Solar Proton Activity over the Solar Cycle 24 and Associated Space Radiation Doses

Abstract

The least number of proton events and ground-level enhancements was recorded in the solar cycle 24 which corresponds with the least smoothed sunspot number compared to the last three previous solar cycles. This was attributed to the weak sun’s polar field and decreasing strength of the interplanetary magnetic field at the start of the solar cycle. The majority contribution to background radiation dose within our earth’s atmosphere is galactic cosmic rays and trapped particles in the Van Allen Belts. However, solar proton events cause sudden spikes in radiation doses, and this depends on the fluence and energy spectra of the events. While these doses are least detected in the lower atmosphere, they have significant radiation damage to spacecraft electronic components and astronauts on long space missions and at higher atmospheric altitudes. Therefore, the prediction of such events and estimation of their effective radiation damage is an important consideration for planning long space missions and spacecraft design materials.

1. Introduction

The sun’s activity varies over the 11 years of a solar cycle with active and quiet phases. Over this period, several energetic particles are released into the interplanetary medium and accelerated towards the earth along the interplanetary magnetic field lines by several mechanisms discussed by Reames [1]. The sudden increase of energetic particle flux is detected by particle detectors on several satellites as solar proton events (SPEs). The proton events can also be detected by neutron monitors on ground stations as suppression of galactic cosmic ray (GCR) intensity (Furbish decrease), mostly observed on geomagnetic storm days [2]. For such proton events, they are defined as ground level enhancements (GLEs) and are associated with a ‘hard’ spectrum of energies >500 MeV [3]. Gradual SPEs are driven by CMEs and interplanetary shocks with a high proton to electron ratio while impulsive events are X-ray flashes associated with solar flares lasting for a few hours to about a day and with a high electron to proton ratio [1, 4, 5]. A few energetic protons from SPEs can escape the geomagnetic shielding and penetrate through the open magnetic field lines at the polar regions of the earth’s atmosphere. These deposit their energy on spacecraft and satellite components at higher orbital altitudes causing space hazardous effects like spacecraft charging, single event upsets, and high radiation levels [6, 7]. The high radiation doses can also be detected by dosimeters onboard aircraft flying at high aviation altitudes mostly during intense proton and GLE events despite the atmospheric mass shielding [8, 9, 10]. GCRs consist of protons, alpha particles, and heavier nuclei which interact with the atmospheric constituents to produce secondary particles that can be observed at the ground level station by neutron and muon detectors. They have high energies ranging between 1 and 20 GeV, most of these energetic particles are deflected away by the magnetosphere and earth’s magnetic field. The few penetrating particles contribute to background radiation levels within the earth’s atmosphere. The important property of GCRs is the magnetic rigidity which defines their ability to resist bending in a magnetic field; this value varies from 0 at the poles to about 17 at the equatorial latitudes [11]. Therefore, the radiation doses due to GCRs are more enhanced at the poles than at the equator. Solar activity is a modulation factor for the propagation of GCRs and this is defined by an anti-correlation relationship between GCR intensity and sunspot number [12].

In this paper, we study the proton event, solar and cosmic ray activity during the solar cycle 24 and relate them to space radiation doses. We study and discuss the effect of particle fluence and energy spectra on dose contribution and some of the effects of high space radiation doses. This is an important consideration for planning long space missions and designing appropriate shielding materials for spacecraft manufacturers.

2. Solar proton events and data

The energetic proton events are detected by several detectors onboard the GOES spacecraft at Geosynchronous orbit [13] and the list of proton events is available at NOAA SWPC (https://ngdc.noaa.gov/stp/space-weather/interplanetary-data/solar-proton-events/SEP%20page%20code.html). The integrated energies for 5-min averages >10 MeV, measured in particle flux units (pfu) defined the proton flux. The start of the SPE was considered when the first three consecutive fluxes were ≥10 pfu and the end was the last time when the flux was greater than or equal to 10 pfu. The last three events in 2017 were picked from the list of Solar Proton Event Archive by European Space Agency (ESA) (https://space-env.esa.int/noaa-solar-proton-event-archive/). Each proton event was verified in the GLE list available at Neutron Monitor Database Event Search Tool (NEST) (https://www.nmdb.eu/nest/gle_list.php) to confirm if it’s associated with the GLE.

2.1 SPE relationship with solar and GCR activity

SPEs are accompanied by shock and flare components and their observation depends on the connectivity of the observer to the flare site. The X-ray flares are used to classify the size of the events with X and M-class being the most powerful and occurring more frequently during the active phase of the solar cycle. SPEs occurrence peaks during the solar maxima with frequent spikes of proton flux occurring due to super active regions associated with large successive eruptions. Out of the total 42 recorded SPEs, only 6 were associated with weak C-class flare eruption while the rest were associated with X and M-class flares. The C-class flares associated SPEs occurred mostly during the ascending phase of the solar cycle. Most energetic events with pfu exceeding 100 pfu occurred during the solar maxima/active periods (86% of the events with >100 pfu occurred from 2012 to 2015). Figure 1 shows the bar graph of the yearly number of SPEs from 2010 to 2017 with the yearly averaged sunspot number. The number of SPEs for the four consecutive solar cycles were 59, 72, 79, and 42 for solar cycles 21, 22, 23 and 24 respectively (https://ngdc.noaa.gov/stp/space-weather/interplanetary-data/solar-proton-events/SEP%20page%20code.html) while the maximum smoothed sunspot number is 232.9, 212.5, 180.3, and 81.8 for the four solar cycles respectively (https://wwwbis.sidc.be/silso/datafiles) [14]. The sharp decline of sunspot number from solar cycle 23–24 corresponds with the least number of SPEs in solar cycle 24. The occurrence rate of GLE events follows a similar pattern with 18%, 20%, 22%, and 2% of the total GLE events (72) for the 4 solar cycles respectively. The least numbers of SPEs and GLEs were recorded in the solar cycle 24 with the least maximum smoothed sunspot number. Solar cycle 24 has been ranked as the fourth weakest of the 24 solar cycles since 1755 and the weakest in 100 years (https://www.weather.gov/news/201509-solar-cycle) [15]. This was caused by the impaired growth of the polar field. The simulation of the solar surface field by Jiang and Schüssler [16] showed emerging low latitude bipolar regions with an opposite orientation of the magnetic polarities in the north–south direction. This was the cause of growth impairment of the polar field and hence resulted in a weak magnetic field throughout the solar cycle 24. Figure 2 shows the variation of sunspot, proton flux at energy levels of >10, >30 and > 60, X-ray intensity, and GCRs intensity recorded by Oulu neutron monitor from 2010 to 2017. The intensity of GCRs recorded by neutron monitors shows an anti-correlation relationship with the rate of particle flux entering the earth’s atmosphere and sunspot number [12, 17].

Figure 1.

A bar graph showing the yearly averaged sunspot number and yearly number of SPEs from 2010 to 2017.

Figure 2.

The variation of sunspot number (a), proton flux at three energy levels (b), integrated X-ray flux (c), and cosmic ray intensity recorded at Oulu neutron monitor station (d) over solar cycle 24.

2.2 Ground level enhancements

GLE events have more concerns with space radiation due to a harder spectrum and large energy intensity. Mewaldt et al. [18] estimated an average power-law index of about 3.18 at energies of >40 MeV and energy intensity of >50 MeV for GLE events in solar cycle 23. He also found out that these events exhibited larger ratios of $\mathrm{Fe}/\mathrm{O}$ ratios and $3\mathrm{He}$-rich element composition. Table 1 shows the detailed properties of the two proton events associated with GLE that occurred during the solar cycle 24. The properties of the GLE events were obtained from Gopalswamy et al. [19]—GLE 71 and Cohen and Mewaldt [20] and Gopalswamy et al. [21]—GLE 72. It’s observed that the GLE events are associated with high CME shock speeds, powerful X-ray flares and originated from the western hemisphere. This provides a better magnetic connection between the sun’s active region and the observer along the sun-earth line [1, 22, 23].

#	GLE 71	GLE 72
Start time	2012 05/170210	2017 09/101625
End time	2012 05/170430	2017 09/160455
>10 MeV maximum pfu	255	844
Region	11,476	12,673
Location	N11W76	S14W74
Flare class	M5.1	X8.3
CME speed (km/s)	1582	3163
Flare time	01:25	15:35
CME time	01:48	16:00
Type II	Yes	Yes
Max energy	>700	>700
Shock height	3.06 Rs	3.92 Rs
% enhancement	18.6	6

Table 1.

A table showing the properties of proton events associated with GLE in the solar cycle 24.

3. Radiation doses within the near-earth space environment from SPEs

Energetic particles driven by CME shocks and flares and GCRs that escape the magnetospheric shielding through the polar regions get trapped by the earth’s magnetic field in the Van Allen Belts (VAB). The outer belts consist of energetic electrons with energies up to $\approx$10 MeV while the inner belt consists of electrons and protons with energies ranging from KeV to $\approx$10 MeV [24, 25]. Over the South Atlantic Anomaly (SAA) which extends from South America to West Africa, the VAB dips down to about 200 km into the upper atmospheric region; the increased energetic particle flux in this region contributes the largest radiation dose during low inclination flights and space missions. The secondary particle showers from the ionization of GCRs contribute mostly to the background radiation levels within our earth’s atmosphere throughout the solar cycle. Several spacecraft and satellites launched into different earth orbits and altitudes have inbuilt particle detectors for detecting the energetic particle flux. The energy deposited by these particles is evaluated by models developed using either radiation transport equations or Monte-Carlo calculations to calculate the amount of absorbed energy as radiation dose to the spacecraft.

Figure 3 shows the dose recorded by a micro-dosimeter from 2009 to 2017 on CRaTER (https://prediccs.sr.unh.edu//data/craterProducts/doserates/data/2017365/doserates_micro_2017365_alldays_allevents.txt). The sudden increase/spikes in dose were due to protons events which were more frequent during the solar maxima periods. The CRaTER is an experiment on the Lunar Reconnaissance Orbiter (LRO), a NASA spacecraft orbiting the moon. CRaTER characterizes the radiation environment with measurement of effects of ionizing energy loss on silicon solid detectors (three pairs of thin and thick silicon detectors and a micro dosimeter) due to penetrating energetic particles and GCRs. The micro dosimeter is on the analog electronics board with the aluminum shield of a thickness of $2.28\mathrm{g}$/cm² facing space [26].

Figure 3.

Dose recorded by a micro-dosimeter on CRaTER from 2010 to 2017.

Figure 4 shows the correlation relationship ($R$) between dose from 2010 to 2017 and GCR intensity and sunspot number. GCR intensity was got from Oulu Neutron monitor in the NEST database (https://www.nmdb.eu/nest/) and sunspot number was got Omniweb database (https://omniweb.gsfc.nasa.gov/form/dx1.html). Figure 4a clearly shows that there is a high positive correlation between the dose and the GCR intensity while Figure 4b shows a significant negative correlation between dose and sunspot number. Solar activity modulates GCRs with 90% of the GCRs filtered out by heliospheric solar wind [27]. This effect decreases with solar wind pressure decrease and this results in change in the controlling heliospheric transport processes. The heliospheric transport processes (diffusion, convection, and adiabatic acceleration) are dominated by cosmic rays during solar minima however this shifts to solar wind dominated transport processes during the solar maxima. This suppresses GCRs propagation, hence a decrease in count rate is observed on the ground neutron detectors drinng solar maxima. The correlation relationship between dose and GCRs and SSN conforms to an anti-corelation relation observed between GCRs and sunspot in Figure 2, this has been studied by a number of researchers [28, 29]. It should be noted that the modulation of GCR intensity does not only depend on solar activity but also on interplanetary magnetic field which further modifies the transport processes [29]. We can observe from Figure 4b that the sunspot number has a less effect on space radiation but only a factor that describes solar activity variation and therefore has a positive relationship with occurrence of proton events.

Figure 4.

Correlation plot between micro-dosimeter reading from 2010 to 2017 and cosmic ray intensity (a) and sunspot number (b).

3.1 Effects of radiation doses

Radiation damage effects are discussed extensively by several scientists including [7, 9, 30] and references therein. In transistors, the accumulation of total ionizing dose (TID) builds up positive charges on the gate oxide of field-effect transistors causing threshold shifts and off-state leakage currents. At very high dose rates, shifts become large enough to exceed the threshold, this results in defective and poor responsiveness of the critical spacecraft component [30]. In semiconductors, the deposited energy by non-ionization processes is quantified as the displacement damage dose (DDD). The DDD displaces electrons from their initial positions forming vacancies and interstitials—Frenkel pairs. The localized groupings of Frenkel pairs result in the formation of material defects which alters the material properties [6, 31]. This affects satellite components mostly utilizing semiconductors; the most observable effect is the loss of maximum power output due to the degradation of solar cells. Hands et al. [30] evaluated degradation performance on solar cells by evaluating cumulative damage effects of TID and DDD on a Galileo satellite and concluded that the most damage effects of extreme space weather come from trapped electrons rather than solar protons.

Astronauts on long space missions may be affected by cumulative radiation effects and exposed to short periods of high radiation levels during proton events. The radiation effects on humans are either deterministic resulting from immediate effects due to high sudden dose levels or stochastic associated with long-term effects of abnormal tissue growth and cancer arising from cumulative doses. Previous studies using space environment radiation models have calculated an average total body dose of $\approx$ 0.5–1.5 μGy/min for space missions which may increase in case of SPE occurrence [32, 33]. In comparison, the dose rates are $\approx {10}^8-{10}^{10}$ times lower than the medical exposure dose rates (8, 4 and 2 mGy for abdominal computed tomography, mammography, and radiotherapy sessions respectively). However, radiation-induced cataracts, DNA damage, cancer risks, cell damage, chromatin decondensation are possible biological effects from repeated doses and nonlinear low dose effects [33]. The estimation of biological space radiation damage is more complex attributed to the probabilistic nature of SPE occurrence, very low doses, individual susceptibility and tissue response to radiation, quantification of secondary radiation, and limited dose data for spacecraft crews on long space missions to develop more representative models for radiation effects.

Particle fluxes of secondary cosmic radiation at 18 and 12 km flight altitude are about 500 and 300 times greater than at sea level respectively [34]. These are the main sources of background radiation at aviation flight altitudes. However, very large and intense SPEs and GLEs may increase the radiation exposure levels to exceed the recommended threshold safety levels (effective dose limits of 20 mSv/year averaged over five years for radiation workers and 1 mSv/year for the public) by the international radiation protection community and the International Civil Aviation Organization (ICAO) [35]. Dosimeter measurements and model simulations have shown an increase in radiation levels during specific proton and GLE events, for example, radiation dose levels on Lufthansa flights between Munich and Chicago rose from 3.4 to 5.7 μSv/hr. during GLE 65 [36] while on Qantas 747 flight from Los Angeles to New York City, the radiation levels increased from 3.4 to 4.7 μSv/hr. during the GLE 66 [37], simulated effective dose rates at 12 km altitude were 5.8 μSv/hr. for GLE 70 and 4.5 μSv/hr. for non-GLE SPE (9th November 2000) [8]. Mishev and Usoskin [38] estimated about 100 μSv for a polar flight at 12.5 km altitude for 3.5 hours from during GLE 72 using rigidity spectrum. While the modeled and simulated doses obtained may not be the actual effective radiation doses to the aircrew but it’s indicative of a possibility of radiation hazard on the aviation industry associated with SPEs. A survey on radiation exposure to Canadian pilots showed an annual dose of about 3 mSv [39]. However, the estimation of total effective radiation doses to the aircrew requires consideration of flight path, time of flight, flight altitude, the position of the aircrew within the aircraft, and material component of the aircraft (shielding design). The International Commission on Radiation Protection (ICRP) recommendations in 2007 recommended that the aircrew be considered as occupationally exposed workers with the implementation of practical regulatory measures including monitoring programs and individual dose assessment for radiation protection by aviation operators [40]. Several warning systems of aviation dosimetry [10, 41, 42] have been developed in addition to S-scale provided by NOAA SWPC alerts the aviation operators to initiate the safety protocols which include rerouting, flying at low altitudes, avoiding polar regions, and flight delay. However, they are challenged by the probabilistic occurrence of SPEs and the description of the broad energy spectrum from low to high energies [10].

3.2 Quantification of space radiation doses from SPEs

Energetic particles incident on spacecraft and satellite materials deposits energy by ionization process, these particles include proton, trapped protons, and electrons in the VAB, secondary photons, and GCRs. The energy absorbed is measured as TID of which cumulative effects can result in device failure of critical spacecraft electronics and biological damage to astronauts. TID is quantified by absorbed dose (SI unit called rad, that is 1 gray (Gy) = 100 rads = 1 J/kg). The absorbed dose ($D$) is a function of fluence spectrum at different energy levels ($\varnothing$) and the mass stopping power of the material ($\mathrm{d}E/\rho \mathrm{d}x$) through which the incident particle penetrates (Eq. (1)); this is dependent on the orbit altitude, time taken in the orbit, and spacecraft/satellite orientation [43].

$$D=\varnothing \frac{\mathrm{d}E}{\rho \mathrm{d}x}\left(1.6\times {10}^{-10}\right)\mathrm{Gy}$$E1

The Bethe-Bloch formula [44] defines stopping power using relativistic quantum mechanics (Eq. (2)):

$$\frac{\mathrm{d}E}{\mathrm{d}x}=\frac{4\pi z^2{k}_o^2{e}^4}{m{v}^2}\left[ln\frac{2{\mathit{mv}}^2}{I}-ln\left(1-\frac{v^2}{c^2}\right)-\frac{v^2}{c^2}\right]$$E2

where $z$ is the atomic number of the heavy particle, $e$ is the magnitude of the electron charge, $m$ is the electron rest mass, $c$ is the speed of light, $I$ is the mean excitation energy of the medium, $v$ is the velocity of the particle and $k_{\mathrm{o}}$is the Boltzmann constant.

3.3 Fluence and energy spectra dose relationship

Space Environments and Effects section at the European Space Agency maintains an archive of SPEs with plots and fit data for event fluence and radiation effects (https://space-env.esa.int/noaa-solar-proton-event-archive/) from 1997 up to date. The event selection is based on the NOAA/GOES-p_ 5 min averaged >10 MeV p + flux exceeds 2.0 p+/cm²/s/sr and ends when the flux returns to below 1 p+/cm²/s/sr. The estimation of radiation effects is done using the best fit from a comparison of three forms of fit that is exponential in power rigidity$(A\ast {\mathrm{e}}^{-\raisebox{1ex}{$\mathrm{Rm}$}\!\left/ \!\raisebox{-1ex}{$B$}\right.}$), exponential in energy $\left(A\ast {\mathrm{e}}^{-\raisebox{1ex}{$E$}\!\left/ \!\raisebox{-1ex}{$B$}\right.}\right)$ and power law $\left(A\ast E^B\right)$, where $A$ and $B$ are constants, $\mathrm{Rm}$ is magnetic rigidity and $E$ is energy. Figure 5a–f shows the correlation relationship ($R$) of the SPEs integrated fluence spectra from 2010 to 2017 at different proton energies of 1, 5, 10, 30, 50 and 100 MeV with the dose behind 0.05 mm Al shield in rads. The $R$ values show that the integrated fluence spectrum at all proton energies has a positive correlation with the magnitude of the dose. However, the integrated fluence at low energies from 1 to 30 MeV is more correlated to dose with high $R$ values compared to higher energies from 50 to 100 MeV. This observation is a manifestation of the contribution of slowed-down proton spectra to dose build-up on a material.

Figure 5.

Scatter graphs with line of best fit showing correlation relationship between integrated fluence spectrum and dose at different energy levels; (a) 30 MeV, (b) 50 MeV, (c) 100 MeV, (d) are 1 MeV, (e) 5 MeV and (f) 10 MeV.

4. Discussion

4.1 Fluence dose relationship

Dose distributed over a material results from a specific number of particles incident on a given material surface. Using the simplified equation of dose,$D={\varnothing }_p\left(E\right)S_{\mathrm{p}}\left(E\right)$, where $S_{\mathrm{p}}$ corresponds to non-ionizing energy loss (NIEL) for DDD and stopping power for TID. The incident primary fluence interacts with the material through soft and hard knock-on collisions resulting in slowed down secondary charged particles. This develops a charged particle equilibrium with the absorption of the secondary fluence as the radiation energy escapes [45]. There exists a linear relationship existing between proton damage coefficients and determined NIEL for the case of GaAs cells [46]. The observable relationship results from the DDD increasing with the increase in proton fluence. Therefore, proton fluence at any energy level is an input to calculating degradation curves and or characteristic curves from which relative proton damage coefficients can be determined.

4.2 Energy spectra dose relationship

While the less energetic protons incident on the material is shielded off, the highly energetic protons (>100 MeV) are slowed down by the shielding of the material to low energies. The slowed-down protons initiate atomic and molecular collision; this process needs a build-up from low energies to exceed the threshold energy for atomic displacement to occur. Once the threshold energy is exceeded, the material properties start to degrade with absorption and particle production. Comparing the rate of energy loss, the NIEL peaks faster at much lower energies than the stopping power. While different materials have different threshold energies for atomic displacement, low proton energies (<1 MeV) are enough to build the threshold energies for atomic displacement for most shielding materials and this plays the largest contribution to damage production. The defining functional form of the energy spectrum below 1 MeV behaves as a reciprocal of stopping power calculated using the continuous slowing down approximation (CSDA) theory and is independent of the shield thickness. Theoretically, this is the range of the proton ($\mathrm{Ra}$) which defines the mean distance a proton travels in the matter before it stops (Eq. (3)) below adopted from [6].

$$\mathrm{Ra}=\underset{E_{\min }}{\overset{E_{\max }}{\int }}-\frac{\rho \mathrm{d}x}{\mathrm{d}E}\mathrm{d}E+\mathrm{Ra}\left(E_{\min }\right)$$E3

Where $\mathrm{Ra}\left(E_{\min }\right)$ is the measured range at minimum energy ($E_{\min }$) always taken at 1 MeV due to large enough data available for any proton event at the energy level.

Messenger et al. [47] found that 90% of the total dose to the shielded device was a contribution of relatively low proton energies up to about 12 MeV and only 10% was from proton energies from 12 to 500 MeV for an event on 19th October 1989. Extremely large SPEs associated with GLEs rarely occur, these contain high proton energies extending to GeV. The energy spectrum from such events is known to be ‘hard’ and it can penetrate deeper thicknesses of the shielding material. Mertens and Slaba [48] calculated the cumulative dose from a set of 65 historical SPEs associated with GLEs using the power-law function and noted that 25% of the peak total dose was contributed from energies >500 MeV corresponding to a spectrally hard SPE of February 1956. The double power law of the band distribution function has been recommended to best describe the energy spectrum of a GLE event over the broad energy ranges [18, 49]. Also, Xapsos et al. [50] found that the Weibull distribution describes the proton event energy spectra for the smallest values and is appropriate on broad energy ranges.

5. Summary

The occurrence of solar proton and GLEs events follow solar cycle described by sunspot number. A least number of SPEs were recorded in solar cycle 24 with the least smoothed sunspot number compared to the previous three solar cycles. Every proton energy starting from the lowest to the highest energy level contributes to radiation dose and the slowed-down protons at low energies are critical in the assessment of space radiation damage. I also noted that the shield of a material is an important component to consider when evaluating proton damage effects. It’s important to use the most appropriate distribution function that can describe the energy spectrum from low to high energy levels for any SPE event. However, this is a challenge with the GLE associated SPEs where there is an underestimation at high energies of GeV by most distribution functions. A further study regarding the probabilistic occurrence of SPEs, with the associated properties of solar flares, CMEs, GLE events, and energy spectrum is necessary, this is important to improve modeling and prediction capabilities of such events.

Acknowledgments

I acknowledge several data providers which include NOAA SWPC, ESA, NEST, SILSO, Omniweb, and NASA.