Comparison of dose and risk estimates between ISS Partner Agencies for a 30-day lunar mission

Abstract

The International Partner Agencies of the International Space Station (ISS) present a comparison of the ionizing radiation absorbed dose and risk quantities used to characterize example missions in lunar space. This effort builds on previous collaborative work that characterizes radiation environments in space to support radiation protection for human spaceflight on ISS in low-Earth orbit (LEO) and exploration missions beyond (BLEO). A “shielded” ubiquitous galactic cosmic radiation (GCR) environment combined with––and separate from––the transient challenge of a solar particle event (SPE) was modelled for a simulated 30-day mission period. Simple geometries of relatively thin and uniform shields were chosen to represent the space vehicle and other available shielding, and male or female phantoms were used to represent the body’s self-shielding. Absorbed dose in organs and tissues and the effective dose were calculated for males and females. Risk parameters for cancer and other outcomes are presented for selected organs. The results of this intracomparison between ISS Partner Agencies itself provide insights to the level of agreement with which space agencies can perform organ dosimetry and calculate effective dose. This work was performed in collaboration with the advisory and guidance efforts of the International Commission on Radiological Protection (ICRP) Task Group 115 and will be presented in an ICRP Report

Introduction

Analysis tools used in this work are typically used to perform operational dosimetry for International Space Station (ISS) crews. The analyses were extended to prepare for human exploration missions to the moon and beyond with this study serving as a benchmark. The parameters currently used in ISS operations to describe radiation environments, dosimetry (organ absorbed dose, dose equivalent, effective dose), and risk of late deleterious effects (lifetime attributed cancer mortality, risk of exposure-induced death, risk of exposure-induced incidence of cancer, total radiation risk, mean lifetime reduction, LACM, REID, REIC, TRR, MLR) are described at a high level herein and elsewhere [23] in this volume, as are the career exposure limits imposed by each Agency for their astronauts. For NASA, those career limits that apply to ISS apply as well to all human spaceflight missions beyond low-Earth orbit. CSA and ESA are reviewing their career limits.

Several governmental space agencies including the partner agencies of the ISS (Canadian Space Agency (CSA), European Space Agency (ESA), Federal Space Agency, Russian Federation (RSA), Japan Aerospace Exploration Agency (JAXA), and National Aeronautics and Space Administration (NASA), as well as China’s CNAS and Russia’s Roscosmos and commercial enterprises are working collaboratively or independently on robotic missions in preparation for crewed cis-lunar or lunar surface missions. The NASA-led Artemis Accords is a growing collaboration of 24 nations at this time that is accelerating the need to develop “best practices” for safe human presence in BLEO [2]. The destination of the November 2022, Artemis I mission around the moon was “cis-lunar”, that is, orbiting under the influence of lunar gravity but not on the lunar surface. The Orion spacecraft carried three human-torso phantoms and numerous radiation detectors to measure the radiation environment inside the vehicle. (Of particular interest to radiation protection, one of the two female phantom torsos wore a vest to demonstrate the ability to shield crew from the ambient radiation environment with far less volume and mass than would be required to surround the entire vehicle.) The Artemis I mission will be followed with Artemis II in which three NASA astronauts and a CSA astronaut will again travel into the influence of lunar gravity on a cis-lunar mission in preparation for a lunar landing at the southern pole of the moon. NASA also envisions construction of an outpost named Gateway [7] that will orbit the moon and serve as a staging station that supports lunar surface missions and exploration missions to Mars. Private/commercial enterprises are also planning crewed missions to support exploration and tourism.

Spaceflight within the Earth’s dipole-driven magnetosphere requires traveling through regions of elevated radiation “belts,” mostly energetic yet “trapped” protons and electrons. At the relatively low orbital altitude (∼420 km) of the ISS, the South Atlantic Anomaly is traversed several times a day and may contribute roughly half of the total relevant exposure of the ISS crew. The belts are far more intense at higher altitudes and can extend out several Earth radii. BLEO mission planners can mitigate travel through the trapped belts by relying on the shielding provided by the vehicle, by rapidly traversing these belts, or, by traveling along trajectories at very high latitudes. “Free space” is a term used by much of the space radiation protection community to refer to locations inside or outside our solar system where the radiation environment is outside the influence of Earth’s geomagnetosphere and beyond the protective shadowing of Earth, the moon, or any celestial body. The omnipresent galactic cosmic radiation (GCR), comprised of energetic protons, He nuclei, and heavy ions of nuclear charge up to 26 (Fe) and higher, are ubiquitous, omnidirectional, and easily penetrate all spacecraft shields. The risk of radiogenic cancer induction increases as GCR exposure slowly cumulates in space and ultimately limits the allowable mission duration.

Solar proton events (SPEs) are formed of solar protons and heavier ions accelerated by eruptive phenomena such as coronal mass ejections or more narrowly directional “solar flares”. The occurrence of such an event is unpredictable but a large SPE is more likely to happen throughout the high activity phase of the solar cycle when the magnetic poles reverse polarity over a roughly 22-year-long cycle. Probabilistic assessments have been performed to inform us of the likelihood that an impactful event can occur during a mission of a specific duration [13], [8] and phase of solar activity [11]. Some SPEs may be higher intensity, but softer in energy, and thus more easily shielded than is GCR. For astronauts to be sufficiently shielded from exposure to soft events, it is important that habitable volumes of the vehicle do not have very thin areas that create “hot spots” where protons that have kinetic energies below ∼50 MeV can leak in. Even the harder, high-energy SPEs are more controlled with thick shields compared with GCR. In this work, the relatively thin shields that were chosen to reflect their effectiveness against an example SPE and relative weak ability to shield GCR.

At this time, no pharmaceutical radioprotectants or mitigators are validated or prescribed for ISS crews, making radiation shielding and mission duration and timing the most effective radiation protection countermeasures.

In this work, a comparison of the ionizing radiation absorbed dose and risk quantities used by various space agencies to characterize example missions in space is described. The results provide insights to the level of agreement with which space agencies perform organ dosimetry and calculate effective dose.

Methods and parameters for example missions

A cis-lunar mission was defined having 30 days duration and located in free space near to but with no shadow shielding from the Earth or Moon. A second scenario was defined as a lunar surface mission that included eight days in free space (outside planetary influences) and an additional 22 days on the lunar surface. To simplify this intracomparison and focus on the environments outside the Earth’s influence, the exposures that occur during transit through the trapped proton and electron belts of Earth were not considered. These may add ∼10% or more to the total absorbed dose for this example mission, although the exposure level is highly dependent upon the vehicle's shielding characteristics and outgoing and Earth-bound re-entry trajectories.

The radiation field assumed included two sources, one from galactic cosmic radiation (GCR) and the other from a hypothetical solar particle event (SPE). For both the cislunar and lunar surface missions, the GCR spectrum for January 2013, a period of (moderate to weak) maximum solar activity was chosen. The intracomparison reflects operational procedures by specifying a time period rather than reference spectra. Spectra for ten of the 58 prominent isotopes of nuclear charges 1≤ Z ≤ 58 that are used in NASA analyses are shown, for example, in Fig. 1.

Figure 1.

Example spectra of incident GCR isotopes used by NASA for the 1–30 January, 2013 time frame.

A reference SPE that occurred in October 1989 was chosen which in fact consisted of a series of several events that occurred on October 19, 22, and 24. For the primary proton spectra of these events, the combined reference spectrum as proposed by Townsend et al. [30] was used, based on the band function parameterization of Tylka et al. [31], as shown in Fig. 2. That choice is consistent with guidance from the National Research Council [17] that NASA should consider abandoning use of the August, 1972 SPE as a “hard event” design spectrum and expand the use of the October, 1989 event, which has was higher proton fluence above 100 MeV. The October, 1989 SPE series is used as the likelihood to the 99^th percentile event for a 30-day mission and approximately 90 percentile event for a 365-day mission [4]. For the lunar mission it was assumed that the SPE occurred during the mission segment when the vehicle is on the lunar surface.

Figure 2.

Reference SPE proton spectrum proposed by Townsend, et al. [30] as represented by the Band function parameterizations that were modeled by Tylka, et al. (2010).

For the cis-lunar mission it was assumed that the GCR field was modified by a simplified shielding configuration comprised of an aluminium sphere with a thickness of 5 or 20 g/cm² (Fig. 3a). At the lunar surface, regolith was used to represent 2π shielding of the moon and an aluminium slab with a thickness of 5 or 20 g/cm² was used to represent shielding material provided by the surface vehicle (Fig. 3b). A review of reference lunar regolith composition [19] and elemental mass percentages were referenced [26]. The chosen vehicle thicknesses provide very thin shielding compared with designed “solar storm shelters” or very large craft such as the ISS where shielding can be well over 100 g/cm² of material in some directions

Figure 3.

Shielding geometry for example (a) cis-lunar mission and (b) lunar surface mission.

Absorbed doses were calculated to red bone marrow (also referred to as blood forming organs (BFO) in adults), lenses of the eyes, skin, and colon. In addition, radiation quality factors used in operations by the participating space agencies were used to calculate weighted doses to the organs and tissues of interest. Some participants relied on fluence-to-dose conversion factors to derive weighted doses to organs and tissues while others used transport models. Each space agency used the risk metrics usually applied in their assessments including cancer incidence and/or mortality as appropriate. Risks were calculated for a male and female astronaut (never-smoker) for an age at exposure of 35 years. The likelihood of an SPE occurrence during the 30-day mission was not considered in this intracomparison.

Individual approaches applied by space agencies

The models, tools, references, and endpoints used by each agency to perform dose and risk assessments are summarized in Table 1.

Table 1.

Models, Tools, References, and Endpoints Used by Agencies for Dose and Risk Assessment. CSA/CNL and ESA results are not presented in this manuscript since they do not represent current practices of those agencies.

Agency	GCR Model	Radiation Transport and Shielding	Dose Assessment	Risk Metrics
JAXA	[15]; W-index 48.6	MC Code PHITS	Fluence dose conversion coefficients, DT/Φ and QICRP60, ICRP Publication 123	Lifetime attributed cancer mortality (DDREF=2);
NASA	Badhwar-O’Neill [29]	NASA HZETRN Transport Code, 1-dimensional v.2020); [26]; Human Phantoms MAX and FAX	Point estimate (rather than mean) of organ absorbed doses calculated for this intracomparison; modified ICRP effective dose framework; QNASA, environment-specific wT from NSCR-2012 model; no DDREF applied to NASA effective dose	(1) Cancer mortality metric: NASA Space Cancer Risk Model NSCR-2012 [5], Cancer mortality and incidence metrics: REID and REIC; Baseline cancer rates for non-smokers of tobacco from 2015 ([3]; DDREF distribution centred on 1.5(2) Reference to NASA’s non-cancer guidelines for both short-term and career exposures
RSA	[15]; Sunspot number 48.3	MC Code GEANT4 v4.10.5p02 [1]	Fluence dose conversion coefficients, ICRP60	Lifetime total radiation risk (%); Mean lifetime reduction (year)

CSA and ESA

Canadian Space Agency (CSA) sought external advice on dose and risk calculation from Canadian Nuclear Laboratories (CNL) for the purpose of this exercise. CNL completed a full analysis of organ doses, effective dose, and an additional, conventional cancer risk parameter, lifetime attributable risk.

Similarly, ESA fully participated in this intracomparison for the same purposes and reported a prospective risk parameter, RADS, as well [32].

Those results from CSA and ESA though very relevant, are not included here since they do not report the current practice for those agencies.

JAXA

For the GCR, JAXA used the particle energy spectrum as described in [15], specifying the mean modulation index W=48.6 to represent the January, 2013 time period. Radiation transport and shielding calculations were done with the Monte Carlo PHITS code [20], [21], [22]. Fluence to dose conversion coefficients (D_T/Φ) and radiation quality factors (Q_ICRP60) as given in Tables A.1-A.31 of ICRP Publication 123 were used [10]. As a health endpoint, JAXA used the lifetime attributed cancer mortality (in %) as described elsewhere in this volume [23].

NASA

The GCR environment was described by NASA with the most recent version of the Badhwar-O’Neill model [29]. The NASA HZETRN code v.2020, [28] was used to perform one-dimensional transport of ions of nuclear charges 1–28. Only the GCR analyses for the lunar surface example mission include neutron albedo and light-ion backscatter from the lunar [26], [24]. The elemental mass percentages reported by Jolliff et al. [12] and Slaba et al. [26] were used to represent the composition of the regolith. The human phantoms MAX and FAX [14], [25] were used to represent either a male or female astronaut. Vehicle shielding was characterized by tracing 1002 rays. To calculate organ dose equivalents and NASA effective dose, a modified ICRP framework was used including NASA radiation quality factors and NASA-derived and tissue/organ weights, w_T, following the NASA Space Cancer Risk Model, NSCR 2012 [5]. Substantial differences exist between NASA’s model and ICRP Publication 103 (2007) that, combined, reduce NASA’s effective dose [23]. For example, hereditary effects from gonad exposures are not included in the NSCR-2012 analysis; skin cancer is omitted; sex-specific organs are specified for males only (prostate) and females only (uterus, ovaries, breasts); and the “remainder” organs are reformulated. Rather than apply fluence-to-equivalent-dose conversion factors to externally incident neutrons, the neutrons’ directly ionizing secondary particles are transported to each organ or tissue of interest and the appropriate NASA quality factor is applied at the dose point of interest. The REID-based tissue/organ weights were defined not from a uniform environment, but, instead, for a representative environment for ISS missions that includes trapped protons and GCR. Table 2 compares the w_T values used by NASA for “GCR Exposures”, a separate set used for SPE (“Non-GCR Exposures”), and the sets defined by ICRP. In all cases, astronauts were considered to be “Never Smokers”.

Table 2.

Tissue/organ weights used by NASA for estimating effective dose. Footnote b describes the values used in this analysis.

	ICRP Publications		NSCR-2012 Methodology
	ICRP 60	ICRP 103	GCR Exposures				Non-GCR Exposures
	Any exposure type or smoking status		Never Smokers		Average Population		Never Smokers		Average Population
Organ/Tissues	both sexes	both sexes	male	female	male	female	male	female	male	female
Oral Cavity	–	–	0.00331	0.00313	0.01582	0.00620	0.003	0.003	0.015	0.006
Salivary glands	–	0.01	–	–	–	–	–	–	–	–
Esophagus	0.05	0.04	0.0210	0.00731	0.0506	0.01033	0.019	0.007	0.048	0.010
Stomach	0.12	0.12	0.1379	0.0898	0.0896	0.0630	0.125	0.086	0.085	0.061
Colon	0.12	0.12	0.1181	0.0971	0.1033	0.0589	0.107	0.093	0.098	0.057
Liver	0.05	0.04	0.0839	0.0553	0.0707	0.0486	0.076	0.053	0.067	0.047
Lung	0.12	0.12	0.212	0.336	0.305	0.512	0.192	0.322	0.289	0.495
Skin	0.01	0.01	–	–	–	–	–	–	–	–
Breast	0.05	0.12	–	0.0867	–	0.0548	–	0.083	–	0.053
Uterus	–	–	–	0.0345	–	0.0227	–	0.033	–	0.022
Ovary	0.2	0.08	–	0.0345	–	0.0227	–	0.033	–	0.022
Prostate	–	–	0.0386	–	0.0221	–	0.035	–	0.021	–
Bladder	0.05	0.04	0.0684	0.0470	0.0791	0.0341	0.062	0.045	0.075	0.033
Brain	–	0.01	0.0243	0.0167	0.0169	0.0103	0.022	0.016	0.016	0.010
Thyroid	0.05	0.04	0.00221	0.00418	0.00211	0.00310	0.002	0.004	0.002	0.003
Leukemia	0.12	0.12	0.21	0.1	0.15	0.07	0.284	0.138	0.194	0.100
Bone surfaces	0.01	0.01	–	–	–	–	–	–	–	–
Other (NSCR)c	–	–	0.0805	0.0867	0.0939	0.0816	0.073	0.083	0.089	0.079
Gonads	0.2	0.08	–	–	–	–	–	–	–	–
Remainder (ICRP)d	0.05	0.12	–	–	–	–	–	–	–	–

^aThe list includes organs for which ICRP provides explicit tissue weighting factors and organs for which RAE reports doses, under the NSCR methodology. RAE provides doses for all listed organs except salivary glands, bone surfaces, breast tissue in males, and gonads, which are treated as explained in text.

^bWeights for GCR exposures are applied to space exposures dominated by GCR. Weights for non-GCR exposures are applied to space exposures dominated by SPE and trapped photons and come from Table 6.3 of NASA TP-2012-217375. For GCR exposures weights are calculated to account for a lower quality factor QF for leukemia and, for solid tumors, weights are obtained by adjusting the non-GCR values by normalization factors of (0.79/716) and (0.90/0.862) in never smokers males and females, and factors of (0.85/0.806) and (0.93/0.900) in males and females from general population, respectively.

^c“Other” refers to organs and tissues included in doses reported by the NASA/SRAG radiation transport codes for remaining organs (i.e., organs other than those with explicitly calculated doses). For these organs and tissues there are no explicit risk models in the NASA/SRAG risk assessment methodology (Section 3).

^d ICRP 60 remainder tissues: adrenals, brain, upper large intestine, small intestine, kidneys, muscle, pancreas, spleen, thymus, uterus. ICRP 103 remainder tissues: adrenals, extrathoracic (ET) region, gallbladder, heart, kidneys, lymphatic nodes, muscle, oral mucosa, pancreas, prostate (males), small intestine, spleen, thymus, uterus/cervix (females).

NASA’s radiation quality factor formalism is dependent upon the nuclear charge and energy of ions such that it is not a single-valued function of linear energy transfer (LET). Relative to the quality factor that weights densely ionizing radiations for the generation of solid tumours, NASA has reduced the quality factor for radiation-induced leukaemia’s. In operational practice, NASA calculates statistical distributions of organ absorbed dose, dose equivalent, and NASA effective dose, and reports mean values. The analysis considers the uncertain distributions of several input parameters including the environment, radiation transport and shielding. Those are sampled along with uncertain parameters that describe the radiation quality, which is considered to dominate the overall uncertainty. No dose/dose rate modifying factor was applied to NASA effective dose. For the purpose of comparison in this work uncertain parameters were single-valued so that deterministic or “point estimates” of NASA effective dose values rather than mean values are reported.

Cancer risk probability distributions were calculated using the NSCR-2012 model [5] and reported in terms of statistical distributions of Risk of Exposure-Induced Death (REID) and Risk of Exposure-induced Cancer (REIC). Transfer of cancer risk from the population of atomic bomb survivors in Japan [18] to the astronaut cohort relies on a combination of additive and relative risk models. Non-targeted effects were not explicitly modelled. Baseline cancer rates were derived from 2015 U.S. adult “background” rates adjusted for never-smokers. The dose and dose rate effectiveness factor (DDREF) is Monte Carlo-sampled as one of the uncertain parameter distributions for REID and REIC, and is centered on 1.5.

The parameters used by NASA for “early” and “late” non-cancer effects are based on Gray-equivalent [Gy-Eq] (for the skin, eyes lenses, BFO, heart) or absorbed dose [Gy] to the hippocampus (for CNS), for comparison with guidance in NASA’s standards for 30-d, annual, and career exposures ([16]. However, Gray-equivalent results are not included in this work. Instead, absorbed dose and dose equivalent are reported for BFO, eye lenses, skin, CNS and the heart and nearby arteries.

RSA

RSA used the GCR model as described in Mathiä et al. (2013) with sunspot number 48.3 and applied the Monte Carlo code GEANT4 v4.10.5p02 [1] for radiation transport and shielding calculations. Fluence to dose conversion coefficients included Q_ICRP60 (ICRP 1991). As risk metrics, RSA used the lifetime total radiation risk (%) and the mean lifetime reduction (years). For details, see the section “RSA Current Practice of Dose and Risk Assessment” in Shavers, et al. [23], this volume.

Results of dose assessments

Table 3, Table 4, Table 5, Table 6 summarize the dose estimates for the cis-lunar and lunar missions including GCR and SPE contributions. Each agency reported results for “GCR only” and “GCR+SPE” and for both shielding geometries (5 g/cm² and 20 g/cm² Al). Each Agency reported doses (absorbed dose and dose equivalent) to the blood forming organs (BFO), lenses of the eyes, skin, and colon. In addition, NASA reported doses to heart adjacent circulatory system, and hippocampus (useful for estimating effects to the cardiovascular system and the central nervous system).

Table 3.

Cis-lunar Mission, GCR only; m/f – male/female; it is noted that “NASA effective dose” was calculated using agency-specific weighting factors instead of using those recommended in ICRP Publication 103 (2007). Furthermore, “effective dose” is calculated here separately for males and females (JAXA and NASA), while the ICRP effective dose is averaged across sexes.

	5 g/cm2			20 g/cm2
Cis-lunar, GCR	JAXA m/f	NASA1)m/f	RSA2)	JAXA m/f	NASA1)m/f	RSA2)
Effective dose (mSv)	28.7/28.8	20.8/23.5	31.73)	25.6/25.6	18.8/20.4	27.53)
BFO dose equiv. (mSv)	27.5/27.9	12/12	29.8	24.9/25.1	12/12	26.6
BFO abs. dose (mGy)	10.2/10.2	8/8	10.5	11.2/11.1	9/9	11.5
Eye dose equiv. (mSv)	35.4/35.4	30/31	38.1	29.0/29.0	25/25	31.2
Eye abs. dose (mGy)	10.1/10.1	8/8	10.6	11.2/11.1	8/8	11.7
Skin dose equiv. (mSv)	35.8/36.0	32/32	38.4	29.5/29.6	25/25	31.3
Skin abs. dose (mGy)	10.2/10.2	8/8	10.6	11.3/11.3	9/9	11.7
Colon dose equiv. (mSv)	27.0/27.4	23/23	26.7	24.6/24.9	21/21	25.3
Colon abs. dose (mGy)	10.3/10.2	8/8	10.4	11.2/11.2	9/9	11.4
CNS equiv. dose (mSv)4)	-	26/27	-	-	22/22	-
CNS abs. dose (mGy)4)	-	8/8	-	-	9/9	-
Heart equiv. dose (mSv)	-	26.0/25.6	-	-	24.1/23.9	-
Heart abs. dose (mGy)	-	8.9/8.9	-	-	9/9	-

NASA reports “NASA effective dose”, as described in the text.

Doses are given as average for male and female

Mean tissue dose is given as an upper limit for effective dose

Dose to hippocampus

Table 4.

Cislunar Mission, GCR+SPE; m/f – male/female; it is noted that “NASA effective dose” was calculated using agency-specific weighting factors instead of using those recommended in ICRP Publication 103 (2007). Furthermore, “effective dose” is calculated here separately for males and females (JAXA and NASA), while the ICRP effective dose is averaged across sexes.

	5 g/cm2			20 g/cm2
Cislunar, GCR+SPE	JAXA m/f	NASA1)m/f	RSA2)	JAXA m/f	NASA1)m/f	RSA2)
Effective dose (mSv)	450/436	405/574	6552)	145/145	163/198	1893)
BFO dose equiv. (mSv)	359/377	325/383	419	133/137	118/128	165
BFO abs. dose (mGy)	237/248	278/327	317	73.4/76.0	98/106	112
Eye dose equiv. (mSv)	998/998	1,763/1,757	1,530	204/204	398/411	261
Eye abs. dose (mGy)	684/684	917/929	1,090	115/115	168/173	184
Skin dose equiv. (mSv)	1070/1080	2470/2550	1,640	204/206	352/352	262
Skin abs. dose (mGy)	722/737	1280/1330	1,170	115/116	187/188	186
Colon dose equiv. (mSv)	342/354	399/468	321	130/133	169/183	126
Colon abs. dose (mGy)	225/233	238/278	248	72.4/74.0	92/99	84.8
CNS equiv. dose (mSv) 4)	-	663/722	-	-	233/245	-
CNS abs. dose (mGy) 4)	-	394/428	-	-	127/133	-
Heart equiv. dose (mSv)	-	485/441	-	-	219/211	-
Heart abs. dose (mGy)	-	267/242	-	-	104/99	-

NASA reports “NASA effective dose”, as described in the text.

Doses are given as average for male and female

Mean tissue dose is given as an upper limit for effective dose

Dose to hippocampus

Table 5.

Lunar Surface Mission, GCR only; m/f – male/female; it is noted that “NASA effective dose” was calculated using agency-specific weighting factors instead of using those recommended in ICRP Publication 103 (2007). Furthermore, “effective dose” is calculated here separately for males and females (JAXA and NASA), while the ICRP effective dose is averaged across sexes.

	5 g/cm2			20 g/cm2
Lunar surface, GCR	JAXA m/f	NASA1)m/f	RSA2)	JAXA m/f	NASA1)m/f	RSA2)
Effective dose (mSv)	19.1/19.1	14.8/16.9	15.82)	16.8/16.8	13.5/14.8	13.83)
BFO dose equiv. (mSv)	18.3/18.5	8.2/8.4	14.9	16.3/16.4	8.2/8.2	13.3
BFO abs. dose (mGy)	7.12/7.10	5.6/5.6	5.3	7.33/7.31	6.0/6.0	5.8
Eye dose equiv. (mSv)	23.2/23.1	20/20	19.0	19.2/19.2	17/17	15.6
Eye abs. dose (mGy)	7.15/7.14	4/4	5.3	7.38/7.37	5/5	5.8
Skin dose equiv. (mSv)	23.3/23.5	23.6/23.7	19.2	19.5/19.5	18.9/18.9	15.7
Skin abs. dose (mGy)	7.19/7.17	5.7/5.7	5.3	7.45/7.43	6.1/6.1	5.8
Colon dose equiv. (mSv)	18.0/18.3	15.9/16.7	13.3	16.1/16.2	14.7/15.0	12.6
Colon abs. dose (mGy)	7.16/7.15	5.6/5.6	5.2	7.37/7.36	6.0/6.0	5.7
CNS equiv. dose (mSv) 4)	-	18.9/19.4	-	-	16.2/16.3	-
CNS abs. dose (mGy) 4)	-	5.5/5.5	-	-	5.9/5.9	-
Heart equiv. dose (mSv)	-	17.7/17.4	-	-	16.3/16.1	-
Heart abs. dose (mGy)	-	6.0/6.0	-	-	5.5/5.5	-

NASA reports “NASA effective dose”, as described in the text.

Doses are given as average for male and female

Mean tissue dose is given as an upper limit for effective dose

Dose to hippocampus

Table 6.

Lunar Surface Mission, GCR+SPE with SPE on lunar surface; m/f – male/female; it is noted that “NASA effective dose” was calculated using agency-specific weighting factors instead of using those recommended in ICRP Publication 103 (2007). Furthermore, “effective dose” is calculated here separately for males and females (JAXA and NASA), while the ICRP effective dose is averaged across sexes.

	5 g/cm2			20 g/cm2
Lunar surface GCR+SPE	JAXA m/f	NASA1)m/f	RSA2)	JAXA m/f	NASA1)m/f	RSA2)
Effective dose (mSv)	173/170	207/292	3282)	60.5/60.7	85.4/104	94.63)
BFO dose equiv. (mSv)	144/151	165/194	210	55.8/57.4	61.1/66.3	82.7
BFO abs. dose (mGy)	87.6/91.7	140/165	159	27.1/27.9	50.4/54	55.9
Eye dose equiv. (mSv)	341/341	934/933	763	83.9/83.8	249/258	131
Eye abs. dose (mGy)	221/221	462/468	545	39.2/39.2	87/90	92.2
Skin dose equiv. (mSv)	360/365	1,240/1,280	821	84.5/85.0	182/183	131
Skin abs. dose (mGy)	231/236	644/665	586	39.3/39.8	95.3/95.5	93.1
Colon dose equiv. (mSv)	139/143	204/239	160	54.9/55.8	89/96.1	62.8
Colon abs. dose (mGy)	84.2/87.0	120/140	124	26.9/27.3	48/51.3	42.4
CNS equiv. dose (mSv)4)	-	337/367	-	-	122/128	-
CNS abs. dose (mGy)4)	-	198/216	-	-	65/68	-
Heart equiv. dose (mSv)	-	257/237	-	-	116/112	-
Heart abs. dose (mGy)	-	141/130	-	-	54/52	-

NASA reports “NASA effective dose”, as described in the text.

Doses are given as average for male and female

Mean tissue dose is given as an upper limit for effective dose

Dose to hippocampus

Effective dose was reported by JAXA and NASA reported the NASA effective dose, as these dose metrics are currently used in operational radiation protection. It should be noted, however, that weighting factors other than those recommended in ICRP Publication 103 [9] were used by NASA. RSA reported “mean tissue dose” as an upper limit to effective dose.

GCR: It is of interest to note that in cis-lunar space (Table 3) and on the lunar surface (Table 5) absorbed dose from GCR in individual organs were roughly consistent across space agencies within about ± 20%. Absorbed doses remained relatively constant or increased when the shield thickness increases from 5 to 20 g/cm². In contrast, differences in dose equivalents across space agencies were somewhat more pronounced than those in absorbed dose to individual organs (in particular to BFO dose equivalent), due to differences in defined radiation weighting factors or quality factors. Dose equivalents in shallow-located tissues and organs, in contrast, usually decreased ∼20% when the shield thickness was increased. The dose equivalent of deep-seated organs also decreased with increasing shield thickness, but less so, resulting in effective dose reduction of ∼10%. (On the lunar surface, the NASA BFO dose equivalent remained roughly constant). Similarly, on the lunar surface, effective dose from GCR fell by 15% each case except that NASA effective dose fell 10% for males.

GCR+SPE: Inclusion of the reference SPE to the cis-lunar and lunar surface environments saw much greater effectiveness of the thicker shield to protect the inhabitants, as shown in Table 4, Table 6. This uniform reduction (by a factor of two or three) of absorbed dose was expected due to the softer kinetic energy spectrum of SPE protons relative to GCR ions. With strong reduction of organ absorbed dose, the organ dose equivalents, and effective dose were also reduced.

The combined differences in the physical transport and formalism for effective dose led to differences between the Agencies of 30% or more when considering GCR only in Cis-lunar environment (free space, Table 3), and nearly 40% when an SPE was included in the comparison. Those differences between results are smaller for the lunar surface analysis (Table 5).

Results of risk analysis

Various risk metrics are used by the space agencies involved in this exercise, so while results are informative, they are not perfectly comparable. Although not used operationally, lifetime attributable risk (LAR) and risk of exposure induced death (REID) for both genders were calculated (though not reported here) for CSA by CNL. Also, ESA investigated radiation attributed decrease of survival (RADS) to explore the potential use of that metric. JAXA reported lifetime attributed cancer mortality (%) for males and separately for females. NASA reported mean and 97.5^th percentile probability statistics for risk of exposure-induced death (REID) and risk of exposure-induced cancer (REIC), which are sensitive to the sex and age at exposure (35y, in this example). RSA reports unique parameters total radiation risk of mortality (TRR) [%] and mean lifetime reduction (MLR) [ΔT, years] for uni-gender (male or female). Table 7 summarizes the reported risk quantities from JAXA, NASA, and RSA.

Table 7.

Risks; LACM – lifetime attributed cancer mortality; REID - Risk of Exposure-Induced Death; REIC - Risk of Exposure-induced Cancer; TRR – total radiation risk; MLR – mean lifetime reduction; m/f – male/female; Scenario 1 – cislunar, GCR; Scenario 2 – cislunar, GCR+SPE; Scenario 3 – lunar surface, GCR; Scenario 4 – lunar surface, GCR+SPE

	5 g/cm2			20 g/cm2
	JAXA m/f	NASA m/f	RSA	JAXA m/f	NASA m/f	RSA
Scenario 1	Cis-lunar GCR
LACM (%)	0.130 / 0.161	-	-	0.116 / 0.143	-
REID (%)1)		0.081 (0.23) / 0.12 (0.35)			0.073 (0.21) / 0.10 (0.30)
REIC (%)1)		0.19 (0.53) / 0.28 (0.80)			0.17 (0.48) / 0.24 (0.68)
TRR (%)	-	-	0.3	-	-	0.3
MLR (y)	-	-	0.1	-	-	0.1
Scenario 2	Cis-lunar GCR + SPE
LACM (%)	2.11 / 2.53			0.662 / 0.823
REID (%)1)		1.61 (4.61) / 3.0 (8.90)			0.65 (1.83) / 1.0 (3.02)
REIC (%)1)		3.60 (9.98) / 7.70(21.8)			1.48 (4.08) / 2.45 (6.90)
TRR (%)	-	-	6.6	-	-	1.9
MLR (y)	-	-	2.1	-	-	0.6
Scenario 3	Lunar surface GCR
LACM (%)	0.0862 / 0.107			0.0759 / 0.0937
REID (%)1)		0.057 (0.16) / 0.086 (0.257)			0.052 (0.15) / 0.075 (0.221)
REIC (%)1)		0.134 (0.37) / 0.20 (0.58)			0.12 (0.34) / 0.18 (0.50)
TRR (%)	-	-	0.2	-	-	0.1
MLR (y)	-	-	0.1	-	-	<0.1
Scenario 4	Lunar surface GCR + SPE
LACM (%)	0.793 / 0.963			0.275 / 0.341
REID (%)1)		0.83 (2.39) / 1.56 (4.66)			0.34 (0.967) / 0.54 (1.58)
REIC (%)1)		1.85 (5.15) / 3.96 (11.32)			0.78 (2.14) / 1.29 (3.62)
TRR (%)	-	-	3.3	-	-	1.1
MLR (y)	-	-	1.1	-	-	0.3

Mean (97.5% percentile)

For each of the risk metrics reported by JAXA, NASA, and RSA, increased shielding shows the same trending that was observed for effective dose: males saw a reduction of 10% to 15% of the effective dose, REID, and REIC for the GCR-only environment, and 60-70% reduction (65% to 70% for females) when including the reference SPE.

Discussion

For each of the four exposure scenarios, the dosimetry results can be compared in terms of organ absorbed doses and dose equivalents, and effective dose. Various factors contributed to the differences observed, including the modelling of transport, and shielding, quality factor parameterization, and assigned tissue weighting factors. JAXA and NASA reported risk results for males or females while RSA provided a single value for males and females combined. The risk metrics differed between space agencies, yet for each environment they trended in a similar manner with respect to shield thickness over the two shielding thicknesses considered.

GCR Only: In general, for the GCR scenario organ absorbed doses were rather consistent across space agencies, despite some differences in modelling of transport and shielding. Organ absorbed doses were relatively constant with slight increases when the aluminium shield thickness was increased from 5 g cm^-2 to 20 g cm^-2. Organ dose equivalents stayed constant or dropped ∼10% to ∼20% with increased shielding thickness. For each depth in cis-lunar orbit, the effective doses reported (in Table 3) by the agencies agreed within <35%, calculated as a ratio of the lowest to highest result. Particle transport methodologies and geometries (slab vs spherical) may contribute to these differences in shielding effectiveness.

Impact of the shielding thickness. For the GCR environment in cis-lunar orbit (treated as “free space”, no shadow shielding of the moon), the effective doses reported in Table 3 were only moderately reduced by additional shielding effective dose ranged between ∼21 mSv (NASA) to ∼32 mSv (RSA) at 5 g/cm² and fell to ∼19 mSv (NASA) and 28 mSv (RSA) at 20 g/cm². At both thicknesses, the RSA result was nearly 50% higher than NASA. The organ absorbed doses compared more closely between agencies, usually within 20% to 30%, again with RSA highest and JAXA in closer agreement with RSA. The lower absorbed dose reported by NASA may originate from multiple sources, including differences in the GCR model predictions during solar maximum, the choice to use the 1D transport option of HZETRN, and difference in nuclear interaction models. For the lunar surface scenario, HZETRN transport included albedo neutrons generated in downstream interactions. That raised the effective dose reported by NASA to closer agreement with RSA while JAXA’s effective dose result is highest. Slaba et al. [27] showed that, for GCR, effective dose eventually increases with depth behind moderately thick (>20 g cm^-2) aluminium shields, a result dominated by the penetrating high-energy protons and multiplicity of successive generations of protons, neutrons, and light ions created in the shield material. Increased shielding reduced the risk metrics for JAXA (LACM) and NASA (REID and REIC) by 10% to 15%.

GCR+SPE: The most significant disagreement between Agency-reported absorbed doses and dose equivalents was a factor of ∼1.8 for the skin when shielding was only 5 g cm^-2. Quality factor accounts for only a small part of that since <Q_skin,_JAXA> = 1.5, <Q_skin,NASA> = 1.9, and <Q_skin,RSA> = 1.4. The large disagreement is probably attributable to the difference in the modelled skin thickness and dose evaluation locations adopted by each agency; JAXA evaluated the mean doses in the target area using the fluence-to-dose conversion coefficients, which could be lower than the corresponding doses at the central location due to the thicker effective shielding thickness. When the SPE is included in the cis-lunar analysis, increasing the shield thickness 5 g cm^-2 to 20 gcm^-2 of aluminium reduced the organ absorbed doses and dose equivalents substantially and frequently improved the effective dose agreement between Agencies. The effective doses for JAXA and NASA males are lower than reported by RSA. Increasing shielding to 20 g cm^-2 of aluminium on the lunar surface dropped the effective doses by 70% for JAXA and RSA, 65% for NASA females (cis-lunar), and 60% for NASA males. For females, RSA reports the highest effective dose at 5 g cm^-2 while NASA is highest amongst Agencies at 20 g cm^-2.

Increased shielding reduced each risk metric (LACM, REID, REIC, TRR) by 65% to 70%, much more than seen for GCR. Such a level of reduction was expected since the energy spectrum for the reference SPE is softer and thus more easily shielded than GCR.

Male/female Results. It is of interest to consider the sex-specificity of the effective dose and risk metrics. Since the results for females are quantitatively higher, these ratios are displayed as female:male (f/m) in Table 8 for JAXA and NASA. JAXA reports essentially no sex-dependence in effective dose, and a f/m ratio ∼1.2 for LACM for each example scenario. RSA averages effective dose and total radiation risk (TRR) over both sexes so the ratios are 1:1 by definition. The NASA female to male (f/m) effective dose ratio is 1.13 when only GCR is thinly shielded, then falls to 1.09 at 20 g cm^-2. (These compare with the female:male ratio ∼1.08 for effective dose aboard the heavily shielded ISS that NASA typically sees.)

Table 8.

Ratio of female to male (f/m) results for each example mission scenario reflects the differences between the sex-specific results. It is noted that “NASA effective dose” was calculated using agency-specific weighting factors instead of using those recommended in ICRP Publication 103 (2007). Furthermore, “effective dose” is calculated here separately for males and females (JAXA and NASA), while the ICRP effective dose is averaged across sexes.

	5 g/cm2		20 g/cm2
Cis-lunar, GCR	JAXA	NASA	JAXA	NASA
Effective dose ratio (f/m)	1.00	1.13	1.00	1.09
LACM ratio (f/m)	1.24		1.23
REID ratio (f/m)		1.48		1.37
REIC ratio (f/m)		1.47		1.41
Cislunar, GCR+SPE
Effective dose ratio (f/m)	0.97	1.42	1.00	1.21
LACM ratio (f/m)	1.20		1.24
REID ratio (f/m)		1.86		1.54
REIC ratio (f/m)		2.14		1.66
Lunar Surface, GCR
Effective dose ratio (f/m)	1.00	1.14	1.00	1.10
LACM ratio (f/m)	1.24		1.23
REID ratio (f/m)		1.51		1.44
REIC ratio (f/m)		1.54		1.50
Lunar Surface, GCR+SPE
Effective dose ratio (f/m)	1.00	1.41	1.00	1.10
LACM ratio (f/m)	1.21		1.24
REID ratio (f/m)		1.88		1.59
REIC ratio (f/m)		2.14		1.65

The f/m effective dose ratio reported by NASA increases to ∼1.4 at 5 g cm^-2 when the reference SPE is included, and that ratio dropped with increased shielding. It is important to note that NASA ([16] uses only the effective dose calculated for females to compare with the NASA astronaut career limit; REID and REIC values are reported to inform the astronauts and other stakeholders. Exposure to the reference SPE increases the f/m REID ratio to as high as ∼1.9 and the REIC ratios are slightly higher than that. The f/m REID ratio decreased slightly with increased shielding.

These results indicate that use of sex-averaged effective doses and risks, although used for radiological protection on Earth, may have to be reconsidered when it comes to radiological protection in space.

Conclusions

Comparison of common dose metrics for these example cis-lunar and lunar surface missions described in this study shows the level of agreement across space agencies on the calculation of organ doses, dose equivalents and effective dose, and highlights differences that arise from implementation approaches of the ISS Partner Agencies. Differences between the Agency results arise from combined differences between nuclear physics interactions methods and particle transport methods, self-shielding of human phantoms, quality factor parameterizations and tissue/organ weights.

The example missions specify a weak GCR environment during a certain time frame (January, 2013) and a reference SPE spectrum that represents a rare and intense event. For the GCR scenario, calculated results disagreed the most for absorbed dose in deep-seated organ and tissue (colon, BFO) but the disagreement was rather small. In contrast, the level of disagreement was larger for shallow organs and tissues (eye and skin) when an SPE was included (see JAXA results); the skin absorbed dose varied by well over a factor of two. The trends of organ absorbed doses and organ dose equivalents are dissimilar when increasing shield thickness from 5 to 20 g cm^-2.

Agreement on effective dose was best (between 17% and 35%) when an SPE was not included. When an SPE was included the differences between results were larger (between 23% and 48%). Differences are caused by numerous factors with radiation transport, quality factor, and tissue/organ weights playing dominant roles. The radiation quality factor and its uncertainty are subject to significant revision [5], [6] and have a direct impact on calculated effective dose and risk. Tissue/organ weights are also important to response. The lung, for example, dominates the effective dose in the NASA model and the uncertainty in dose response for lung cancer is significantly broad. Dose or dose rate effectiveness factors were not explicitly applied to effective dose the results of NASA and JAXA reflect sex-specific effective dose; however, the results of JAXA have very small differences between female and male.

Relative to the shielding thicknesses, the results of risk metrics, LACM, REID, REIC trended in a similar manner though quantitatively more sensitive. LACM, REID and REIC reflected female/male specificity.

During a mission, circumstances might impose a higher-than-anticipated exposure or exposure rate. For instance, a mission may be extended in duration or temporarily repositioned to a higher altitude in Earth orbit. An SPE can suddenly increase anticipated exposures above pre-mission projections. Although the likelihood of one or more SPEs occurring was not considered here, that likelihood increases for longer duration missions. A metric that has not been formalized is the probability of a solar particle event occurring during a specific time period or mission segment. That probability is dependent on both the timing and duration of the mission segment. Kim et al., [13], Jiggens et al., [11] and others have considered the probability of SPEs of magnitude (total proton fluence above 10 or 30 MeV) for different time intervals

For persistent reasons, exposure limits have not been fully harmonized between ISS Partner Agencies. A metric similar to Permissible Mission Duration used by NASA for ISS missions may be useful for planning and management of multinational crews on a BLEO mission. The quantity would define an “allowable duration” of a specific segment of an exploration-class mission. For example, permissible segment duration (PSD) could be defined to estimate the maximum duration that a specific phase of a specific mission could be extended before any crewmember’s cumulative exposure exceeds an exposure limit. PSD could be useful in the operational setting particularly when Agencies impose different risk modelling algorithms, different risk uncertainty tolerances, or even different exposure limits. For medical privacy concerns, the PSD metric could be useful to de-identify the most “dose restricted” individual of a crew.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.