Alfvén wave generation and electron energization in the KiNET-X sounding rocket mission Open Access

A grand challenge problem in magnetospheric physics is the coupling of large-scale wave energy to small-scales where plasma can be heated and auroral electrons accelerated (Holmgren , 1980). Turbulent cascades from large-scale energy and momentum injection scales to dissipative (kinetic) scales are ubiquitous in space plasmas. Understanding the flow of energy and momentum in, e.g., magnetosphere–ionosphere coupling, is therefore critically dependent on the physics of kinetic-scale dissipation. The KINetic-scale Energy and momentum Transport eXperiment (KiNET-X) was an active space plasma experiment launched from Wallops Island, VA on May 16, 2021 where barium clouds of known (i.e., upper and lower limits) energy and momentum were injected at the ion kinetic scale and coupling to the ambient plasma was measured. The scientific objectives of KiNET-X were to (1) quantify the coupling of a barium cloud to the ambient plasma medium with multipoint, in situ plasma measurements and by observing the evolution of the cloud with ground-based optics; (2) identify physical processes that support parallel electric fields (i.e., electron inertial effects, classical collisions, and/or turbulent wave particle interaction); (3) investigate the interplay between energy and momentum transport and the formation of parallel electric fields; and (4) observe and quantify the possible electron acceleration.

KiNET-X made plasma measurements adjacent to and within an Alfvén wing (i.e., realm of perturbations to the ambient plasma) of an injected thermite barium release. Two barium thermite canisters were detonated $∼$ 30 s apart at altitudes of about 400 and 350 km. A plasma diagnostic array included four free-flying “Bob” payloads containing thermal ions (two leading) and electron instruments (two trailing) were positioned above the release, but within the magnetic flux tube subtended by the release cloud. Figure 1 illustrates the first fraction of a second after a release, showing the evolution of the Alfvén wing with respect to the barium cloud and the instrumented payload. The inter-instrument separation was comparable to the perpendicular gradient scale lengths of the cloud, which were comparable to the oxygen ion inertial length (⁠ $∼$ km). The main payload provided DC and quasi-DC electric and magnetic fields, AC electric fields (waves), pitch-angle resolved energetic electrons (i.e., $∼$ 100 eV), thermal electrons, thermal ions, and plasma number density. Ground-based (Bermuda) and aircraft-based (north of Bermuda and the releases) optical observations recorded the motion of the releases. The combined dataset provided basic information on the magnitude and scale of the Alfvénic perturbations and quantified the coupling. The  Appendix summarizes the KiNET-X instrument suite.

KiNET-X was an experiment similar in design to the critical ionization velocity (CIV) experiments, CRIT I and II, (Torbert , 1992; Brenning , 1991a, 1991b) and Active Plasma Experiment (APEX) North Star (Erlandson , 2002; Lynch , 2001; Pfaff , 2004; Delamere , 2004) but with a fundamentally different design and scientific objectives. The CRIT experiments utilized shaped charge barium releases (i.e., a directed high-velocity beam) below the solar terminator to isolate non-photo ion production and thus generated minimal momentum loading. APEX utilized an aluminum shaped charge injection aimed at instrumented payloads which could not measure Alfvén waves. Our science was motivated largely from optical observations of several thermite barium releases (i.e., spherically expanding clouds about the trajectory motion) made as part of the NASA Combined Radiation and Release Experiments Satellite (CRRES) in 1990 where relevant in situ measurements were not available. Several of the CRRES experiments were designed to study cloud/ionosphere coupling via momentum loading with fully sunlit barium clouds. The CRRES releases demonstrated that the ion clouds propagated much farther across the geomagnetic field than would be expected if the cloud was ideally coupled (i.e., if momentum loss occurred through freely radiating Alfvén waves) with the background ionosphere. The effect was termed “skidding” (Huba , 1992; Delamere , 2000). This skidding illustrated that the dense ion clouds were partially decoupled from the ambient plasma.

In this article, we summarize the results of the KiNET-X sounding rocket mission, highlighting the generation of Alfvén waves and energized electrons. The electron heating and acceleration has numerous implications for previous active plasma experiments. We discuss implications for previous critical ionization velocity experiments and coupling experiments. Additionally, we suggest that the electron energization was directly linked to Alfvén waves.

The CIV effect occurs when a neutral gas, with a kinetic energy exceeding the ionization potential, is ionized while passing through a magnetized plasma (Alfvén, 1954; Newell, 1985). While KiNET-X was not designed to study the CIV effect, electron energization is a fundamental process that operates in all injection experiments. In the CRIT I and II sounding rocket experiments (Torbert , 1992), shaped charges were used to create high-velocity jets (>10 km/s), and free-flying instrumented payload sections were placed in and near the jet to make plasma measurements. From a comparative perspective, KiNET-X can potentially elucidate the fundamental physics common to these legacy active plasma experiments.

Electron energization is a critical mechanism in the CIV effect (Newell, 1985; Piel, 1990; Torbert, 1990; Brenning, 1992), providing the requisite electron impact ionization. While the relatively low ionization yields of the CIV experiments are not comparable to a fully sunlit experiment in terms of robust momentum loading, the non-photoionization mechanisms can play a very important role in the evolution of the electrodynamic interaction. For example, seed ionization is critical to enhance the momentum loading from resonant charge exchange (i.e., Ba + Ba⁺). Furthermore, electron energization can be a sink in momentum and energy transport by the Alfvén waves. Early CIV theories suggested that the heavy ions would run ahead of the magnetized electrons, setting up a space charge (Sherman, 1973). The electrons would subsequently $E×B$ drift, leading to a modified two-stream instability that would excite lower hybrid waves. Wave excitation leads to electron energization, and the heated electrons can provide further ionization, completing the feedback loop. Rather than invoking lower hybrid waves to heat the electrons, we will show in this article that Alfvénic energization is also a likely candidate for heating electrons.

We note that the magnitudes of the space charge and related electric fields are not arbitrarily large. On scales much larger than the Debye length, space plasmas are to an excellent approximation quasi-neutral, and the perpendicular electric fields, in this case, should simply be the convection electric field. However, significant ambipolar electric fields can also exist with the parallel potential roughly equal to the driving perpendicular potential which, in a steady state, implies skidding (Delamere , 2002). In the CRIT I experiment, Brenning (1991a) showed that indeed $E⊥∼−vBa×B$ as one would expect, where $vBa$ is the cloud injection velocity. This is consistent with the expectation for the fluid-scale, magnetohydrodynamic source of Poynting flux, or Alfvén waves. However, large gyroradius effects are important during the early phase of the injection and asymmetries arise due to Hall currents, but the requisite $J·E<0$ condition remains for the MHD driver (Delamere , 1999).

In 11 CIV experiments, the ionization yields have ranged between 0% and 30% (Wescott , 1975; Haerendel, 1982; Deehr , 1982; Wescott , 1986a; 1986b; Newell, 1985; Torbert , 1992; Stenbaek-Nielsen , 1990; Wescott , 1990; 1994). These disparate results, to date, remain largely unexplained. If electron heating is the key mechanism via the Alfvénic driver, then initial seed ionization is critical for Alfvén wave generation. It is possible that each experiment exhibited different initial conditions giving rise to the disparate results.

The KiNET-X experimental concept is based on: (a) results and insights gained through the extensive analysis of image data of thermite barium releases acquired during the NASA Combined Release and Radiation Effects Satellite (CRRES) mission in 1990 and 1991 (Bernhardt, 1992), and (b) numerical simulation using a three-dimensional hybrid code designed for the specific task of modeling thermite barium releases (Delamere , 2000). The heritage of the KiNET-X instrumentation suite was derived from the APEX North Star experiment that investigated coupling between the ionosphere and a high-velocity plasma (Erlandson , 2002; Lynch , 2001; Pfaff , 2004; Delamere , 2004).

As a part of CRRES, a number of barium thermite releases were made at orbital speed (10 km/s) and at altitudes between 400 and 500 km. The releases were made at twilight, with the releases in sunlight, and the ground in darkness to allow optical observations. One of the first observational results was the unexpected offset between the release point and the final location of the edge of the ion cloud. Huba (1992) demonstrated that an assumed $E×B$ drift motion was due to the polarization of the dense ion cloud. We note that the barium cloud was considered to be optically thin and collisionless for $t>0.2$ s, simplifying the problem to an ion cloud coupled to the ambient ionospheric plasma via Alfvén waves. However, the cloud may not be optically thin or in photo-equilibrium for the first few seconds (Drapatz, 1972), further complicating the interpretation of skidding. Currents driven by the Alfvénic disturbance accelerate the ambient plasma and decelerate the ion cloud; momentum is therefore transferred from the ion cloud to the ambient medium until the ion cloud is stopped (Delamere , 2002).

A subsequent detailed analysis of the optical data (assuming an optically thin cloud to photoionization) showed that the observed $E×B$ drift distance of the ion cloud was significantly larger than would be predicted from ideal Alfvénic coupling. Modeling efforts by Huba (1992) and Delamere (2000) predicted that the efficient momentum coupling with the ambient plasma should stop the cloud within 10–20 km of the release point. However, the observations indicate that a substantial portion of the barium plasma cloud skidded for roughly 50 and 100 km for the small and large releases, respectively (Delamere , 1996). Figure 2 shows an ion cloud profile taken from unfiltered optical observations of the small G11A barium release. The x-axis is distance along the release trajectory from the release point. The calibrated ion profile (solid line) is the integrated column intensity along the magnetic field in a bin of one pixel in width. The shaded region represents contamination to the signal from neutral emissions, and the dashed line illustrates the possible continuation of the ion profile into the region containing neutral emissions (expected to be an exponential decay as the neutral cloud is ionized, or $dNi/dt=(N0/τ)e−t/τ$⁠, where $Ni$ is the number of ions, $N0$ is the number of released neutrals, and $τ$ is the photoionization timescale). The skidding distance is defined by the peak in the ion cloud profile at roughly 50 km from the release point. Numerical estimates for freely propagating Alfvén waves predict less than 10 km for the $E×B$ motion. A similar analysis of the large G9 CRRES release give observed and estimated drift distances of >100 and 17 km, respectively. Some process must, therefore, have attenuated the Alfvén waves, or, in other words, the ion cloud must not be fully coupled to the ambient ionospheric plasma. Delamere (2002) demonstrated that parallel electric fields are responsible for this decoupling.

The APEX aluminum plasma cloud was a dense “slug” (i.e., few Al⁺ gyroradii in size), injected perpendicular to the magnetic field at $∼$ 25 km/s and toward a plasma diagnostic payload (PDP) and an optical sensor payload (OSP) with a magnetometer separated by roughly $∼$ 1 km. The altitude of the release was at 363 km. The cloud (i.e., center of the diamagnetic cavity) was observed to slow down between PDP and OSP, indicating a momentum transfer from the ambient plasma. The slowing was consistent with hybrid simulations of the release, suggesting that even at this small kinetic scale, momentum was transferred from the ambient plasma (Delamere , 2004). Limitations in the optical observations precluded a more detailed analysis of the perpendicular cloud motion, or skidding.

We now summarize the results of the KiNET-X experiment, highlighting signatures of Alfvén wave propagation, parallel electric fields, and electron energization. Both barium canisters successfully detonated at 400 and 350 km, respectively, on the down leg of the trajectory. Figure 3 shows the two ion clouds as viewed from the airborne optics platform, showing the two ion clouds (violet) and the green neutral cloud of the second release. The motion is from upper right to lower left, and the sharp edge of the ion clouds is in close proximity to the release location. More details regarding the ion yields and temporal evolution of the clouds are given by Barnes (2024).

The results of the KiNET-X experiment for the main payload instruments are summarized in Figs. 4 and 5 for the first and second releases, respectively. From top to bottom, we show electron density (Langmuir probe), barium ion density (Petite Ion Probes), DC magnetic and electric fields, AC waves, thermal electron temperature (ERPA), and field-aligned electrons (EPLAS). A coordinate system about the local magnetic field vector such that the zonal direction is defined by $B×r$⁠, where $B$ is the local magnetic field vector, and $r$ is the vector from the center of the Earth to the spacecraft. The zonal component is positive in the eastward direction. The meridional component is the zonal direction $×B$⁠. The field-aligned or parallel component is along $B$⁠.

The overall similarity of the observation shows that both injections were qualitatively similar. However, there are a few key differences. In the second injection, we see increased fluctuations in the density and in the DC magnetic and electric fields. The maximum magnetic perturbation for each injection was similar, $∼$ 60 nT in the meridional direction (i.e., north–south). The ratio of electric to magnetic field magnitude was roughly equal to the ambient Alfvén speed, confirming the generation of an Alfvénic disturbance during the first $∼$ few tenths of a second. The parallel electric fields along the spin axis are measured with two parallel double probes of equal length on either side of the payload, as described in the  Appendix. Since the spin axis was oriented along the magnetic field direction to within 1°, these measurements correspond to electric fields “parallel to the magnetic field,” denoted by “parallel 1” and “parallel 2” in the fourth panel of Figs. 4 and 5. The parallel components show brief excursions (0.1–0.2 s) corresponding to values as high as 10 mV/m at the onset of the releases, before relaxing to values ranging from 0 to 5 mV/m within the subsequent 0.5–1 s flight time. The values in the two measurements are not identical, possibly due to asymmetric contributions from photoelectrons associated with the non-uniform optical properties of the releases as they swept past the payload, displaying different signatures in the two double probes situated 180° apart.

The electron beam was observed about 1 s after the release and could be associated with field-aligned currents at the leading edge of the electrodynamically active portion of the cloud. We do not have an explanation for why the beam is not strictly field aligned nor do we believe that this is an instrumental limitation (see discussion in the  Appendix). Given the short duration of the beam together with the absence of a beam in the first release, we conclude that the beams are localized and perhaps intermittent. Regardless, strong parallel potentials will have significant implications for momentum transport, and we consider this observation to be compelling evidence for skidding.

Active plasma experiments can be used to strongly perturb the space plasma environment. During the early phase of an injection experiment (e.g., few seconds), the plasma cloud can excite a variety of waves rather than acting as tracer particles. Early experiments relied primarily on ground-based optics to diagnose the plasma interaction, but advances in optical sensors have dramatically improved imaging capability of both the ion and neutral components of the injected cloud. Advances in plasma (fields and particles) instruments have enabled a new generation of possible experiments from the sounding rocket platform.

One of the lingering puzzles in active plasma experiments is understanding ion yields. The neutral clouds are considered to be optically thin to ionizing solar radiation, but this may not be the case within the first $∼$ second (Drapatz, 1972). Certainly, the photo-equilibrium state of the barium neutral atoms is not instantaneous, resulting in, potentially, reduced ionization rates during the first few seconds (Barnes, 2024). It is during this early time that the mass density of new-born ions can exceed the ambient density, resulting in strong perturbations. In the case of KiNET-X, additional ionization from electron impact is not expected as 0.2 eV electrons are much too cold (see Fig. 7 and discussion below). However, for comparison with other experiments, electron impact ionization could provide significant additional ionization, leading to a positive electrodynamic feedback. The CRRES CIV experiments (G-13 Ba and G-14 Ba) produced respective ion yields of 0.15% and 1.48% (Hampton, 1996), yet the release conditions were very similar. A possible explanation in the context of a nonlinear and positive feedback is the seed ionization. Sources of seed ionization include associative ionization, charge stripping, and charge exchange. Any initial ionization that generates a significant Alfvénic perturbation can subsequently heat the thermal electrons leading to additional electron impact ionization. Future studies should carefully consider the sensitivity of the electrodynamic interaction to variations in the initial ionization processes.

The results of the KiNET-X mission have certainly motivated a careful reanalysis of our understanding of electron energization in previous experiments. Torbert (1992) showed significant electron heating in the CRIT II experiment. The shaped change injection produced speeds of neutral barium vapor > 10 km/s. CRIT II electron measurements showed elevated fluxes in the 1–2 eV range and 10 eV range with ambient ionospheric electron temperatures expected to be a fraction of an eV. The ionization potential of Ba is roughly 5 eV. Clearly, the 10 eV electrons are capable of producing ionization. Figure 7 shows an estimate of the electron impact ionization rate coefficient, k, as a function of electron temperature, computed using the electron impact ionization cross sections, $σ$⁠, from Dettmann and Karstensen (1982) integrated with a Maxwellian distribution, such that $k=⟨σv⟩$⁠. Here, we see that a 2 eV thermal electron distribution has a rate coefficient of roughly $2×10−8$ cm³/s due to contributions from the tail of the distribution $>5$ eV. The ambient electron density for CRIT II was $∼5×105$ cm⁻³, and the neutral barium yield was 10²⁵ atoms. Thus, the ion production rate is $kNBane∼1023$ s⁻¹, which would be consistent for the ion yield if the energized electrons persisted for 1–2 s. We suggest, based on the concurrent observation of Alfvén wave propagation and thermal electron energization, that future studies and experiments focus on the Alfvénic electron energization mechanism.

Electron energization is a key aspect of inertial Alfvén waves, having both resonant (Fermi) and non-resonant components (Kletzing, 1994). The latter current-carrying component results in a residual heating of the thermal electrons that persists even after the passage of the wave (e.g., Watt 2005, 2006; Damiano 2019; Coffin 2022). This thermalization can be major sink of the wave energy (e.g., Coffin (2022), and although the perpendicular variations cannot be determined from these single point measurements, the IAW could be a potential source of the thermalized electron populations evident in the observations. The strong parallel electric field inferred from the electron beam in the second release is suggestive of “skidding,” which would require that the release cloud was decoupled from the ambient plasma. Here, we note that the beam appears after the primary Alfvénic activity at 628.3 s and could be related to a longer-term decoupling of the cloud. The fact that such a beam was not observed in the first release suggests either no skidding occurred or that the beam was spatially localized and observation is based on a chance encounter.

We conclude as follows:

• The KiNET-X experiment generated an Alfvénic pulse.

• A field-aligned electron beam was observed in the second release, consistent with a field-aligned potential equivalent to the perpendicular cross-cloud potential. Strong parallel electric fields (⁠ $E∥/E⊥∼1$⁠) suggest that the ion cloud was skidding.

• Coincident with the Alfvénic activity was a two- to threefold increase in the thermal electron temperature. The electron heating is roughly equal to the energy density of the Alfvén pulse suggesting strong wave/particle interactions.

• Future experiments to address prompt ionization and Alfvénic energization might benefit from CRIT I/II-like shaped charge injections or orbital injections (i.e., $∼$ 10 km/s).

This article was supported by NASA (Grant Nos. 80NSSC18K0797 and 80NSSC21K2009). The KiNET-X science team is grateful for the support of Matthew Blandin and Kylee Branning during the launch.

The authors have no conflicts to disclose.

P. A. Delamere: Investigation (equal); Project administration (equal); Writing – original draft (equal); Writing – review & editing (equal). A. Otto: Conceptualization (equal). M. Moses: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). C. Moser-Gauthier: Data curation (equal); Formal analysis (equal). K. A. Lynch: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). M. Lessard: Data curation (equal); Formal analysis (equal); Investigation (equal). R. Pfaff: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). M. Larsen: Investigation (equal); Resources (equal). D. L. Hampton: Data curation (equal); Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). M. Conde: Conceptualization (equal). N. P. Barnes: Data curation (equal); Formal analysis (equal); Writing – review & editing (equal). P. A. Damiano: Formal analysis (equal); Writing – review & editing (equal).

The KiNET-X data shown in Figs. 4–6 are accessible at Delamere (2024) (https://figshare.com/s/bb015eaeef8d00e90113).

The main payload carried a suite of eight Petite Ion Probes (PIPs), along with two small telemetered “Bob” payloads each carrying a PIP. A full description of the Isinglass/Hesh Bob payloads can be found in Roberts (2017a; 2017b). The PIPs provided measurements of the thermal plasma, as estimates of the ion bulk moments, given constraining plasma measurements from other onboard or nearby sensors. The PIP sensor is a small collimated retarding potential analyzer (RPA) used in a subsonic regime inside the payload sheath (Frederick-Frost , 2007; Fisher , 2016; Roberts , 2017b). Thermal ions are used to look for variations in plasma quantities because, in the low altitude dark ionosphere (and indeed even in sunlight at these altitudes), payloads float to negative payload potentials of one to several volts (Siddiqui , 2011). The complex response of the PIPs to the accelerated (through the sheath) and rammed (from payload velocity and plasma drift, both subsonic) thermal ion population means that the current-to-voltage (I–V) curves can be used for extracting plasma parameters (Fisher , 2016; Roberts , 2017b).

Barium density data were obtained by forward-modeling PIP current-vs-screen bias voltage (IV) curves for a two-species Maxwellian plasma to measured main payload PIPs' IV curves at times when a PIP was within 30° of RAM. The main payload PIPs observed the barium from the first release starting just under 2 s after canister detonation, with a maximum density of approximately $1×1010$ m⁻³ observed around $∼$ 597 s. In contrast, the main payload PIPs first observed barium from the second release about 1 s after detonation, with the maximum density of $∼6.5×1010$ m⁻³ observed around $∼$ 630.25 s.

Electron retarding potential analyzers (ERPA) on the main payload as well as the trailing bobs determined the background electron temperature. The ERPAs are quite similar to a Faraday Cup with a swept retarding potential placed at its entrance for energy selection. Electrons with energies less than the retarding potential are rejected and those with higher energies pass through the screen and are collected on an anode below. The current collected by the anode is measured by a low noise electrometer circuit. The electron distribution function is obtained by sweeping the potential of the entrance screen and recording the differential flux at each step. This is accomplished with a ramping up and then down of that potential through a range that is wide enough such that it accommodates possible payload charging. The ERPA covers an energy range of about 0–3 eV with steps of 0.06 eV.

Possible sources of uncertainty in the determination of electron temperature could be due to a number of factors, including effects of misalignment to the background magnetic field, modulation of that field misalignment due to coning of the rocket, flux dependence, and digitization resolution affected by the resolution of the steps during the sweep. We also note that the temperature calculation is unaffected by the payload potential, as the potential does not affect the width or shape of the differential spectrum. We assume errors associated with misalignment to the background magnetic field to be insignificant compared to the others mentioned. Flux dependency is also of minimal significance since it is most prevalent at high energies where the counts drop off into the noise; we do not include this region in our fit, and therefore it does not affect our temperature calculation. The largest source of uncertainty listed above is digitization resolution (discretization of the distribution function). While this is without a doubt the most significant source of uncertainty, the nature of our temperature calculation diminishes its impact on the final temperature result. Ultimately, the error bars calculated as described here lead to a mean error of less than 5% in the temperature determination.

The Electron PLASma Instrument (EPLAS) measures electron precipitation and is a version of a basic electrostatic analyzer top-hat detector with 2D planar field of view (Carlson , 1982; Young , 1988). The detector covered 10 eV to 14 keV in 32 one msec energy steps, with 30 pitch angle bins and 32 energy sweeps/s. At typical payload translational speed of $∼$ 1 km/s, this provided a 2D distribution function every 32 m, comparable to expected scale sizes of field-aligned currents. The main payload was aligned to the background magnetic field and the EPLAS was deployed so that its acceptance plane was nominally aligned to the field in order to obtain pitch angle information of the distributions. If the payload had a significant coning motion, it would be possible that some precipitation might be blocked. In the extreme case, where the payload would be in a cartwheel motion, the EPLAS aperture would then be limited to acquiring distribution functions twice per spin. We include this information for completion—no significant coning was observed, and it seems clear that the instrument had access to the entire population.

The DC and AC vector electric fields were measured using the standard double probe technique (Pfaff, 1996). In this manner, spherical sensors with embedded pre-amps were extended on three independent double probes of length 6.17 m (tip-to-tip) in the spin plane. Two double probes provided orthogonal spin plane measurements, while a third double probe was oriented parallel to one of these components, yet separated along the spin axis by 3 m. This configuration provided two redundant measurements of the third (axial) electric field component on either side of the payload. As the spin axis was aligned to 1°–2° of the magnetic field direction, these measurements provided the parallel electric field components. As a result, the experiment contained a full, three dimensional electric field measurement that completely parameterized the vector electric field (DC to 32 kHz) including electric fields parallel to the magnetic field direction.

In order to help identify plasma modes and their dispersion relations, the electric field AC instrument was configured in a wave interferometer mode. In this manner, spaced receivers composed of sphere to payload potential differences were gathered to provide wavelength and phase velocity information of the electrostatic modes (Pfaff and Marionni, 1998). Beyond the VLF regime, the electric field experiment also included high frequency channels to record wave electric fields to completely characterize the plasma wave environment associated with the releases, which included a burst memory which gathered vector AC fields (three components) sampled at 8 Msample/s.

A vector flux gate magnetometer was used to measure the magnetic field. This magnetometer provided triaxial measurements sampled with an 18 bit A/D (simultaneous with the DC electric field data). The magnetometer resolution was better than $±$ 1 nT, more than sufficient to detect perturbations associated with the barium injection.

A fixed-biased Langmuir probe was flown in order to observe the relative plasma number density and its fluctuations. In addition to pre-launch calibration curves and theory, the Langmuir probe was normalized using simultaneous ground-based ionosonde data, as well as plasma wave data where applicable.

Ground-based optics at Bermuda and optics onboard a NASA aircraft were used to capture the distribution of neutral particles as they exit the thermite canister, inventory the total number of ions produced (the yield) and examine the subsequent structuring and position of the ion cloud. The two sites not only provided the data needed to triangulate the release position and subsequent ion cloud distribution but also the combination of the rocket trajectory and aircraft (A/C), positioned north of Bermuda, allowed us to examine the Ba ion cloud morphology both parallel and perpendicular to the magnetic field.

A pair of video cameras, two filtered time-lapse cameras, and several DSLR cameras were used. The video cameras (Sony $α$ 7S) were operated at 30 frames per second, and set at two different sensitivities to provide the dynamic range needed to cover the extreme intensity range from release to subsequent Ba⁺ cloud. The video cameras were fitted with 50 mm lenses which provided $∼$ 100 m spatial sampling with a 40° × 30° field of view. The ion inventory cameras were fitted with filters to observe the 455.4 nm Ba⁺ resonance emission, used in several Ba release experiments to determine the total volume emission rate, and therefore total number of emitters. Both sites also had DSLR time lapse coverage of the event to examine the long-term motion of the ion and neutral cloud providing the ambient winds and electric field. In addition, two EMCCD cameras at Bermuda observed at 557.7 and 630.0 nm to look for emissions due to electron excitation.