The Geant4 Lunar Albedo Computed Environment Model with Temperature Variations (GLACE-T)

Approved for public release. OTR 2025-00924. README for the GLACE-T Model (GLACE with Temperature Variations) Mark D. Looper, The Aerospace Corporation Version 1.0, July 1, 2025 1. Di&erences from the GLACE Model In previous work, we performed extensive Geant4 simulations (Agostinelli et al., 2016) of secondary energetic particles, conventionally called albedo, that are produced when the lunar surface is struck by energetic ions arriving from space. The model tabulated all secondary (and backscattered primary) particles except neutrinos that resulted from isotropic bombardment of a lunar regolith target by all ion species from Z = 1 to 28 (H to Ni). There were sixteen lunar-surface target geometries, one representing dry ferroan anorthosite (FAN) and the rest representing a surface layer of hydrogenor waterbearing FAN over a dry substrate, with diVerent concentrations and diVerent layer thicknesses. The results were collected into a large set of JSON files, with individual primary-particle species and energies tabulated separately so that a user can convolve the model output with any desired incident cosmic-ray or solar-particle ion spectra, and the fileset was released on the Zenodo open archive as the Geant4 Lunar Albedo Computed Environment (GLACE) model (Looper, 2025). The impetus for the development of the GLACE model was the interpretation of energetic-particle measurements from the Cosmic Ray Telescope for the EVects of Radiation (CRaTER) aboard NASA’s Lunar Reconnaissance Orbiter (LRO) (Spence et al., 2010), which measures particles in the MeV to GeV energy range. Because the temperature of the simulated regolith target would have no eVect on the spectra of these high-energy particles, GLACE did not take any special account of that, and the simulations were done using the Geant4 default value of 273.15 K from the IUPAC Standard Temperature and Pressure (STP). However, as we subsequently considered the possibility of measuring energetic electrons that would result from the decay of albedo neutrons, the details of the thermal-neutron spectra became of interest. We therefore performed an additional, smaller set of simulations for a dry regolith target with varying temperature distributions, and we are releasing the results as a supplement here, which we are calling GLACE-T. Because the GLACE-T simulations and analyses were based on those from GLACE, most of the detailed description in the README file for GLACE also applies to GLACE-T. We do not repeat that description here; reference should be made to the README file that is available on Zenodo alongside the GLACE model data files (Looper, 2025), and we discuss herein only what has been changed. The biggest diVerence is that, for GLACE-T, we only tabulated albedo neutrons, since target-material temperature would have no eVect on other albedo species, and we only tabulated neutrons coming oV the lunar surface, not at 20 km altitude or at depth as for GLACE. The incident-particle species and energy bins are the same as for GLACE, and the energy and angular bins for the albedo neutrons are the same. However, we performed only a single Geant4 run for each case of target and incident-particle species and energy, and so we do not include standard deviations of the results calculated from multiple runs as we did for GLACE. The target geometries for GLACE-T were also layered; however, in this case all layers consisted of dry FAN, with a lower layer at 250 K and upper layers of nine thicknesses (1 mm to 10 m, spaced logarithmically) having one of six temperatures (100 K to 350 K, spaced linearly). With the upper layer also at 250 K the diVerent thicknesses would be redundant, so we only include a 10-m thickness for that temperature case. Finally, the GLACE README file warns that diVerent versions of Geant4 could produce slightly diVerent results that might distort some of the subtle changes in albedo particle distributions that result from diVerent regolith hydrogenation, if simulation results from diVerent versions of Geant4 were compared. Since we were only interested in intercomparison of the GLACE-T results, not in comparison with the older GLACE results (in particular, we did not include hydrogen or water in the temperature-varying GLACE-T simulations, so we could only compare with the dry GLACE results anyway), we performed the GLACE-T simulations using the then-current version of Geant4, geant4-v11.3.0, rather than the one used for the original GLACE simulations, geant4.10.07.p01 (10.7.1). 2. Structure of the Stored Model Output 2.1 File Names and Organizations Like GLACE, the GLACE-T model is stored as a large set of JSON-formatted ASCII files, which can readily be imported by many data-analysis software packages. These model files are accompanied by some sample results showing realistic neutron fluxes produced using the model, stored as both JSON files and PNG plots, which users can compare with the results of their own calculations to check their importing and use of the model components. All these files are collected into 46 zipfiles, one for each combination of the nine target upper-layer thicknesses and six upper-layer temperatures as described in Section 1, minus the redundancies where the upper and lower temperatures are both 250 K as also discussed there. Each zipfile is about 223 MB compressed, 1.6 GB uncompressed. 2.1.1 Energy/Angle Kernels The GLACE-T model, like GLACE, is stored on Zenodo as a set of JSON files collected into directories and zipped. Because all GLACE-T result files relate to neutrons coming oV the lunar surface, there are fewer files and one less subdirectory level than for GLACE. The model’s files for the energy/angle response kernels 𝑅!" " 𝐸!,𝐸",𝜃" & , as defined in the README file for GLACE, will have names and filepaths like dry_1mm_100k_neutrons/ion_28_58.json The contents of each file will be discussed in Section 2.2, but the directory and file names encode information as well. The base directory here, “dry_1mm_100k_neutrons,” is the directory that was zipped into “dry_1mm_100k_neutrons.zip,” and all such base directories have similarly structured names. The four pieces of the name, separated by underscores, tell what target geometry and particles are addressed by the contents of that directory and zipfile. All these names begin with “dry,” since the GLACE-T simulations did not include hydrogen. Next is the thickness of the upper layer, one of “1mm,” “3mm,” “1cm,” “3cm,” “10cm,” “30cm,” “1m,” “3m,” or “10m,” followed by the temperature of that upper layer in degrees Kelvin, one of “100k,” “150k,” “200k,” “250k,” “300k,” or “350k”. L a st is “neutrons” for all directories and their zipfiles. The filenames themselves identify the incident ion species, with atomic number Z and mass number A for the most abundant isotopes of ions from Z = 01 to 28 (H to Ni), plus 3He; the ion in the example filename above is 58Ni. Thus a single zipfile contains the kernel files for all primary ion species that address the production of neutrons at the lunar surface for a given target temperature profile. 2.1.2 Energy/Angle Examples In each of the subdirectories described in Section 2.1.1, alongside the “ion_ZZ_AA.json” files, are six files that can be used to exercise the model and check that it has been imported into users’ code correctly. There are two JSON files containing spectra from the Badhwar-O’Neill 2020 cosmic-ray model (Slaba & Whitman, 2020) for periods corresponding to the least (solar minimum) and greatest (solar maximum) solar modulation identified in LRO/CRaTER observations over the past solar activity cycle (Looper et al., 2020); the two files, “solmin_gcr.json” and “solmax_gcr.json,” are the same in each of the directories enumerated in Section 2.1.1. Convolving each of the primary-ion spectra in these files (3He is not included in the Badhwar-O’Neill 2020 model output) with the ion’s respective kernel, as discussed in the README file for GLACE, and summing over all primary ions gives the total energy/angle distribution for the neutrons that would be produced from the target geometry addressed by that directory. These distributions are stored in JSON files and plotted as PNG images with names incorporating the solar cycle designation: each directory will contain “solmax_neutrons_surface.json” and “solmin_neutrons_surface.json,” and also *.png files with corresponding names. 2.2 Contents of Model Files The formats of the JSON files for the GCR spectra and examples as discussed in Section 2.1.2 are the same as for GLACE (Tables 4 and 5 in that README file), except that “OBSERVER_ELEVATION” and “SECONDARY_PARTICLE_SPECIES” are always “At lunar surface” and “Neutrons” in the GLACE-T files. Table 1 below gives the format of the kernel files for the individual primary ions; it diVers from the format in Table 1 of the GLACE README file only in that the “RESPONSE_SIGMA” row is omitted, as discussed in Section 1. Table 1 – Energy/Angle Kernel Files Variable Name Variable Type Description LUNAR_TARGET String One of the 46 target layer geometries OBSERVER_ELEVATION String “At lunar surface” only PRIMARY_ION_SPECIES String Values of Z and A for incident ion SECONDARY_PARTICLE_SPECIES String “Neutrons” only GEANT4_ENVIRONMENT String Geant4 version and physics list used GLACE_T_VERSION String Version of GLACE-T model PRIMARY_ENERGY_UNITS String Units of PRIMARY_ENERGY_BINS SECONDARY_ENERGY_UNITS String Units of SECONDARY_ENERGY_BINS SECONDARY_ZENITH_ANGLE_UNITS String Units of SECONDARY_ZENITH_ANGLE_BINS RESPONSE UNITS String Units of RESPONSE PRIMARY_ENERGY_BINS [201] Edges of 200 𝐸!bins SECONDARY_ENERGY_BINS [601] Edges of 600 𝐸" bins SECONDARY_ZENITH_ANGLE_BINS [91] Edges of 90 𝜃"bins RESPONSE [200,600,90] Simulated values of 𝑅!"$𝐸!,𝐸",𝜃"& 3. Acknowledgements This work was supported by NASA under contract NNG11PA03C. 4. References Agostinelli, J., et al., “Recent Developments in Geant4,” Nucl. Inst. And Meth. In Phys. Res. A, 835, pp. 186-225, DOI: 10.1016/j.nima.2016.06.125 (2016) Looper, M. D., “The Geant4 Lunar Albedo Computed Environment (GLACE) Model” [Data Set], Zenodo, DOI: 10.5281/zenodo.8343472 (2025) Looper, M. D., J. E. Mazur, J. B. Blake, H. E. Spence, N. A. Schwadron, J. K. Wilson, A. P. Jordan, C. Zeitlin, A. W. Case, J. C. Kasper, L. W. Townsend, and T. J. Stubbs, “Long-Term Observations of Galactic Cosmic Ray LET Spectra in Lunar Orbit by LRO/CRaTER,” Space Weather 18 (12), e2020SW002543, DOI: 10.1029/2020SW002543 (2020) Slaba, T. C., and K. Whitman, “The Badhwar-O’Neill 2020 GCR Model,” Space Weather 18 (6), e2020SW002456, DOI: 10.1029/2020SW002456 (2020) Spence, H. E., et al., “CRaTER: The Cosmic Ray Telescope for the EVects of Radiation Experiment on the Lunar Reconnaissance Orbiter Mission,” Space Sci. Rev., 150, pp. 243284, DOI: 10.1007/s11214-009-9584-8 (2010)