CALET Search for electromagnetic counterparts of gravitational waves during the LIGO/Virgo O3 run

The CALorimetric Electron Telescope (CALET) on the International Space Station (ISS) consists of a high-energy cosmic ray CALorimeter (CAL) and a lower-energy CALET Gamma ray Burst Monitor (CGBM). CAL is sensitive to electrons up to 20 TeV, cosmic ray nuclei from Z = 1 through Z $\sim$ 40, and gamma rays over the range 1 GeV - 10 TeV. CGBM observes gamma rays from 7 keV to 20 MeV. The combined CAL-CGBM instrument has conducted a search for gamma ray bursts (GRBs) since Oct. 2015. We report here on the results of a search for X-ray/gamma ray counterparts to gravitational wave events reported during the LIGO/Virgo observing run O3. No events have been detected that pass all acceptance criteria. We describe the components, performance, and triggering algorithms of the CGBM - the two Hard X-ray Monitors (HXM) consisting of LaBr$_{3}$(Ce) scintillators sensitive to 7 keV to 1 MeV gamma rays and a Soft Gamma ray Monitor (SGM) BGO scintillator sensitive to 40 keV to 20 MeV - and the high-energy CAL consisting of a CHarge-Detection module (CHD), IMaging Calorimeter (IMC), and fully active Total Absorption Calorimeter (TASC). The analysis procedure is described and upper limits to the time-averaged fluxes are presented.


Introduction
The importance of simultaneous or near-simultaneous multimessenger observations has long been recognized (e.g., Mészáros et al. 2019 andBurns et al. 2019 for recent reviews), and was clearly demonstrated over three decades ago by the correlated neutrino and optical observations of SN1987A (Bionta et al. 1987;Hirata et al. 1987;Shelton 1987;Alexeyev et al. 1988). The observations of the gravitational radiation event GW170817 from a binary neutron star merger (Abbott et al. 2017a) together with the short gamma-ray burst (GRB) GRB 170817A Savchenko et al. 2017) and the optical transient AT2017gfo (Abbott et al. 2017b) have now made it possible to draw physics conclusions about the connection between short GRBs and neutron star mergers, the origin of heavy r-process elements, the speed of gravitational waves, and the nature of kilonovae. Additional observations of short GRBs associated with gravitational-wave events will especially provide information about neutron stars and their mergers, the nature of the GRB jet, and the neutron star equation of state (Burns 2020;Pian 2021). A number of joint gamma-ray/gravitational-wave searches have been carried out since the 2017 event involving LIGO/Virgo and Fermi-GBM (Hamburg et al. 2020;Stachie et al. 2020), Swift (Page et al. 2020;Abbott et al. 2021), INTEGRAL (Mereghetti et al. 2018), AGILE (Verrecchia et al. 2019), and Calorimetric Electron Telescope (CALET; Yamaoka et al. 2017;Adriani et al. 2018). In no case since GRB 170817A/GW170817, however, has there been a confirmed positive GRB signal in association with a LIGO/Virgo gravitational-wave event.
The Japanese-Italian-US CALET cosmic-ray/gamma-ray telescope (Figure 1) was launched to the International Space Station (ISS) on 2015 August 19 and has been in operation on the Exposed Facility of the Japanese Experiment Module (JEM-EF) of the ISS since 2015 October (Asaoka et al. 2018a;Adriani et al. 2019Adriani et al. , 2020Adriani et al. , 2021. The main detector of the CALET payload is a calorimeter (CAL) to observe high-energy cosmic rays and gamma rays above 1 GeV. In addition, the Gamma-ray Burst Monitor (CGBM) covers the gamma-ray energy range from 7 keV to 20 MeV.
The ongoing CGBM and CAL searches for GRBs are described by Kawakubo et al. (2021). The searches for CGBM and CAL counterparts to Advanced LIGO and Advanced Virgo gravitational-wave events during LIGO/Virgo observing runs O1 and O2 are described in Yamaoka et al. (2017) and Adriani et al. (2018). The present paper describes the search by the CALET gamma-ray detectors for prompt GRBs associated with gravitational-wave events during LIGO/Virgo observing run O3. In Section 2, we describe the low-energy and high-energy CALET gamma-ray telescopes. In Section 3 we discuss the analysis procedures and present the results of the CALET GRB counterpart search during LIGO/Virgo observation run O3. Finally, we discuss the results and conclusions in Section 4.

CALET Gamma-Ray Burst Monitor (CGBM)
The CAL and CGBM instruments have gamma-ray sensitivity in different energy ranges. CGBM is primarily responsible for observing prompt emission from GRBs (Kawakubo et al. 2019(Kawakubo et al. , 2021. CGBM consists of two Hard X-ray Monitors (HXMs) and the Soft Gamma-ray Monitor (SGM). Each HXM module consists of a 6.6 cm diameter × 1.3 cm deep and a 7.9 cm diameter × 1.3 cm deep LaBr 3 (Ce) scintillator; SGM consists of a 10.2 cm diameter × 7.6 cm deep BGO scintillator. Both are viewed by photomultiplier tubes (PMTs). Outputs of the PMTs are amplified by two amplifiers with a ratio of gains ∼30 and pulse heights are acquired individually as high-gain and low-gain pulse heights in order to provide the required dynamic range. The two subsystems provide sensitivity to X-rays and gamma rays over the energy range 7 keV-1 MeV (HXM) and 40 keV-20 MeV (SGM), as shown in Figure 2. The detectors have fields of view ∼3 sr (HXM) and ∼8 sr (SGM).
CGBM normally collects two types of continuous monitor data suitable for temporal analysis and spectral analysis: time history (TH) data with 0.125 s resolution and eight energy channels (four channels in high gain and four channels in low gain) and pulse height (PH) data with 4 s resolution and 512 channels (102 channels with high gain and 410 channels with low gain). Energy ranges for the high-gain and low-gain channels in TH data are listed in Table 1. Monitor data are collected every 0.125 s and transmitted to the ground every second. In addition, an onboard trigger system uses the signalto-noise ratio (S/N) calculated every 0.25 s to detect increased count rates: Here Δt is the foreground (signal) integration time; N tot is the number of counts integrated over Δt in the selected energy range; and N BG is the number of background counts in the selected energy range integrated over the background time interval Δt BG preceding Δt.   S/Ns are calculated on board continuously every 0.25 s for Δt BG = 16 s and four signal integration times (Δt = 0.25 s, 0.5 s, 1 s, or 4 s) over the energy ranges 25-100 keV for HXM and 50-300 keV for SGM. An onboard trigger threshold is set at S/N = 8.5 for each HXM and 7.0 for SGM. If any calculated S/N exceeds the trigger thresholds, CGBM captures ∼700 s of individual event data starting ∼200 s prior to the trigger and consisting of event arrival times with a time resolution of 62.5 μs and ADC pulse heights corresponding to the individual energy deposits in each detector. Each event consists of two ADC pulse heights measured by high gain and low gain. When a CGBM onboard trigger occurs, (1) the CGBM event-by-event data with fine time resolution are recorded; (2) the energy threshold of the CAL is reduced from its normal 10 GeV threshold to 1 GeV for approximately 10 minutes (low-energy gamma-ray or LEG mode) to increase the CAL sensitivity to GeV photons from GRBs; and (3) two optical images with an exposure of 0.5 s are taken by the Advanced Stellar Compass (ASC) star sensor (Hudeca et al. 2011) to detect optical flashes during the prompt emission phase. The CGBM event data buffer can store event data from four successive triggers on board. When a fourth trigger occurs before the buffer is emptied, the onboard trigger is disabled until event data are downlinked and deleted. The downlinks of the CGBM event data and optical images are performed three times a week.
A trigger alert system running on the ground server at the Tsukuba Space Center of the Japanese Aerospace Exploration Agency (JAXA) analyzes the real-time monitoring data including housekeeping data, status information, and data settings when a CGBM onboard trigger occurs (Asaoka et al. 2018b). When the alert system notices a CGBM onboard trigger, the real-time TH data are analyzed and a GCN notice is delivered. If real-time data are unavailable due to loss of connection between the ISS and ground, hourly data distributed with at most one hour delay can be used for ground analysis.
As an example, the time histories of GRB 200521A observed in the three detectors of CGBM are shown in Figure 3 (Yoshida et al. 2020). GRB 200521A was also detected by ASIM, AGILE, INTEGRAL SPI-ACS, and Konus-Wind (Marisaldi et al. 2020;Svinkin et al. 2020;Verrecchia et al. 2020). A sky map of GRBs detected by CGBM through 2021 November is shown in Figure 4. Since CGBM has no capability to localize the GRBs, GRB positions were based on the reports to the Gamma-ray Coordinates Network (GCN) 42 by other GRB instruments. Out of 271 GRBs detected by CGBM, 195 were localized by other instruments.
Charged particles trapped by the Earthʼs geomagnetic field cause the CGBM count rates to increase at high geomagnetic latitudes and through the South Atlantic Anomaly (SAA). Therefore, CGBM high voltages are turned off at high geomagnetic latitudes and during SAA passages in order to avoid false triggers and excessive PMT currents due to increased particle fluxes. As a result, the cumulative CGBM live time over the period 2015 October-2021 November is ∼60%.
The Active Thermal Control System (ATCS) of the JEM-EF maintains temperature variations for CAL to within a few degrees. However, CGBM temperatures are not controlled by the ATCS. CGBM temperature variations depend on both the solar beta angle and solar altitude. Temperature corrections can be made on the ground using prominent background lines at 34 keV, 511 keV, 1.47 MeV, and 2.2 MeV.

CALET Calorimeter (CAL)
The primary purpose of CAL is to observe high-energy electrons, protons, and nuclei. In addition, CAL is also  sensitive to gamma rays at 1 GeV-10 TeV (Cannady et al. 2018). The primary CAL trigger mode for cosmic rays and gamma rays is the High-Energy (HE) mode with a threshold of 10 GeV for the observed energy. CAL typically takes data in HE mode, but when the geomagnetic latitude is below 20°( except around the SAA) or following a CGBM trigger, CAL adds LEG mode with the gamma-ray trigger threshold lowered to 1 GeV (Cannady et al. 2018).
CAL identifies particles based on the ionization energy deposited in three separate detector systems: (1) The Charge Detector (CHD) located at the top of CAL consists of two orthogonal layers made of 14 plastic scintillator paddles with individual element charge resolution for particles from electrons and protons to ultraheavy nuclei with Z ∼ 40. (2) Below the CHD, a finely segmented preshower Imaging Calorimeter (IMC) consists of eight double layers of 1 mm 2 cross-sectional scintillating fibers, arranged in belts along orthogonal directions, interspersed with seven layers of tungsten with a total thickness of 3 radiation lengths (r.l.). IMC can observe tracks of incident particles and provides an independent charge determination, fine-grained tracking information, and an image of the initial stage of the shower development.
(3) The Total Absorption Calorimeter (TASC) located at the bottom of CAL consists of lead tungstate (PbWO 4 ) bars arranged in 12 layers with a total thickness of 27 r.l., sufficiently deep to absorb electron showers totally up to TeV energies.
Incident high-energy gamma rays produce Compton electrons and e + e − pairs in the IMC, resulting in electromagnetic showers. Gamma-ray events are distinguished by the absence of signals in CHD and the top layers of IMC and by the consistency of the observed shower profile in the IMC and TASC with an electromagnetic rather than a hadronic shower. The three CAL subsystems are shown in Figure 5 together with examples of the CAL response to electrons, protons, nuclei, and gamma rays. Incoming gamma-ray directions are checked to eliminate events that might have crossed ISS structures in the CAL field of view (Cannady et al. 2021). The detector performance is characterized by Monte Carlo simulations compared in detail to a series of accelerator calibrations (Asaoka et al. 2017) and regularly monitored in flight with cosmic-ray data.
Photon energies are determined by summing the deposited energies in the TASC. At 10 GeV, the energy resolution is 3%. The CAL field of view for gamma rays is ∼45°from the zenith.
The analysis algorithms (CC Track mode for the LEG trigger and EM Track mode for the HE trigger) are described in Cannady et al. (2018). The effective area is shown for both CC and EM Track mode in Figure 6(a) as a function of gamma-ray energy for four separate zenith angle ranges. The effective area reaches ∼400 cm 2 for energies up to ∼50 GeV, where the identification of zero-charge particles in CHD and IMC begins to be affected by the backscatter of higher energy particles.
The angular resolution is shown in Figure 6(b) separately for CC and EM modes as a function of the gamma-ray energy and N p , the number of IMC layers used in the tracking. The angular resolution in Figure 6(b) is defined as the value C 68 such that 68% of simulated events have reconstructed directions deviating from the true direction by an angle α < C 68 . C 68 is 0°. 5 or better for all energies above 1 GeV for all but the shortest tracks.
A sky map for 5 yr of CAL observations above 1 GeV including steady gamma-ray sources and bright transient events (e.g., CTA102) is shown in Figure 7. Sources are marked by Figure 5. Schematic of CAL calorimeter showing CHD, IMC, and TASC subsystems and characteristically identifiable behavior of electrons, protons, nuclei, and gamma rays passing through the detector. Gamma rays are identified by the absence of charge in the CHD and IMC and by the shape and profile of the electromagnetic shower in the TASC. Signal amplitudes are shown according to the right-hand color scale in terms of minimum ionizing particle (MIP) energy deposits. green or blue circles depending on extragalactic or Galactic origin, respectively. Although the CAL energy range is far above the typical energy range of GRB prompt emissions, CAL provides the possibility of detecting high-energy gamma-ray emission from GRBs similar to events observed up to nearly 100 GeV by Fermi-LAT (Ajello et al. 2019b).

CALET Search for GW Events during the LIGO/Virgo Third Observing Run
CALET was still in its commissioning phase at the time of the initial LIGO detection of GW150904. Since 2015 October, however, CALET has actively searched for electromagnetic counterparts of gravitational-wave events. The CGBM and CAL searches for Advanced LIGO and Advanced Virgo events during O1 and O2 have been described by Yamaoka et al. (2017) and Adriani et al. (2018) respectively. Here we describe the CALET search since the start of O3 in 2019. The LIGO/Virgo collaboration (LVC) reported 56 gravitational events (not including retracted events) in O3. Also, LVC and Fermi-GBM reported one subthreshold event (Abbott et al. 2019). The 57 events are shown in Table 2. All information in Table 2 is based on GCN circulars and The Gravitational-wave Candidate Event Database (GraceDB) 43 operated by LIGO. Here "possible source" shows the most probable source as listed in GraceDB. CALET has searched for electromagnetic counterparts of gravitational waves in both the CGBM and CAL data and results have been reported in GCN circulars ( Table 2). We describe the details of the CGBM and CAL analyses separately in Sections 3.1 and 3.2.

CGBM Analysis
As of the end of 2021 November, CGBM has detected 271 GRBs, with 12% of the CGBM GRBs classified as short GRBs (Kawakubo et al. 2021). CGBM observations of the O3 events are summarized in Table 3. Here T 0 is the time of the gravitational-wave event reported in GraceDB and listed in Table 2. Since the duty cycle of CGBM is ∼60% due mainly to passages through the SAA, CGBM high voltages were sometimes turned off at gravitational-wave trigger times. The high voltage status at the trigger time of each gravitationalwave event was recorded in the housekeeping data and the status of the high voltages is shown in Table 3 under the CGBM trigger. Here "disabled" means the onboard trigger system was disabled at T 0 either because the CGBM high voltages were off or the CGBM event buffer was full. "No trigger" means the onboard trigger system was not triggered during the time interval T 0 ± 60 s even if the onboard trigger system was enabled. The monitor data were inspected on the ground for each O3 event to confirm that no potential candidate event occurred within T 0 ± 60 s. For each event, the summed LIGO/Virgo localization probability above the CGBM horizon was calculated using the LIGO/Virgo sky maps from GraceDB. If the summed LIGO/Virgo localization probability above the horizon was greater than 1%, a targeted signal search was then performed using the TH data. Light curves were constructed from the TH data for each CGBM detector with 0.125 s time binning in twenty energy bands for T 0 − 60 s to T 0 + 60 s. S/N was recalculated for each light curve with an expanded set of search parameters (Table 4) to search for a significant signal within T 0 ± 60 s. In the ground analysis, Δt BG was taken from both sides of Δt with one exception. In the case of the high voltage turning on or off within ±60 s of the trigger time, Δt BG was taken only from the period when the high voltage was on. S/N was therefore calculated every 0.125 s for 1440 separate conditions summarized in Table 4. Table 3 shows the highest S/N from the set of S/Ns calculated for all conditions such that the foreground intervals Δt include T 0 . "Conditions for s T max, 0 " shows the conditions resulting in s T max, 0 . s max,60s shows the highest S/N calculated for all conditions in T 0 ± 60 s. "Conditions for s max,60s " shows the conditions corresponding to s max,60s . Finally, T max is the start time of the foreground interval relative to T 0 when the S/N was equal to s max,60s .
Background varies over an orbit mainly due to the variation of the trapped charged particle flux and activity of bright X-ray sources in the field of view. Since the estimation of background counts in the S/N calculation is based on the summation of the observed counts before and after the foreground time interval, estimated background counts during Δt are sometimes CAL angular resolution C 68 for the two trigger conditions as a function of gamma-ray energy and the number of IMC layers used for tracking. The smooth curve shows the fitted resolution function described by Cannady et al. (2018). underestimated or overestimated depending on the background variation during the orbit, affecting the calculated S/Ns. The distribution of S/Ns is not described by a normal distribution. As an example, Figure 8 shows the distribution of S/Ns calculated using SGM high-gain data for one day. The histogram includes S/Ns calculated every 0.125 s continuously with all conditions and gain settings. Individual bins are therefore not independent of each other. As seen in the figure, the distribution extends up to S/N ∼ 7. On this particular day, a small tail is present at high S/N due to an increased counting rate in two successive orbits at high latitude. Including this contribution of high S/N events due to low-energy charged particles, the fraction of events with S/N > 7 is 1.6 × 10 −7 . We set the threshold for the off-line trigger at S/N = 7 and require that candidate events show up in multiple detectors (HXM1, HXM2, and SGM) and multiple energy channels, that the event arrival time falls within T 0 ± 60 s, and that CAL CHD and IMC see no simultaneous increase in the low-energy charged particle rate.
As a check on the reasonableness of these event selection criteria, we have performed a search for events simulating gravitational-wave counterparts by searching for events either one orbit prior to or one orbit after the times of the 57 events listed in Table 2-i.e., at times and positions where the instrument is at approximately the same latitude and in approximately the same pointing direction as at the time of the actual gravitational-wave event. One event was seen with S/N = 7.13, but only in HXM2-i.e., no multichannel events were found that successfully masqueraded as candidate CGBM counterpart events.
In the time windows T 0 ± 60 s corresponding to real LIGO/ Virgo events, one LIGO/Virgo event (S200112r) was found with a nearby CGBM S/N > 7 (Figure 9). In this case, the time of highest S/N (T max ) was T 0 − 1.41 s. However, the signal can be seen in only the lowest energy channel of HXM2; HXM1 and SGM detected no significant signal. We conclude that this signal is likely a random fluctuation and not likely to be physically associated with S200112r.
Since no acceptable GRB candidate associated with any of the LIGO/Virgo GW events was found in the CGBM data, we estimated upper limits for the X-ray/gamma-ray flux. For each GW event, we simulated typical GRB spectra in the TH data using the "fakeit" command of XSPEC version 12.10.1 44 (Bambi & Nampalliwar 2020) assuming typical spectra and durations for short GRBs. We used the CGBM response matrix database which includes CGBM response files calculated for 5°p itch in zenith angle and azimuth. For the assumed input spectrum, we used a Band function and a power law with exponential cutoff and standard values for photon indices and E peak characteristic of short GRBs: For the Band function, we used α = −0.46, β = −2.98, and E peak = 413 keV (Poolakkil et al. 2021) and for the cutoff power law we used α = −0.42 and E peak = 566 keV Racusin et al. 2017). We assume a burst duration of 1 s. Based on the LIGO/ Virgo sky maps, we took the direction of the source to be that direction within the CGBM field of view for which the localization probability was maximum and applied the relevant CGBM response matrices. Tables 5-7 show the resulting timeaveraged flux upper limits at the level of 7σ calculated separately for HXM1, HXM2, and SGM in the 10 keV-1 MeV energy range. Here P corresponds to the summed probability in the field of view of each detector (HXM1, HXM2, or SGM), α and δ are the highest probability directions in Equatorial coordinates of the GW sources in the field of view of each detector, and θ and f are the zenith and azimuth angles of incident photons striking the detector from the direction of the GW source. (Since the CGBM angular response is calculated on a 5°grid, θ and f are tabulated with 5°precision).

CAL Analysis
Results of the CAL observations of the 57 events reported by LIGO/Virgo for O3 are shown in Table 8. Of the 57 events, 20 were in the CAL field of view. Of those, 13 occurred while CAL was in HE mode and seven occurred when CAL was in LEG mode. In no case did CAL detect any events from the allowed region within ±60 s of the LIGO/Virgo T 0 . The pointing direction of the center of the CAL field of view is given in the columns labeled α and δ. "Coverage" is the Figure 7. Map of the sky above 1 GeV observed by CAL in Galactic coordinates (Cannady et al. 2021). The point-spread function is determined for each photon event and an order 9 Healpix map (∼7′ resolution) is filled with the summed smeared signals from each event. Total exposure is accumulated in bins of energy for each pixel of an order 6 Healpix map (∼55′ resolution). The resulting counts map is shown in logarithmic scaling.
fraction of the overlapping region of the LIGO/Virgo localization map covered by the CAL field of view during the interval T 0 ± 60 s.
Based on the LIGO-Virgo sky maps and taking ISS structures in the field of view into account for each event, we calculate an effective area for each direction as a function of  Disabled L L L L L L S190421ar No trigger 0% L L L L L S190425z Disabled No energy. CAL observations of the GeV sky reflect a combination of individual galactic and extragalactic sources together with both galactic and extragalactic diffuse emission in good agreement with Fermi-LAT Pass 8 observations (Cannady et al. 2018(Cannady et al. , 2021. Based on this, an expected number of background events is calculated for each time interval and direction in the CAL search for counterparts; typically, the expected N BG ∼ 0.1 or less. In the case of a null event, we assume a power-law spectrum for a potential gamma-ray burst with a single power-law photon index of −2 (Poolakkil et al. 2021), taking into account the CAL sensitivity as a function of energy, and estimate an upper limit (90% confidence level) on the gamma-ray flux based on 2.44 events above the expected background. Assuming a burst duration Δt = 1 s, upper limits are calculated for the time interval T 0 ± 60 s in units of erg cm −2 s −1 for the energy range 10-100 GeV (with the CAL in HE mode) or 1-10 GeV (with the CAL in LEG mode). The maximum time-averaged flux for an individual pixel in the LIGO/Virgo localization area is listed as the upper limit given in the final column of Table 8. The lowest upper limit in Table 8 is for the CAL observation of the sky at the time of S190408an. No CGBM onboard trigger occurred near the GW event time; no excess count rate was seen in the ground analysis of the HXM and SGM data within 60 s of T 0 , and no CAL gamma-ray events were detected within 60 s of the GW trigger time. Figure 10 shows the map of Disabled No trigger 60% 3.31 HXM2, Low, ch1 5.59 HXM2, Low, ch1 −1.29 Δt = 4 s, Δt BG = 8 s Δt = 0.5 s, Δt BG = 8 s S200129m Disabled No  90% confidence level upper limits measured by CAL during the time interval T 0 − 60 s to T 0 + 60 s. The pointing direction of the CAL during the observation is marked by the cyan contour extending upward from the lower right to the middle of the diagram. The magenta cross marks the zenith direction at T 0 . The green area near the extreme upper edge of the CAL 90% upper limit region marks the localization contours reported by LIGO/Virgo. The red and blue circles are the HXM and SGM fields of view ignoring the effects of the ISS structures at T 0 , respectively.

Discussion and Conclusions
The neutron star merger event GRB 170817A was detected by Fermi-GBM and INTEGRAL ∼1.7 s after the gravitational-wave event at a 10-1000 keV flux level (based on the Fermi-GBM  measurement) of (5.5 ± 1.2) × 10 −7 erg cm −2 s −1 . The observed properties of the gamma-ray event were those of an ordinary although subluminous short GRB.
CALET is currently approved to continue operating through 2024. We gratefully acknowledge JAXAʼs contributions to the development of CALET and to the operations aboard the JEM-EF on the International Space Station. We also wish to express our sincere gratitude to Agenzia Spaziale Italiana (ASI) and NASA for their support of the CALET project. In Japan, this work was supported in part by JSPS KAKENHI grant Nos. 26220708, 19H05608, 17H02901, 21K03592, and 20K22352 and by the MEXT-Supported Program for the Strategic Research Foundation at Private Universities (2011-2015) (No.S1101021) at Waseda University. The CALET effort in Italy is supported by ASI under agreement 2013-018-R.0 and its amendments. The CALET effort in the United States is supported by NASA through Grants 80NSSC20K0397, 80NSSC20K0399, and NNH18ZDA001N-APRA18-000. A part of this research is made possible by use of data obtained from DARTS at ISAS/JAXA.