant increases in solar radio emissions, x-rays, and, occasionally, detectable levels of gamma rays and neutrons from the Sun. A large body of research in the 1980s led to the classification of SPEs into two types, âgradualâ and âimpul- siveâ (Reames, 1990; Cliver and Cane, 2002; Tylka and Lee, 2006). These terms are now generally understood as shorthand for two distinctive particle-acceleration mechanisms. In gradual SPEs, which have large intensities at energies relevant to astronaut radiation safety, shocks driven by fast CMEs are the dominant accelerator. The particle acceleration in impulsive SPEs, however, is believed to be due to magnetic reconnection processes, similar to those that go on in solar flares. Compared with gradual SPEs, impulsive SPEs are characterized by small intensities, short durations, low energies that do not penetrate typical shielding, and observability only over a narrow range of solar longitudes that are well connected to the spacecraft. Impulsive SPEs are also characterized by distinctive patterns of enhancements in heavy ions (Reames, 2000a; Reames and Ng, 2004; Mason et al., 2004). Impulsive SPEs are unimportant to astronaut safety because of their low particle fluxes. However, impulsive SPEs contribute indirectly to the SPE radiation hazard by enriching the suprathermal seed population from which CME-driven shocks accelerate particles to high energies. All further discussion of SPEs in this report refers to gradual SPEs. See NOAA Space Environment Center, Solar Cycle 24 Prediction, available at http://www.sec.noaa.gov/SolarCycle/SC24/index.html. â
26 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION FIGURE 2-5â National Oceanic and Atmospheric Administration/NASA Solar Cycle 24 prediction panel results, showing two potential outcomes for the next solar maximum (data through March 31, 2007). SOURCE: NOAA Space Environment Center, âSolar Cycle 24 Prediction,â available at http://www.sec.noaa.gov/SolarCycle/SC24/index.html. Dramatic increases in the intensity of penetrating particles (with ranges of millimeters up to tens of centimeters) can begin within minutes to tens of minutes of the onset of solar activity. During these early minutes, the particle flux is generally âanisotropic,â Figure 2-5 more particles come from one direction than from another. The peak R01155, meaning that direction is not necessarily toward the Sun but providedlies caption direction of the interplanetary magnetic Downloaded from URL generally in along the field, which varies. The flux becomes essentially isotropic (with no preferred direction) within tens of minutes to Fixed image, not changeable hours, depending on particle energy. Peak flux may occur minutes to days after onset, also generally depending on energy. The flux can be quite large throughout the event, although the flux at energies above several hundred MeV is generally not a significant contribution to the total fluence (flux integrated over the event). Figure 2-6 shows examples of the time evolution of some very large SPEs from Cycle 23. SPEs typically persist for hours to days, depending on energy. The observed time profiles and energies produced by the CME- driven shock are determined by the evolving nature of the shock, the sweep of the observerâs magnetic connection point across the shock front as the shock moves outward from the Sun, and the properties of the interplanetary medium through which the shock and the energetic particles propagate. Many of the largest SPEs are part of multi-event episodes, produced as a single solar active region rotates across the face of the Sun. These episodes have the potential to constrain operations for many successive days. Although high-energy particles are generally produced when the CME-driven shock is still far from Earth, some SPEs have a substantial secondary peak when the CME-driven shock passes over the Earth (typically 18 to 30 hours after event onset). These secondary increases
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 27 FIGURE 2-6â Examples of solar proton timelines from the Geostationary Operational Environmental Satellite (GOES) at >30Â MeV and >100 MeV. Panels (a) and (b) show multi-event episodes. Panels (c) and (d) show single events. The second increase in panel (d) is an example of an energetic storm particle (ESP) event, associated with the arrival of a powerful coronal mass ejection-driven shock at Earth. SOURCE: National Geophysical Data Center. R01155, Figure 2-6 fixed image, not changeable are often referred to as energetic storm particle (ESP) events. In most cases, ESP events are seen only at energies that do not pose a radiation hazard. But a few times per solar cycle, the shockâs arrival at Earth also brings very large fluxes at very high energies, extending beyond ~100 MeV. These rare, powerful events are the most severe transient radiation environment to which Exploration astronauts may be exposed. Energy Spectra The energy distribution also varies substantially from event to event. In general, the most significant energy range is from a few tens of MeV to a few hundred MeV. The drop in the energy spectrum is an important feature. âSoftâ events have a larger proportion of particles with lower energy. âHardâ events have more than the average proportion of high-energy particles. Figure 2-7 shows the proton energy spectrum from twoÂ events illustrating the
28 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION FIGURE 2-7â Examples of âsoftâ (1998 April 20) and âhardâ (1989 September 29) solar proton spectra. SOURCE: Reames, 2000b. Copyright 2000; used with permission from the American Institute of Physics. R01155, Figure 2-7 Fixed image, not changeable difference between hard and soft spectra.Â Both SPEs wereÂ associated with fast CMEs (~1,800 km/s) on the Sunâs west limb. Behind 10 g/cm2 aluminum shielding, the softÂ event (April 1998)Â would contribute approximately 0.02Â cGy per hour, while the hard event (September 1989) dose-equivalentÂ rate would have been 1.0 cGy per hour (neglecting body self-shielding and secondary neutrons produced in the shielding). Composition On average, protons comprise more than 90 percent of the energetic ions produced in an SPE. For this and other reasons, protons are the primary concern when evaluating potential SPE radiation hazards. However, the processes that accelerate protons to high energies also accelerate heavier ions. Moreover, the relative abundances of the various heavy-ion species vary significantly from event to event, as well as with energy and with time during an event. These abundance variations have proven to be powerful probes of the acceleration and transport processes by which SPEs are produced. As an example of this variability, Figure 2-8 shows the energy-dependence of the event-integrated Fe/O (iron/oxygen) ratio in two very large SPEs associated with ostensibly similar flares and CMEs. Whereas the two events are nearly identical out to ~10 MeV per nucleon, at the highest measured energies the Fe/O ratios differ by nearly two orders of magnitude. Various hypotheses have been put forward to explain the behavior in FigureÂ 2-8. However, analysis and modeling of observations like these are contributing to an emerging consensus on the way in which flares and CMEs contribute to the production of large SPEs: whereas CME-driven shocks are the ultimate energy source for potentially hazardous solar energetic particles, the reconnection processes associated with flares provide a distinctive contribution to the seed particles that are promoted to high energies by the action of the shock (Mason et al., 1999; Desai et al., 2003, 2006; Tylka et al., 2001, 2005; Lee, 2007). Strong correlations have also been found between heavy-ion characteristics and spectral shapes (such as those of Figure 2-7), and recent work
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 29 FIGURE 2-8â Event-integrated Fe/O ratio versus kinetic energy in two very large solar particle events. The data are normalized to the nominal coronal value given by Reames (1995). Different symbols distinguish measurements from various instruments on the Advanced Composition Explorer and Wind. Coronal mass ejection (CME) speed, longitude of the source of the flare, and x-ray/optical flare size are also noted for each event. The source longitude is the angular distance from a standard Âmeridian R01155, Figure 2-8 (0Â degrees heliographic longitude), measured from east to west (0 degrees to 360 degrees) along the Sunâs equator. Optical flares in H-alpha are usually accompanied by radio and image, not changeablehigh-energy particle emissions. The optical Fixed x-ray bursts and occasionally by brightness and size of the flare are indicated by a two-character code called âimportance.â The first character, a number from 1 to 4, indicates the apparent area; the second character indicates relative brilliance: B for bright, N for normal and F for faint. (See http://www.ngdc.noaa.gov/stp/SOLAR/ftpsolarflares and http://www.spaceweather.com/glossary/Âflareclasses_optical.html.) X-ray flares are classified in three categories: X-class flares are big; they are major events that can trigger planetwide radio blackouts and long-lasting radiation storms. M-class flares are medium-sized; they can cause brief radio blackouts that affect Earthâs polar regions. Minor radiation storms sometimes follow an M-class flare. Compared to X- and M-class events, C-class flares are small, with few noticeable consequences here on Earth. (See http://www.spaceweather.com/glossary/flareclasses. html.) SOURCE: Tylka et al., 2006. suggests that these correlations can be understood in terms of the evolving nature of the CME-driven shock as it moves outward from the Sun (Tylka and Lee, 2006; Sandroos and Vainio, 2007). As a result, insights derived from heavy ions show promise for explaining and eventually modeling the event-to-event variability that lies behind the engineering and operational challenges of the SPE radiation hazard. Because of the higher rate of energy deposition of heavy ions when traversing matter, it is important to assess whether solar heavy ions might pose a significant radiation hazard in themselves. As a starting point for this dis- cussion, it should be remembered that heavier ions must have higher initial energies in order to penetrate a given depth of shielding. For example, to penetrate 1.0 g/cm2 of aluminum, protons and iron nuclei must have energies of ~30 MeV and ~100 MeV per nucleon, respectively. Given that SPE spectra at the skin of the spacecraft gener- ally fall steeply with increasing energy, the higher-penetration thresholds go a long way in suppressing the dose from solar heavy ions. For a few SPEs, heavy-ion spectra have been measured out to well beyond minimal penetration energies. Tylka and Dietrich (1999) reported measurements of solar heavy ions out to nearly 1 GeV per nucleon. Kim et al.
30 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION (1999) used these results in an analysis of proton and heavy-ion contributions to astronaut radiation exposure in the September 29, 1989, event. They found that for most of the organs and shielding thicknesses that they consid- ered, the total dose equivalents in this event would have exceeded the 30-day exposure limits from NCRP Report No. 98 (NCRP, 1989). In some cases alphas contributed as much as 10 to 40 percent of the dose equivalent. Ions heavier than alphas contributed no more than a few percent to the total dose equivalent. These calculations suggest that solar heavy ions, apart perhaps from alphas, do not make a significant contribu- tion to the SPE radiation hazard for astronauts. However, it should be emphasized that the biological effect of HZE (high atomic number and energy) particles is at present not well understood (see Chapter 3).Â As this understanding evolves, it may become necessary to revisit this conclusion about the potential relevance of solar heavy ions. If solar heavy ions are found to have a potential biological impact, their time structure should also be con- sidered. At energies relevant to a given depth of shielding, the time-intensity profiles for protons can rise much more slowly that that of heavier ions, especially iron. In the event considered by Kim et al. (1999), the relevant proton fluence built up gradually over the course of more than a day, whereas about half of the relevant Fe fluence arrived in just the first few hours of the event. Thus, if solar heavy ions pose a radiation hazard to astronauts, that hazard will generally occur in the first few hours of the eventï£§just at the time when astronauts might be caught âoutsideâ and under minimal shielding. Finally, even if solar heavy ions are a negligible radiation hazard for astronauts, it should be noted that they can have significant impacts on spacecraft electronics. This topic is discussed more thoroughly in Chapter 3. Frequency of Events The NOAA Space Environment Center (SEC) declares that an SPE is underway when the number of protons with energy greater than 10 MeV exceeds 10 cm2 per second per steradian (4Ï steradians is a full sphere). Dur- ing most of the solar cycle there are roughly 10 SPEs per year meeting this criterion. Each event will require the attention of those responsible for astronaut radiation hazards. Some will turn out to be too low in peak flux or total fluence to be an issue for the astronauts. Most will be large enough to have some impact on concurrent lunar missions: from delaying or shortening an EVA up to aborting the mission. One to three times per solar cycle there will be an extremely large event (top 5 percent in peak flux or total fluence). Catalogs of SPEs are tabulated on the basis of either peak flux or event-integrated fluence above some relevant energy thresholds. Both peak flux and fluence are needed to assess radiation hazards for spacecraft systems. For human radiation exposure, however, fluence is generally the relevant quantity. Figure 2-9 shows the frequency of events for recent solar cycles organized by the fluence of all particles with energy greater than 30 MeV. Near-Term Forecasts and Nowcasting There are two classes of SPE forecast tools currently available to support human spaceflight: near-term and ânowcast.â Near-term forecast tools provide estimates of the probability of an event within the next few hours to days. Nowcasts attempt to forecast the energy spectrum and time evolution of the expected flux. Of these, the time to maximum flux and the event duration are of greatest importance for an ongoing mission, because they are used to project how quickly and for how long astronauts must take shelter. The energy spectrum is needed for depth- dose calculations. Finally, the total expected dose is proportional to the total event fluence, given by the integral of the flux over the event duration. The NOAA SEC and the United States Air Force (USAF) are responsible for providing space weather fore- casts. Each center has developed operational software to prepare SPE forecasts and alerts. The USAFâs projections serve specialized Department of Defense (DOD) needs, but its forecasts are also available to NASA through the NOAA SEC. The NOAA SEC provides SPE forecasts to U.S. domestic civil users, including the Space Radiation Analysis Group (SRAG) at the NASA Johnson Space Center, and the International Space Environment Service. NOAA also directly supports the NASA SRAG with Geostationary Operational Environmental Satellite (GOES) measurements. The NASA SRAG combines the GOES data, NOAA forecasts, and operational models of Earthâs magnetic field to predict doses inside the space shuttle and the International Space Station (ISS).
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 31 120% 100% 80% Cycles 19-23 Cumulative % Cycle 19 60% Cycle 20 Cycle 21 Cycle 22 40% Cycle 23 20% 0% 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 1.E+10 Size of Event, protons/cm2 (â¥Ð¤30) FIGURE 2-9â Cumulative distribution of solar particle events. SOURCE: F. Cucinotta, NASA, âRadiation Risk Assessments for Lunar Missions-Shielding Evaluation Criteria,â presentation to the committee December 12, 2006. R01155, Figure 2-9 To print in color When no particle event is underway and when there are no significant x-ray events, neither the SEC nor the USAF has a quantitative model or algorithm to produce multiple-day forecasts for SPEs. NOAA provides a 1- to 3-day probability estimate of energetic-particle activity (along with estimates of the probability of significant x-ray flares and other geomagnetic activity). The particle event forecast is based on qualitative observations of ongoing solar activity and solar active regions and their history. Long-lived active regions occasionally have a pattern of multiple eruptions. NOAA may upgrade the potential for a particle event when a region that produced a particle event on a previous solar rotation (with a period of about 27 days) returns. The lack of multiple-day prediction severely limits risk management alternatives for human spaceflight. In effect, a crew on a spacecraft or a team of astronauts at a remote site on the Moon must react to an event after it begins, limiting the time and options for evasive actions, if warranted, to a few hours. Even after an SPE has been observed to be underway, forecasts of peak flux are good only to within an order of magnitude or worse. The two operational models in use are the NOAA PROTONS model and the USAFâs Proton Prediction System (PPS). Both of these empirical models were developed before the central role of fast CMEs in producing large solar energetic particles was widely recognized. These models predict and characterize SPEs starting from soft x-ray and microwave proxies, as well as optical information on the location of the active region. The use of these proxies contributes to the limited reliability of these models. Nevertheless, these models continue to be essential simply because real-time CME observations are not routinely available. The NOAA PROTONS model predicts the probability of an SPE exceeding its threshold criterion (peak intensity >10 protons/cm 2 per steradian at >10 MeV), and the time of occurrence and the value of the peak intensity. The USAF PPS predicts time, intensity, and spectra.
32 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION Statistical Analyses of SPEs There are now more than 40 years of solar particle event observations outside the mitigating influence of Earthâs magnetic field. These data have been used to compile several statistical models of SPE frequency and intensities, for example in works by Feynman et al. (1993), Xapsos et al. (1998, 2000), and Nymmik (1999). These models are not meant to predict or characterize individual events. Instead, they are very useful in spacecraft design for approximating the number and type of SPEs that may be encountered during a given mission. Various versions of these trend analyses address peak proton flux and total proton fluence, but they do not adequately capture statistics on the spectral hardness of events, the probability of correlated successive events, or the distribution of âall-clearâ periodsï£§consecutive days with no solar event underway. SPACE RADIATION CLIMATOLOGY Nearly all of our knowledge of the space radiation environment comes from the past 50 years of the space age. All of our experience and ârules of thumbâ for operating in this environmentï£§from the slow, nearly constant rate of single-event upsets caused by GCRs, to lifetimes of solar panels that are continually degraded by the Âaccumulated fluence of solar protons, to the probability and duration of all-clear periodsï£§are based on this era. It is tempting to assume that the experience of the past 50 years is typical of the space radiation climate. But there is one datasetï£§albeit more indirect than the satellite measurements of the space ageï£§that offers a longer-term perspective: ice cores, which trap the products of reactions of GCR (McCracken and McDonald, 2004) and solar energetic particles (McCracken et al., 2001a,b) with constituents of Earthâs atmosphere. Painstaking measurements and analysis of this unique archive have pushed the knowledge of the space radiation climate back more than 1,000 years. This long-term record offers three cautionary notes: (1) the experience of the space age is not typical, (2) there were periods in the past when the space radiation environment was significantly harsher and thus would have presented a much more challenging environment for the design and operation of Exploration systems, and (3) significant changes in the space radiation climate can occur on the timescale of decades, the same timescale envisioned for the Exploration initiative. Ice-Core Results on Galactic Cosmic Radiation To quantify the potential implications of changes in the level of GCR modulation, Figure 2-9 shows dose-depth calculations from four different runs of the Badhwar-OâNeill GCR model (OâNeill, 2007). In two of the runs, the modulation parameter was set at values typical of solar maximum and solar minimum in the space age. In the third run, the modulation parameter was reduced to 20 percent of the solar-minimum value, so as to correspond roughly to the highest GCR intensities inferred from the ice-core studies (McCracken et al., 2004). Finally, in the fourth run, the estimated local interstellar GCR spectrum (LIS), without any modulation whatsoever, was used. This run provides an upper limit on how severe the GCR exposure might become (barring a nearby supernova!): Figure 2-10 suggests that the dose rate from GCRsï£§even under substantial shieldingï£§could be ~2 to ~8 times more severe than current expectations if the GCR level were to rise to the levels indicated by the ice cores. Ice-Core Results on Solar Particle Events McCracken et al. (2001a,b; see also Shea et al., 2006) have presented a comprehensive study of very large SPEs in ice-core nitrates since 1561 AD. Among the conclusions from their study: â¢ In the past 450 years, there have been SPEs in which the >30 MeV proton fluence was more than twice as large as the value given by King (1974) for the August 1972 event. (NASA has chosen the King description of the August 1972 event as the design standard for SPE radiation protection for astronauts.) â¢ The rate of such very large SPEs has been six to eight times larger in the past than what has been observed in the space age.
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 33 FIGURE 2-10â Dose rate versus depth of shielding from the Badhwar-OâNeill GCR model (OâNeill, 2007), under four different assumptions about the level of solar modulation. These calculations used a simplified transport code that did not include the dose contribution from projectile and target fragments produced in the shielding. NOTE: Dose rates are at 1 AU; LIS, local interstellar GCR spectrum. SOURCE: Provided by P.M. OâNeill, NASA Johnson Space Center, for this committee. R01155, Figure 2-10 Fixed image, not changeable â¢ The rates of these very large SPEs appear to have an ~80-year periodicity, corresponding to the so-called Gleissberg cycle in sunspot numbers. â¢ The space age has coincided with a minimum in the Gleissberg cycle and in SPE production. McCracken and McDonald (2004) summarize the implications of their studies by saying: We predict that the frequency of large solar proton events may increase from its present low value by a factor of 6 to 8, commencing perhaps in Solar Cycle 24. Should this prediction be correct, the Earth will experience substantially more solar proton events and ground level events than has been our experience since 1950. This will have major implications for space flight and engineering. It is worthwhile to see how Solar Cycle 23, which is just now drawing to a close, compares with previous experience from the space age. Figure 2-11 shows the cycle-integrated >10 MeV and >30 MeV solar proton flu- ences for Cycles 19-23. These results show that Cycle 23 has had the largest >10 MeV solar proton fluence since the start of the space age. This is also true at >30 MeV, although the increment over Cycle 19 is small and prob- ably within the systematic uncertainties. In any case, at both energies, the Cycle 23 fluences are clearly higher than those in Cycles 20 and 21. Total proton fluence is a more robust measure than ânumber of events,â since many âeventsâ actually comprise multiple outbursts, and there â are no universally agreed criteria for how to count âeventsâ in these circumstances.
34 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION FIGURE 2-11â Cycle-integrated solar proton fluences >10 MeV and >30 MeV for Solar Cycles 19-23. Fluence numbers for Cycles 19-22 come from Shea and Smart (1990, 1999). Cycle 23 fluences were derived from the Geostationary Operational Environmental Satellite proton data available at the National Geophysical Data Center. R01155, Figure 2-11 Are these upswings a statistical outlier? changeable Fixed image, not Or are they perhaps the first harbingers of the changes predicted by McCracken and McDonald (2004)? At this point, it is impossible to say with certainty. However, Feynman et al. (1993) present a statistical model of solar proton exposures, which quantifies the probability of observing a par- ticular solar proton fluence in a given mission duration, based on the SPE database from 1963-1991. For each of these energies, that model suggests only a ~7 percent probability for observing cycle-integrated fluences at least as large as those reported for Cycle 23. Finding 2-4. Space radiation climate. Ice-core studies indicate that the past ~50 years may have coincided with a comparatively benign space radiation climate, in terms of both GCR modulation levels and the frequency of very large SPE events. Of particular concern is the possibility of a six- to eightfold increase in the number of very large SPE events, perhaps starting within the next decade. If such an increase were to occur, it would have a major impact on the design and operation of Exploration systems. Recommendation 2-1. Planning for long-term changes in space climate. NASA must ultimately judge how much weight to assign to the cautionary findings from ice cores on a potentially more severe space radiation climate in the future. Given that the Exploration initiative envisions a commitment of the nationâs resources that spans decades, NASA should ensure that the mission architecture has sufficient flexibility and margin to cope with such changes, should they occur. The relevant figures in Feynman et al. (1993) are Figure 4c and 4d. Note that the fluence on the horizontal axis of these figures has been â mislabeled. The units should be cmâ2, not cmâ2-srâ1 (J. Feynman, personal communication with A. Tylka).
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 35 At present, it is unclear whether or not this is the case. NASA should make an assessment of the potential impact of these radiation-climate changes on Exploration missions. Knowledge Gaps Coronal Mass Ejections There is only a partial understanding of what causes a CME to erupt from the Sun and of the necessary and sufficient conditions by which a large, fast CME produces high intensities of high-energy particles of potential relevance to astronaut safety. The broad range of coordinated observations from a fleet of scientific satellites in the past 10 years has yielded substantial progress and focused attention on key unresolved issues. These issues encompass questions about the detailed physics of solar eruptions, particle acceleration, and transport processes, and the conditions under which they transpire. These questions can only be answered by new science missions that will sharpen and broaden our view of these phenomena. The planned Solar Dynamics Observatory will reveal the solar-magnetic topologies that give rise to CMEs. The recently launched STEREO satellites will provide the first three-dimensional view of CME eruptions; they will also map, for the first time, the longitudinal structure of large SPE events, something that is essential for developing predictive SPE capabilities for Mars. The proposed Sentinels mission (Lin, 2006) will provide in situ observations of the radial evolution of SPEs between the Sun and Earth, thereby providing unique data for developing physics-based SPE models. Modeling efforts, such as Boston Universityâs Center for Integrated Space Weather Modeling, will incorporate these new sources of data into physics-based numerical simulations. Current SPE prediction models make no utilization of what is already known about the role of fast CMEs in producing SPEs. A significant empirical correlation between CME speed and SPE intensity has long been known. Recent work has shown that the scatter in this correlation is tightened by identifying those CMEs for which another CME has erupted from the same active region in the previous day (Gopalswamy et al., 2004; Kahler and Vourlidas, 2005). At present, forecasters do not take advantage of these potential predictors, simply because routine real-time CME observations are not available to them. NASA and NOAA are presumably moving aggressively to alleviate this deficiency in preparation for Exploration missions. In the meantime, SPE-prediction algorithms incorporating this knowledge can be developed and tested using the extensive database acquired over the past decade. One- to 3-day forecasts will require substantial improvements in our understanding of the physical causes of CMEs, the particle acceleration and transport mechanisms, the role of a âseed populationâ in generating very high energy particles, and the dynamics of the ambient solar wind. These investigations will need a focused effort and will involve a wider community of researchers. Note that there are additional reasons, beyond NASA, to acquire this understanding, as the CMEs and other space weather effects lead directly to other impacts at Earth. The PROTONS and PPS models were developed under the paradigm that the particle acceleration occurred at the site of the associated x-ray flare. However, there is little correlation between the time, location, and magnitude of a flare and the time, location, and speed of an associated CME eruption other than a general relationship that very active solar regions generate large CMEs and big flares. Hence, current forecasting models do badly in predicting extreme events, and extreme events represent the greatest danger to human spaceflight crews. Observational data on the source CME and associated shock are available only from science platforms, and these data do not often have an operationally useful cadence or timeliness. Another weakness of each of these models is an inability to predict enhancements to the proton flux when a strong shock moving with the CME passes over Earthâpotentially the most dangerous period to a human in space. Operational forecasts also do not include high-energy heavy ions, which may potentially impact spacecraft electronics. There is general recognition within the NOAA SEC and the USAF that a more physics-based model of SPEs is needed to improve on these predictions. Nonetheless, a modest effort to identify better proxies or indicators of SPEs could offer some potential to marginally improve phenomenological SPE forecasts. Examples under investi- gation include the direct detection of interplanetary or solar surface shocks and x-ray spectral signatures that may presage SPEs. It may be possible to monitor the local arrival of energetic solar electrons as a precursor to large solar proton events, with an advanced warning of tens of minutes (Posner, 2007).
36 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION Methods involving artificial intelligence, Bayesian inference, and locally weighted regression have demon- strated promise in providing nowcasting capabilities after energetic particles begin to arrive. These methods are capable of predicting, with reasonable accuracy, total doses and the future temporal evolution of the dose as par- ticles arrive very early in the evolution of the event. Typically the forecasts are within 10 to 20 percent error of the actual profile before half of the event dose was received. However, sometimes one parameter was not adequately forecasted, which causes the event to be either over- or underpredicted (Hoff et al., 2003; Hines et al., 2005; Neal and Townsend, 2005). However, these methods are at present unable to forecast event fluence levels and their assoÂ ciated doses until after particles begin to arrive. Hence, they could provide a much-needed short-term capability for mission operations, but in a merely stopgap role. Computer codes implementing these models are currently research codes and not in the form of operational tools that are usable by mission operations personnel. However, operational tools using these codes are currently under development with funding provided by the NASA Living With a Star Targeted Research and Technology program. Significant new capability exists within the space physics communityâobservations from ACE, SOHO, S Â TEREO, and Wind, and the establishment of multiple interdisciplinary space weather modeling consortia such as the Center for Space Environment Monitoring at the University of Michigan or the Community Coordinated Model- ing Center, a partnership of government agencies centered at Goddard Space Flight Centerâprovide a substantial opportunity to advance the state of the art in CME forecasting and thus better SPE forecasts. The ÂSTEREO mission and the planned Solar Dynamics Observatory will also contribute to the understanding of CMEs and SPEs. This knowledge in turn may help in two ways. First, it may help interpret the climatological data on SPEs, including the distribution of SPEs within the solar cycle and whether there are physical limits on the worst-case SPE. Second, a better understanding of the physics of CMEs and the particle acceleration process may lead to longer-term forecasts of all-clear periods, periods of hours to days when it is extremely unlikely for a significant event to occur. Choice of the Design-Standard Solar Spectrum NASA has adopted the proton spectrum of the August 1972 event, as parameterized in King (1974), as the standard for assessing the ability of shielding to reduce astronaut radiation exposure to acceptable levels during an SPE (NASA, 2006). This proton spectrum is shown in Figure 2-12. Also shown in Figure 2-12 is the spectrum of the October 1989 SPE, as given by Xapsos (2000). Both of these events were the product of three successive solar eruptions over a period of 6 days, as solar rotation carried an active region across the face of the Sun Â(Simnett, 1976; Sauer, 1993). In both cases, each successive eruption produced particle spectra that extended to higher Âenergies, as observed at Earth. There are significant differences in the time structure of the two episodes, Âprimarily associated with the arrival of a powerful shock at Earth on October 20, 1989, something for which there is no comparable feature in August 1972. Nevertheless, the reported event-integrated proton fluences below ~100Â MeV are the same in these two events to within a factor of two. Both of these events are very rare; there is no reason to regard either of them as intrinsically âless likelyâ than the other and therefore less appropriate as a standard for shielding design. There may be engineering reasons for adopting August 1972 as the design-standard event. But in terms of radiation safety, there is no reason to give it greater weight than the October 1989 event. The critical difference between the two events becomes apparent when the spectra are examined above ~100Â MeV, where the October 1989 spectrum is harder: it falls off much more slowly with increasing energy than the exponential form given by King for August 1972 (King, 1974). The potential implication for shielding is Âillustrated in Figure 2-13, which shows the dose from each of these spectra at the center of a spherical shell of aluminum, as a function of the shellâs thickness. In these calculations, the aluminum shielding is followed by 8Â g/cm2 of water to crudely represent the bodyâs self-shielding. The dose calculations in Figure 2-13 do not include the production of secondary neutrons in the shielding and are therefore lower limits on the dose. Although these calculations are highly simplified compared with NASAâs state-of-the-art calculations for realistic spacecraft and habitat configurations, the results in Figure 2-13 nevertheless illustrate the reason for potential concern. If, for example, the objective were to reduce radiation exposure to 20 cGy (corresponding to roughly 30 cSv for a proton quality factor of 1.5), this level could be achieved for the August 1972 spectrum with about 8 g/cm 2 aluminum-
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 37 FIGURE 2-12â Representations of the event-integrated proton spectra for August 1972 (from King, 1974) and for October 1989 (from Wilson et al., 1999, and Xapsos, 2000). Measured data points (open circles of August 1972; filled circles for October 1989) are also shown, as discussed in the text. R01155, Figure 2-12 Fixed iamge, not changeable FIGURE 2-13â Dose-depth curves corresponding to the proton spectra in Figure 2-12. See the text for details. R01155, Figure 2-13 Fixed image, not changeable
38 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION equivalent shielding. For the October 1989 spectrum, the shielding thickness required to reduce exposure to the same level would need to be roughly 50 percent larger, at about 12 g/cm 2. A more rigorous simulation of realistic shielding distributions will produce different numbers, but the relative magnitude of results from the two events may very well be on this order. There is also another reason for hesitancy in adopting the King spectrum (King, 1974) as the final word in assessing the adequacy of astronaut radiation shielding. The data points in Figure 2-12 show the actual proton fluence measurements that were used in generating the King (1974) and Xapsos (2000) fits. All of these measure- ments come from satellites except for one: the highest-energy data point in the August 1972 event at 200 MeV. This data point, which largely determines the shape of the spectral fit, comes from a series of stratospheric balloon launches (Bazilevskaya et al., 1973). One cannot foreclose the possibility of significantly larger systematic error in the balloon data point. If that were the case, the King spectrum might not be a reliable representation of what actually occurred in August 1972, at least not above ~100 MeV. It is also important to use the correct spectral form when making radiation calculations. The King spec- trum for August 1972 (King, 1974) is exponential in energy. The Xapsos (2000) spectrum for October 1989 is a Weibull function in energy. Figure 2-12 also shows an older representation of the October 1989 spectrum (Wilson et al., 1999), which appears to be derived from an exponential fit in rigidity to GOES data points below 100 MeV (Townsend and Zapp, 1999). This older parameterization is clearly inadequate in that it falls below the >300 MeV GOES data points by more than an order of magnitude. As a result, this alternative form is closer to the August 1972 spectrum at high energies and, as shown in Figure 2-13, leads to roughly comparable dose levels under ~10 g/cm2 or more of spacecraft shielding. Any comparative study of radiation exposure that has used an e Â xponential-in-rigidity for the October 1989 spectrum is potentially invalid, especially for shielding thicknesses greater that ~5 g/cm2. Any conclusions drawn from such studies should be re-examined. These same concerns pertain to published spectra for the September 1989 SPE and perhaps other events as well. Spectral extrapolations used in radiation-shielding calculations can be cross-checked against existing proton measurements in the range of ~200-500 MeV. There are several potential sources of such data that may not have been adequately exploited to date. NASAâs Solar Anomalous and Magnetospheric Particle Explorer (SAMPEX) satellite provides solar proton measurements out to ~400 MeV for events since 1992 (Mewaldt et al., 2005a,b). Several GOES satellites have carried the High Energy Proton and Alpha Detector (HEPAD), which provided the data points above 300 MeV in Figure 2-12; data are available for most large SPEs since 1986. However, the HEPAD data are difficult to work with owing to uncertainties in the calibration and background-subtraction procedures (Smart and Shea, 1999) and poor documentation for the online HEPAD data available from the National Geophysical Data Center. Finally, neutron monitors can provide proton fluence measurements above ~500 MeV, as a check on whether a functional form falls too steeply at high energies. However, most published studies of neutron monitor observations to date have not included the absolute proton fluences that are required for radiation calculations. Uncertainty in the spectral form of SPE protons has important implications for onboard particle detectors. These detectors will be an essential component of the radiation protection program for Exploration missions. They should be sized so as to measure the solar proton spectrum accurately out to ~500 MeV. They must also be designed so that dead-time effects and other data-rate limitations, which often plague particle instruments in very large SPEs, do not significantly compromise the reliability of the measurements. Finally, the committee notes that the October 1989 SPE has been used as a design standard elsewhere in NASA. In particular, NASA has twice designated the Cosmic Ray Effects on Micro-Electronics (1996 revision) (CREME96) SPE model (Tylka et al., 1997) as the standard for electronics design in Exploration vehicles (NASA, 2006). This model is based on measurements of the October 1989 event, and its proton spectrum is essentially identical to that of Xapsos (2000). The factor of two is intended to raise the event to the level of a âworst case.â CREME96 also includes solar heavy ions, since spacecraft electronics are potentially vulnerable to both solar protons and solar heavy ions. This older representation was first developed when there were still substantial uncertainties about the reliability and potential relevance of â the high-energy GOES measurements. The committee notes in passing that this erroneous spectrum for October 1989 has been reproduced in other comparisons of SPE spectra (e.g., Turner, 1999; Mewaldt et al., 2005a; NCRP, 2006).
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 39 Finding 2-5. The King spectrum as a design standard. Although the committee recognizes the advantages of adopting a specific solar proton spectrum as the design standard, NASAâs current strategy of evaluating the effi- cacy of an SPE shielding configuration using only the August 1972 King spectrum is not adequate. Under typical depths of shielding for Exploration vehicles, the level of radiation exposure produced by other large events in the historical record could exceed the exposure of August 1972. Finding 2-6. Spectra data fitting. There is no theoretical basis for any of the published spectral fits to large SPEs. The extrapolation to energies beyond 100 MeV must therefore be guided by data. Solar proton spectral forms based on data that do not extend to ~500 MeV may very well give misleading results in evaluations of the efficacy of radiation shielding for astronauts. Recommendation 2-2. SPE design standards. The dose levels made possible by a shielding design should also be calculated using the observed proton spectrum from other large events in the historical record, even if it is not fea- sible to modify the shielding design as a result. The October 1989 event is particularly important in this regard. Recommendation 2-3. Uncertainties in spectra data fitting. NASA should make use of existing data to re- e Â valuate the spectra beyond 100 MeV in large events in the historical record and should assess the impact of uncertainties in the high-energy spectra on the adequacy of radiation shielding designs. Solar Wind Models As noted in the National Research Council report Space Radiation Hazards and the Vision for Space Explo- ration (NRC, 2006), reliable predictions of SPE onset and severity require an understanding of the heliospheric environment through which the energetic particles propagate. The ability to characterize the solar wind and interÂ planetary magnetic field accurately during and after the passage of CMEs is improving. The standard model in use operationally today is the Wang-Sheeley-Arge model (Arge and Pizzo, 2000), an empirical model based on observations of the solar magnetic field. More physics-based three-dimensional models are under development (e.g., Riley et al., 2001; Roussev et al., 2003). These models have not yet been validated with adequate spatial and temporal in situ observations. Data from NASAâs current STEREO mission (to explore longitudinal structure at 1Â AU) and NASAâs proposed Sentinels mission (to explore radial dependencies inside of 1 AU) will be particularly valuable in this regard. TRAPPED RADIATION In addition to galactic cosmic rays and solar energetic particles, particles trapped in Earthâs magnetic field comprise the third major component of the near-Earth ionizing radiation environment. The trapped particles in these âradiation beltsâ that surround Earth include electrons, protons, and heavier ions. At Earth, the trapped electron spectra extend out to about 10 MeV, and trapped proton spectra extend to hundreds of MeV. The trapped proton intensities near Earth are among the highest proton intensities potentially encountered by manned Exploration missions. Trapped protons are an important design consideration for all spacecraft operating in near-Earth orbits. Measurements to date indicate that the energy spectra of the heavy ions are soft and that their intensities above a few tens of MeV per nucleon are too small to be of practical concern. The trapped particle models in current use are the AP-8 for protons and the AE-8 for electrons (Sawyer and Vette, 1976). Although the AP-8/AE-8 models are still widely used for spacecraft design, it is recognized that they are severely outdated because of the secular changes in Earthâs magnetic field since the era when the measurements underlying the models were made. It is also now recognized that radiation belts vary on timescales shorter than just the solar maximum/solar minimum levels described in the AP-8/AE-8 models (e.g., Blake et al., 1992; Li et al., 1993). NASAâs Living with a Star programâs Radiation Belt Storm Probes, which are scheduled for launch in 2012, are expected to provide measurements that will be the basis for new models. In the meantime, there are also efforts aimed at extending the capabilities of existing models and improving their accuracy through re-analysis of
40 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION TABLE 2-1â Average Radiation Doses of the Apollo Mission Flight Crews Apollo Mission Skin Dose (cGy) â 7a 0.16 â8 0.16 â 9a 0.2 10 0.48 11 0.18 12 0.58 13 0.24 14 1.14 15 0.3 16 0.51 17 0.55 a Low-Earth-orbit missionsâdid not go to the Moon. SOURCE: Bailey, 1975. existing datasets (Ginet and OâBrien, 2007; Ginet et al., 2007). Beta versions of the new models (designated AP-9 and AE-9) are scheduled for release in the 2009-2010 time frame. Efforts at improving models of Earthâs trapped radiation are important for other NASA missions, as well as for DOD and for the commercial use of space. The magnetic fields on the Moon and at Mars are vestigial; they are too weak to sustain any trapped-particle population. Accordingly, trapped radiation will not be considered further in this report, except to note that the exploration vehicles must pass through Earthâs radiation belts when departing from and returning to Earth. Un- like crews on the space shuttle and the ISS, whose orbits just clip the lower reaches of the radiation belts, crews on exploration vehicles can potentially be exposed to much larger doses while transiting Earthâs radiation belts. In fact, as shown in Table 2-1, the largest radiation exposure of the Apollo era occurred on Apollo 14, when the return trajectory carried the spacecraft through an intense region of trapped protons. Although the additional dose in the case of Apollo 14 was less than 1 cGy, mission planners should bear this experience in mind. Avoidance of such an exposure is an example of the purpose of attention to the As Low As Reasonably Achievable (ALARA) principle. SECONDARY RADIATION Whenever the local space radiation environment in deep space impinges on the shell of the spacecraft, the solar and galactic cosmic rays penetrate the spacecraft structure and shielding and their physical characteristics are altered by atomic and nuclear collisions with the constituent atoms of the structural and shielding materials. The physical changes in these radiation fields as they pass through bulk material shielding (and through body tissues overlying critical organs) include energy losses, mainly due to atomic collisions and nuclear elastic and inelastic collisions, and changes in particle identities resulting from nuclear fragmentation reactions that produce secondary neutrons, protons, mesons, and heavier charged particles. Fragmentation occurs in both the projectile, which is usually moving at high speeds, and in the target, which is usually stationary or nearly so. Hence, the more ener- getic secondary radiations are generally produced from the projectile nuclei. These alterations in the composition of the radiation fields as they pass through the spacecraft and crew membersâ bodies are described by radiation transport codes, which must include the effects of spacecraft and human geometries. Biological damage, and the concomitant risk to crew members, result from interactions of the altered, transported radiation environment at the local organ or tissue site. This local radiation environment will be some mixture of primary and secondary radiations, which vary depending on the material composition and thickness of the target materials. Figure 2-14 depicts the complexity associated with these internal radiation fields, where the external radiation interacts with the spacecraft structure creating an internal environment that is different in composition from the external one. As the thickness of the spacecraft walls increases, the differences between the internal and external environments
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 41 FIGURE 2-14â Schematic of the space radiation protection problem. SOURCE: Wilson et al., 1997. R01155, Figure 2-14 Fixed image, not changeable increase as well. Because of self-shielding of internal body organs by overlying tissue, the local radiation fields at the internal organs are different from the internal environment within the spacecraft. Dosimetric quantities, such as dose, dose equivalent, and effective dose, which are presumed to be related to biological risk, can then be estimated from the calculated local particle fluxes and energies, in the tissue or organ of interest. These radiation transport methods can also be used to assess damage to spacecraft electronics. One representative calculation of the percentage variation in organ dose-equivalent contributions from incident galactic cosmic rays and their secondary radiations, produced by projectile and target fragmentation processes, is displayed in Table 2-2. The calculations assumed spacecraft aluminum shielding of the indicated area density, TABLE 2-2â Percentage Contributions to the Annual Total Dose Equivalent, Rounded to the Nearest Whole Percent, from Surviving Incident Galactic Cosmic Rays, Their Projectile Fragmentation Products, and Target Nuclear Fragments for Various Organs and Several Thicknesses of Aluminum Shielding Skin Ocular Lens Bone Marrow Type Contribution (g/cm2) (g/cm2) (g/cm2) No aluminum shielding Incident ions 89 86 65 Projectile fragments 9 11 30 Target fragments 2 3 5 5 g/cm2 aluminum shielding Incident ions 90 79 62 Projectile fragments 6 17 35 Target fragments 4 4 3 30 g/cm2 aluminum shielding Incident ions 56 56 49 Projectile fragments 39 39 45 Target fragments 5 5 5 SOURCE: Townsend et al., 1992. Reprinted with permission from the Radiation Research Society.
42 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION and employed the Computerized Anatomical Man model of human geometry. Note that even with no external shielding, secondary radiation fields produced by the overlying body tissue will be responsible for about one- third of the bone marrow dose equivalent, and more than 10 percent of the skin and eye dose equivalents. As the spacecraft shield thickness increases, the increased production of secondary radiations results in an increase in their percentage contributions to the organ dose equivalents. Note that most of the organ dose equivalent behind 30 g/cm2 aluminum shielding results from secondary radiations produced in the spacecraft structure and overlying body tissues of the astronauts. Figure 2-15 displays the dose equivalent (H) in water (a surrogate for tissue) for an incident galactic cosmic radiation spectrum (1977 solar minimum) after passage through the indicated thickness (area density) of aluminum and an additional 10 cm of water. These dose equivalents are fairly representative of bone marrow values. Note that the contribution from ions heavier than neon (Z = 10) decreases rapidly as the shield thickness increases, as does the contribution from lighter ions ranging from lithium (Z = 3) through fluorine (Z = 9). This rapid decrease in the contributions from ions heavier than helium (Z = 2) is due to a combination of their rapid energy losses from Coulomb scattering with atomic electrons and to their breakup (fragmentation) into lighter ions with smaller atomic numbers as they penetrate the shield. Low-energy, heavy particles stop very quickly owing to their large stopping powers. Hence, the values for H are large for thin shields. If the heavy ions did not fragment, the dose from them would increase as they slowed. However, as the higher-energy particles penetrate, they break up into FIGURE 2-15â Annual dose equivalent in centisieverts (cSv) for the 1977 GCR solar minimum spectrum of Badhwar and OâNeill as a function of aluminum shielding area density plus 10 cm of water (to approximate body self-shielding for bone marrow). The bottom curve (labeled Z = 0) displays the contribution due to neutrons. The curve labeled Z = 0, 1 displays the R01155, Figure 2-15 cumulative contributions due to protons and neutrons (The difference between the two curves is the contribution due solely to protons). The curve labeled Z = 0,Fixed image, not changeable adding the Z = 2 particles to the Z = 0, 1 results. 1, 2 displays contributions resulting from The spectrum builds until all contributions for Z = 0 through 28 have been included (top curve). Note that the contributions from Z = 0, 1, and 2 gradually increase as the Al area density increases, whereas the contributions from the heavy ions (Z > 2) decrease with increasing Al area density because they are removed by fragmentation (breakup) processes in the Al shield and in the overlying body tissue. SOURCE: Generated with HZETRN (see Appendix D), quality factors from ICRP (1991).
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 43 lighter fragments, which have smaller stopping powers (less dose per particle) and also lower Q values per particle. As they continue to penetrate, they fragment into smaller ions and eventually no longer contribute to ZÂ >Â 2; instead they now contribute to the ZÂ =Â 1,Â 2 spectra. The total dose equivalent versus depth curve then decreases much more slowly, as the shield thickness increases, because of the large numbers of neutrons, hydrogen nuclei (mainly protons), and helium nuclei (mainly alpha particles) produced by fragmentation. The hydrogen and helium spectra are mixtures of primary and secondary particles. The neutron spectra are all secondary neutrons since there are no neutrons in the incident spectrum. These lighter particles have ranges and collision mean-free paths that are much larger than the more highly charged parent nuclei that produced them, which accounts for this slowly decreasing trend in the dose-equivalent versus shield-thickness curves. Knowledge Gaps There is a serious concern about the purely one-dimensional nature of the space radiation transport codes used to estimate shielding requirements and potential biological risks to crews on deep-space missions, including planned lunar missions. Based on presentations to the committee by John Wilson from the NASA Langley Research Center (âCurrent and Projected Radiation Shielding Approaches and Capabilities,â December 12, 2006) and on the report of the shielding workshop held at the center in January 2007 (NASA/NIA, 2007), significant attention has been paid to âfragmentationâ cross sections (i.e., probability of interaction). There has also been some work reported on âcharge-changing cross sections.â However, there were no references to dependence on angle or energy of the fragments. For an incident-GCR projectile and any material target nuclei, there will be one or more fragments coming out of any nuclear interaction, at different energies and into different directions. The data necessary for calculations of GCR shielding come in various levels. In the presentations to the committee, âfragmentationâ was used to denote the probability of one given fragment, of any energy, being emitted into a small cone pointed in the forward direction (at 0 degrees). This is useful but limited information. For shielding materials that are not too thick, most of the fragments will continue in a fairly narrow cone in the forward direction, so not much is missed by detecting them with a small detector. However, the fragmentsâ energies at any point inside a shielding mate- rial will Âdepend on the interaction location where they were made. As the shielding material gets thicker, some of the fragments will be made by second, third, and further generations of interacting particles. For that reason, measurements of the energy of the outgoing particle in thick shielding are an essential test of the contribution of intermediate reaction products to the final mix. At that point, angular distributions begin to be important. Finally, when a galactic cosmic ray projectile undergoes a nuclear reaction, it breaks up into many pieces, not just one or two. As a result, several fairly heavy fragments may come out close together, at the same time. The number of fragments, or âmultiplicity,â is an important component of knowledge about the reaction. RADIATION FROM NUCLEAR GROUND POWER Under its technology development program, NASA is developing fission reactors for potential use during Exploration missions. Such reactors will be particularly valuable on nonpolar locations on the Moon and will also be particularly valuable on Mars, since solar power will not be sufficient in either location. Presented below is a review of the knowledge of radiation environments associated with nuclear reactors. The engineering challenges associated with the safe operation of nuclear reactors are discussed in Chapter 4. Development and Use of Nuclear Ground Power Although discovered in 1939, the nuclear fission chain reaction was not harnessed until the 1950s for com- mercial use in nuclear reactors. The first electricity-generating nuclear power plant was the Experimental Breeder Reactor-I, at the National Reactor Testing Station in Idaho. It initially produced nearly 100 kW, enough to power the equipment in the small reactor building (Fischer, 1997). To date there are 436 nuclear power plants in the world, operating in 32 countries, with a total capacity of 370,000 MW, providing 16 percent of total electricity generation. Terrestrial nuclear engineering and technology proved safe and accurate, and nuclear engineers gained tremendous
44 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION experience in building and using smaller nuclear reactor units for research, training, education, materials testing, and the production of radioisotopes for medicine and industry, and for propelling ships and submarines. There are 284 research reactors, operating in 56 countries, and more than 220 submarine reactors. Radiation from Terrestrial Nuclear Ground Power Plants The principal concerns and fear regarding the operation of nuclear power plants arise from perceived risk associated with potential exposure to radiation and effects of radiation on health. More than 50 years of nuclear power development and usage have resulted in only two major accidents, the Three Mile Island (TMI) (in the United States in 1979) and Chernobyl (in Ukraine in 1986). The sequence of certain events including equipment malfunctions, design-related problems, and worker errors led to a partial meltdown of the TMI-2 reactor core, but only very small and insignificant off-site releases of radioactivity. The experience from this accident, based on decades-long measurements and analysis, shows that radioactive material was not readily mobilized beyond the immediate internal reactor structure, and thus has helped create better understanding of the predictions related to radioactive releases in the environment. Apart from the Chernobyl accident that released significant amount of radioactive isotopes into the environment, no single nuclear worker or member of the public has died due to radioactive exposure from civil nuclear power plant operation. Most of the radiation exposure for the average person dwelling in the United States comes from natural background radiation (55%), while the main remaining exposure comes from medical procedures (15%), cosmic radiation (8%), and soil (8%) (NCRP, 1987). The accumulated experience in using nuclear technology shows that the overall risk from the normal operation of nuclear power plants (including the fuel cycle and probabilistic prediction of potential design-basis accidents) will contribute around 0.05 percent of the total dose. Several large studies in the United States, Canada, and Europe have found no evidence of any increase in cancer mortality among people living near nuclear ground power plants. For example, the National Cancer Institute of the National Institutes of Health has reported in a fact sheet (available at http://www.cancer.gov/cancertopics/factsheet/Risk/ nuclear-facilities) on the results of a survey that evaluated mortality from 16 types of cancer and found no increased incidence of cancer mortality for people living near 62 power plants in the United States, and also no increase in the incidence of childhood leukemia mortality (Jablon et al., 1991). Radiation Shielding and Control Nuclear reactors are carefully designed with a number of engineering barriers that prevent the release of radioactive material and ensure safe and reliable isolation from the environment. The main safety design criteria are intended to reduce the chance of release of radioactive materials into the environment during normal operation and accidents. Besides a number of primary and backup systems employed to monitor, control, and support the safe operation of a plant, each of the reactors has a well-designed series of engineering and physical barriers with the ultimate goal of preventing discharge of radioactive materials. Regulations and Policy The U.S. Nuclear Regulatory Commission (U.S. NRC) is the federal agency that establishes the policy and controls the radioactive release from nuclear ground power into the environment. The U.S. NRCâs mission is âto protect public health and safety and the environment from the effects of radiation from nuclear reactors, Âmaterials, and waste facilities,â including nuclear materials used in commercial nuclear ground power. The U.S. NRC requires See U.S. NRC, Fact Sheet on TMI Accident, available at http://www.nrc.gov/reading-rm/doc-collections/fact-sheets/3mile-isle.html. â A committee of the National Council on Radiation Protection and Measurements (NCRP) is currently compiling an update of these statistics. â Although updated statistics were not available in time for this report, it is known, for example, that the exposure due to medical procedures has increased by 650 percent (Mettler, 2007). See IAEA Power Reactor Information System, available at http://www.iaea.org/programmes/a2/index.html. â See U.S. NRC Web site, http://www.nrc.gov/about-nrc/radiation.html. â
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 45 that the plant operator monitor the environmental impact of the nuclear power plant by measuring airborne and liquid radioactive effluents and taking radioactivity readings from the air, water, animals, and plants surround- ing the facility (U.S. NRC, 2002). Since the attacks on September 11, 2001, the U.S. NRC has strengthened the regulations on nuclear reactor security. Finding 2-7. Knowledge of radiation from nuclear ground power. Experience with nuclear power on Earth has provided sufficient knowledge to create this capability on the Moon. The remaining challenges are engineer- ing problems, not scientific problems. Experiments to show the operational safety of space and planetary-surface fission power systems, including unique design features such as compactness, light weight, and heat transport and heat rejection in reduced gravity, will be important. REFERENCES Arge, C.N., and V.J. Pizzo. 2000. Improvements in the prediction of solar wind conditions using near-real time solar magnetic field updates. Journal of Geophysical Research 105:10465, doi:10.1029/1999JA900262. Bailey, J.V. 1975. Biomedical Results of Apollo. Section II, Chapter 3 in Radiation Protection and Instrumentation. NASA Headquarters, Washington, D.C. Bazilevskaya, G.A., Y.I. Stozhkov, A.N. Charakhchyan, and T.N. Charakhchyan. 1973. The energy spectra and the conditions of propagation in the interplanetary space for solar protons during the cosmic ray events of August 4 to 9, 1972. 13th International Cosmic Ray Conference 2:1702-1707. Bieber, J.W., J. Chen, W.H. Matthaeus, C.W. Smith, and M.A. Pomerantz. 1993. Long term variations of interplanetary magnetic field with implications for cosmic ray modulation. Journal of Geophysical Research 98(A3):3585-3603. Blake, J.B., M.D. Looper, D.N. Baker, R. Nakamura, B. Klecker, and D. Hovestadt. 1992. New high temporal and Âspatial resolution measurements by SAMPEX of the precipitation of relativistic electrons. Advances in Space Research 18(8):171-186. Burlaga, L.F., J. Perko, and J. Pirraglia. ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ Astro- 1993. Cosmic Ray modulation, merged interaction regions, and multifractals. physical Journal 407(1):347-358. Cane, H.V., G. Wibberenz, I.G. Richardson, and T.T. von Rosenvinge. 1999. Cosmic ray modulation and the solar magnetic field. Geophysical Research Letters 26(5):565-568. Cliver, E.W., and H.V. Cane. 2002. Gradual and impulsive solar energetic particle events. EOS Transactions 83:61-68. Cliver, E.W., and A.G. Ling. 2001. Coronal mass ejections, open magnetic flux, and cosmic-ray modulation. Astrophysical Journal 556(1):432-437. Desai, M.I., G.M. Mason, J.R. Dwyer, J.E. Mazur, R.E. Gold, S.M. Krimigis, C.W. Smith, and R.M. Skoug. 2003. Evidence for a suprathermal seed population of heavy ions accelerated by interplanetary shocks near 1 AU. Astrophysical Journal 588(1):1149-1162. Desai, M.I., G.M. Mason, R.E. Gold, S.M. Krimigis, C.M.S. Cohen, R.A. Mewaldt, J.E. Mazur, and J.R. Dwyer. 2006. Heavy-ion elemental abundances in large solar energetic particle events and their implications for the seed population. Astrophysical Journal 649:470-489. Feynman, J., G. Spitale, and J. Wang. 1993. Interplanetary proton fluence model: JPL 1991. Journal of Geophysical Research 98:13281-13294. Fischer, D. 1997. History of the International Atomic Energy Agency: The First Forty Years. IAEA, Vienna, Austria. Available at http://www-pub.iaea.org/MTCD/publications/PDF/Pub1032_web.pdf. Ginet, G.P., and T.P. OâBrien. 2007. A New Generation Radiation Belt Model, AP9/AE9. NASA Goddard Space Flight Center, Greenbelt, Md., available at http://lws-set.gsfc.nasa.gov/RadSpecsForum.htm. Ginet, G.P., T.P. OâBrien, C. Groves, W. Olson, and G. Reeves. 2007. The AE(P)-9 radiation belt model: Requirements defini- tion and architecture. Presentation at the Space Weather Workshop, Boulder, Co., April 24-27. Available at http://helios. swpc.noaa.gov/sww/index.html. Gopalswamy, N., S. Yashiro, S. Krucker, G. Stenborg, and R.A. Howard. 2004. Intensity variation of large solar energetic particle events associated with coronal mass ejections. Journal of Geophysical Research 109(A12). Hines, J.W., L.W. Townsend, and T.F. Nichols. 2005. SPE dose prediction using locally weighted regression. Radiation Protec- tion Dosimetry 115(1-4):232-235. Hoff, J.L., L.W. Townsend, and J.W. Hines. 2003. Prediction of energetic solar particle event dose-time profiles using artificial neural networks. IEEE Transactions on Nuclear Science 50(6):2296-2300.
46 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION ICRP (International Commission on Radiological Protection). 1991. 1990 Recommendations of the International Commission on Radiological Protection. ICRP Publication 60. Annals of the ICRP 21. Pergamon Press, Elmsford, N.Y. Jablon, S., Z. Hrubec, and J.D. Boice, Jr. 1991. Cancer in populations living near nuclear facilities. A survey of mortality nationwide and incidence in two states. J. Am. Med. Assoc. 265:1403-1408. Kahler, S.W., and A. Vourlidas. ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ 2005. Fast coronal mass ejection environments and the production of solar energetic particle events. Journal of Geophysical Research 110(A12S01), doi:10.1029/2005JA011073. Kim, M.-H.Y., J.W. Wilson, F.A. Cucinotta, L.C. Simonsen, W. Atwell, F.F. Badavi, and J. Miller. 1999. Contribution of High Charge and Energy (HZE) Ions During Solar-Particle Event of September 29, 1989. NASA/TP-1999-209320. NASA, Washington, D.C. King, J.H. 1974. Solar proton fluences for 1977-1983 space missions. Journal of Spacecraft and Rockets 11(6):401-408. Lara, A., N. Gopalswamy, R.A. Caballero-Lopez, S. Yashiro, H. Xie, and J.F. Valdes-Galicia. ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ 2005. Coronal mass ejections and galactic cosmic ray modulation. Astrophysical Journal 625(1):441-450. Lee, M.A. 2007. What determines the composition of SEPs in gradual events? Space Science Reviews 130:221-229, doi:10.1007/s11214-007-9188-0. Li, X., I. Roth, M. Temerin, J. Wygant, M.K. Hudson, and J.B. Blake. 1993. Simulation of the prompt energization and transport of radiation particles during the March 23, 1991 SSC. Geophysical Research Letters 20:2423. Lieberman, B.S., and A.L. Melott. 2007. Considering the case for biodiversity cycles: Reexamining the evidence for periodicity in the fossil record. PLoS Computational Biology 2(8):e759. Lin, R.P. 2006. Solar Sentinels: Report of the Science and Technology Definition Team. NASA/TM-2006-214137. NASA, Washington, D.C. Mason, G.M., J.E. Mazur, and J.R. Dwyer. 1999. 3He enhancements in large solar energetic particle events. Astrophysical Journal Letters 525:L133-L136. Mason, G.M., J.E. Mazur, J.R. Dwyer, J.R. Jokipii, R.E. Gold, and S.M. Krimigis. 2004. Abundances of heavy and ultraheavy ions in 3He-rich solar flares. Astrophysical Journal 606:555-564. McCracken, K.G., and F.B. McDonald. 2004. A phenomenological study of long-term cosmic ray modulation, 850-1957 AD. Journal of Geophysical Research 109(A12103), doi:10.1029/2004JA010685. McCracken, K.G., G.A.M. Dreschhoff, E.J. Zeller, D.F. Smart, and M.A. Shea. 2001a. Solar cosmic ray events for the period 1561-1994. 1. Identification in polar ice, 1561-1950. Journal of Geophysical Research 106:21585-21598. McCracken, K.G., G.A.M. Dreschhoff, D.F. Smart, and M.A. Shea. 2001b. Solar cosmic ray events for the period 1561-1994. 2. The Gleissberg periodicity. Journal of Geophysical Research 106:21599-21610. McDonald, F.B. 1998. Cosmic ray modulation in the heliosphere: A phenomenological study. Space Science Reviews 83(1-2):33-50. Mettler, F.A. 2007. Magnitude of radiation uses and doses in the United States: NCRP Scientific Committee 6-2 analysis of medical exposures. Pp. 9-10 in Advances in Radiation Protection in Medicine. 43rd NCRP Annual Meeting, April 16-17. NCRP, Bethesda, Md. Mewaldt, R.A., C.M.S. Cohen, A.W. Labrador, R.A. Leske, G.M. Mason, M.I. Desai, M.D. Looper, J.E. Mazur, R.S. Selesnick, and D.K. Haggerty. 2005a. Proton, helium, and electron spectra during the large solar particle events of October-November 2003. Journal of Geophysical Research 110:A09S18, doi:10.1029/2005JA011038. Mewaldt, R.A., M.D. Looper, C.M.S. Cohen, G.M. Mason, D.K. Haggerty, M.I. Desai, A.W. Labrador, R.A. Leske, and J.E. Mazur. 2005b. Solar-particle energy spectra during the large events of October-November 2003 and January 2005. Pro- ceedings of the 29th International Cosmic Ray Conference 1:101-104. NASA (National Aeronautics and Space Administration). 2006. Constellation Program Design Specification for Natural EnviÂ ronments. CxP-70023. NASA, Washington, D.C. NASA/NIA (National Institute of Aerospace). 2007. Space radiation measurements and transport physics workshop. Hampton, Va., January 24-25. NASA, Washington, D.C. NCRP (National Council on Radiation Protection). 1987. Ionizing Radiation Exposure of the Population of the United States. Report No. 93. Bethesda, Md. NCRP. 1989. Guidance on Radiation Received in Space Activities. Report No. 98. Bethesda, Md. NCRP. 2006. Information Needed to Make Radiation Protection Recommendations for Space Missions Beyond Low-Earth Orbit. Report No. 153. Bethesda, Md. Neal, J.S., and L.W. Townsend. 2005. Multiple solar particle event dose time profile predictions using Bayesian inference. Radiation Protection Dosimetry 115(1-4):38-42. Newkirk, G., A.J. Hundhausen, and V. Pizzo. 1981. Solar cycle modulation of galactic cosmic rays: Speculation on the role of coronal transients. Journal of Geophysical Research 86:5387-5396.
CURRENT KNOWLEDGE OF THE RADIATION ENVIRONMENT 47 NRC (National Research Council). 2006. Space Radiation Hazards and the Vision for Space Exploration. The National Acad- emies Press, Washington, D.C. Nymmik, R.A. 1999. Probabilitistic model for fluences and peak fluxes of solar particles. Radiation Measurements 30:287-296. Nymmik, R.A., M.I. Panasyuk, T.I. Pervaja, and A.A. Suslov. 1992. A model of galactic cosmic ray fluxes. Nuclear Tracks and Radiation Measurements 20:427-429. Nymmik, R.A., M.I. Panasyuk, and A.A. Suslov. 1996. Galactic cosmic ray flux simulation and prediction. Advances in Space Research 17(2):19-30. OâNeill, P.M. 2007. Badhwar-OâNeill 2007 galactic cosmic ray (GCR) model using Advanced Composition Explorer (ACE) measurements for solar cycle 23. IEEE Nuclear and Space Radiation Effects Conference. Honolulu, Hawaii. July 23-27. Owens, M.J., and N.U. Crooker. 2006. Coronal mass ejections and magnetic flux buildup in the heliosphere. Journal of Geo- physical Research 111(A10). Pohl, M., and J.A. Esposito. 1998. Electron acceleration in supernova remnants and diffuse gamma rays above 1 GeV. Astro- physical Journal 507:327-338. Posner, A. 2007. Up to 1-hour forecasting of radiation hazards from solar energetic ion events with relativistic electrons. Space Weather 5(5):S05001, doi:10.1029/2006SW000268/. Reames, D.V. 1990. Energetic particles from impulsive solar flares. Astrophysical Journal Supplement Series 73:235, doi:10.1086/191456/. Reames, D.V. 1995. Coronal abundances determined from energetic particles. Advances in Space Research 15(7):41-51. Reames, D.V. 2000a. Abundances of trans-iron elements in solar energetic particle events. Astrophysical Journal Letters 540: L111-L114. Reames, D.V. 2000b. The observational consequences of proton-generated waves at shocks. Pp. 79-86 in Acceleration and Transport of Energetic Particles Observed in the Heliosphere, eds. R.A. Mewaldt, J.R. Jokipii, M.A. Lee, E. MÃ¶bius, and T.H. Zurbuchen. AIP Conference Proceedings 528. Melville, N.Y. Reames, D.V., and C.K. Ng. 2004. Heavy-element abundances in solar energetic particle events. Astrophysical Journal 610:510-522. Riley, P., J.A. Linker, and Z. Mikic. 2001. An empirically-driven global MHD model of the solar corona and inner heliosphere. Journal of Geophysical Research 106(A8):15889. doi:10.1029/2000JA000121. Roussev, I.I., T.I. Gombosi, I.V. Sokolov, M. Velli, W. Manchester IV, D.L. DeZeeuw, P. Liewer, G. TÃ³th, and J. Luhmann. 2003. A three-dimensional model of the solar wind incorporating solar magnetogram observations. Astrophysical Journal 595:L57-L61. Sandroos, A., and R. Vainio. 2007. Simulation results for heavy ion spectral variability in large gradual solar energetic particle events. Astrophysical Journal Letters 662:L127-L130. Sauer, H.H. 1993. GOES observations of energetic protons E>685 MeV: Ground level events from October 1983 to July 1992. Proceedings of the 23rd International Cosmic Ray Conference 3:254-257. Sawyer, D.M., and J.I. Vette. 1976. AP-8 Trapped Proton Environment for Solar Maximum and Solar Minimum. NSSDC/ WDCA-R&S, 76-06. NASA Goddard Space Flight Center, Greenbelt, Md., December. Simnett, G.M. 1976. Solar cosmic radiation during August 1972. Space Science Reviews 19:579-610. Shea, M.A., and D.F. Smart. 1990. A summary of major solar proton events. Solar Physics 127:297-320. Shea, M.A., and D.F. Smart. 1999. Patterns of solar proton events over four solar cycles. Proceedings of the 26th International Cosmic Ray Conference 6:374-377. Shea, M.A., D.F. Smart, G.A. Dreschhoff, K.G. McCracken, and H.E. Spence. 2006. Solar proton events for 450 years: The Carrington event in perspective. Advances in Space Research 38:232-238. Smart, D.F., and M.A. Shea. 1999. Comment on the use of GOES solar proton data and spectra in solar proton dose calcula- tions. Radiation Measurements 30(3):327-336. Townsend, L.W., and E.N. Zapp. 1999. Dose uncertainties for large solar particle events: Input spectra variability and human geometry approximations. Radiation Measurements 30(3):337-343. Townsend, L.W., F.A. Cucinotta, and J.W. Wilson. 1992. Interplanetary crew exposure estimates for galactic cosmic rays. Radiation Research 129(1):48-52. Turner, R. 1999. Risk Management Strategies During Solar Particle Events on a Human Mission to Mars or the Moon. ÂANSER, Arlington, Va. Tylka, A.J., and W.F. Dietrich. 1999. IMP8 observations of the spectra, composition, and variability of solar heavy ions at energies relevant to manned space missions. Radiation Measurements 30(3):345-360.
48 MANAGING SPACE RADIATION RISK IN THE NEW ERA OF SPACE EXPLORATION Tylka, A.J., and M.A. Lee. 2006. A model for spectral and compositional variability at high energies in large gradual solar particle events. Astrophysical Journal 646:1319-1334. Tylka, A.J., J.H. Adams, Jr., P.R. Boberg, B. Brownstein, W.F. Dietrich, E.O. Flueckiger, E.L. Petersen, M.A. Shea, D.F. Smart, and E.C. Smith. 1997. CREME96: A revision of the Cosmic-Ray Effects on Micro-Electronics Code. IEEE Transactions on Nuclear Science 44(6):2150-2159. Tylka, A.J., C.M.S. Cohen, W.F. Dietrich, C.G. Maclennan, R.E. McGuire, C.K. Ng, and D.V. Reames. 2001. Evidence for remnant flare suprathermals in the source population of solar energetic particles in the 2000 Bastille Day event. Astro- physical Journal Letters 558:L59-L63. Tylka, A.J., C.M.S. Cohen, W.F. Dietrich, M.A. Lee, C.G. Maclennan, R.A. Mewaldt, C.K. Ng, and D.V. Reames. 2005. Shock geometry, seed populations, and the origin of variable elemental composition at high energies in large gradual solar particle events. Astrophysical Journal 625:474-495. Tylka, A.J., C.M.S. Cohen, W.F. Dietrich, M.A. Lee, C.G. Maclennan, R.A. Mewaldt, C.K. Ng, and D.V. Reames. 2006. A comparative study of ion characteristics in the large gradual solar energetic particle events of 2002 April 21 and 2002 August 24. Astrophysical Journal Supplement Series 164:536-551. U.S. NRC (U.S. Nuclear Regulatory Commission). 2002. Fact Sheet on Environmental Monitoring. Available at http://www. nrc.gov/reading-rm/doc-collections/fact-sheets/env-monitoring.html. U.S. NRC, Fact Sheet on TMI Accident, available at http://www.nrc.gov/reading-rm/doc-collections/fact-sheets/3mile-isle. html. Wilson, J.W., F.A. Cucinotta, J.L. Shinn, M-H., Kim, and F.F. Badavi. 1997. Preliminary considerations. Pp. 1-15 in Shielding Strategies for Space Exploration, eds. J.W. Wilson, J. Miller, A. Konradi, and F.A. Cucinotta. NASA, Washington, D.C. Wilson, J.W., F.A. Cucinotta, J.L. Shinn, L.C. Simonsen, R.R. Dubey, W.R. Jordan, T.D. Jones, C.K. Chang, and M.Y. Kim. 1999. Shielding from solar particle events exposures in deep space. Radiation Measurements 30(3):361-382. Withbroe, G.L., T.J. Bogdan, W.F. Denig, J.F. Drake, P.R. Goode, J.R. Jokipii, S.W. Kahler, I.R. Mann, F.R. Toffoletto, A.J. Tylka, and W. Ward. 2006. Senior Review of the Sun-Solar System Connection Mission Operations and Data Analysis Programs. Submitted to Richard R. Fischer, Director, Heliophysics Division, NASA, Washington, D.C., February 7. Xapsos, M.A., 2000. Characterizing solar proton energy spectra for radiation effects applications. IEEE Transactions on Nuclear Science 47(6):2218-2223. Xapsos M.A., G.P. Summers, and E.A. Burke. 1998. Probability model for peak fluxes of solar proton events. IEEE ÂTransactions on Nuclear Science 45(6):2948-2953. Xapsos, M.A., G.P. Summers, J.L. Barth, E.G. Stassinopoulos, and E.A. Burke. 2000. Probability model for cumulative solar proton event fluences. IEEE Transactions on Nuclear Science 47(3):486-490.