Along‐Trajectory Acoustic Signal Variations Observed During the Hypersonic Re‐Entry of the OSIRIS‐REx Sample Return Capsule Open Access

Earth’s atmosphere encounters extraterrestrial material ranging from microscopic particles to meter‐scale objects, with entry velocities spanning 11.2 to over 72 km/s (Ceplecha et al., 1998). Whereas smaller particles typically ablate entirely at high altitudes, larger meteoroids and asteroids deposit substantial energy along their ablational trajectories, generating shock waves (Bronshten, 1983; Silber et al., 2018). Meter‐size meteoroids represent an important middle ground, bridging the gap between ablation‐dominated particles and larger impactors capable of reaching the surface. These objects present both a scientific and societal interest: they contribute to our understanding of planetary formation, atmospheric entry physics, and planetary defense strategies. However, our ability to perform detailed studies is hindered by the fact that these objects are statistically more rare than smaller, more numerous impactors. For instance, prediction windows are frequently limited to just a few hours (Jenniskens et al., 2021), restricting the ability to plan coordinated observational campaigns ahead of time. Ground‐based observational networks (Devillepoix et al., 2020) are further constrained to landmasses, leaving large regions of the Earth unmonitored, and many detections rely on sensors designed for other purposes, such as U.S. government systems that observe meteoroids incidentally as secondary phenomena (Nemtchinov et al., 1997; Silber, 2024b).

Sample return capsules (SRCs) provide an unparalleled opportunity to complement studies of meteoroid events, offering predictable and well‐characterized conditions that mitigate the challenges posed by the unpredictable nature and limited observational coverage of natural entries. These compact capsules, engineered for interplanetary missions, re‐enter Earth’s atmosphere at velocities exceeding 11 km/s, comparable to the lower range of meteoroid speeds. Unlike meteoroids, which are highly stochastic in their trajectories, compositions, and entry angles, SRCs have well‐known source parameters, follow predictable paths, experience negligible ablation, and do not fragment under nominal conditions (Silber et al., 2023). Their re‐entry timing is known far in advance, which provides ample time for observational campaign preparation. Therefore, it is immensely beneficial to leverage controlled re‐entries with the aim to investigate shock wave formation, energy deposition, and infrasound propagation.

Since the Apollo program, there have been only five interplanetary SRC re‐entries recorded by geophysical instruments, with the Origins, Spectral Interpretation, Resource Identification, and Security–Regolith Explorer (OSIRIS‐REx) SRC being the most recent (Silber et al., 2023; Carr et al., 2024; Fernando, Charalambous, Saliby, et al., 2024; Fernando, Charalambous, Schmerr, et al., 2024; Silber, Bowman, et al., 2024). The prior reentries, Genesis (ReVelle et al., 2005), Stardust (ReVelle and Edwards, 2006), and Hayabusa 1 and 2 (Ishihara et al., 2012; Sansom et al., 2022), have demonstrated that SRC re‐entries can be effectively captured by geophysical sensing techniques, particularly infrasound and seismic instrumentation. Infrasound, or a low‐frequency acoustic wave (<20 Hz), is generated as a byproduct of a shock wave decay. It can propagate efficiently over long distances with minimal attenuation (Evans et al., 1972), making it ideal for studying high‐energy atmospheric events such as bolides (ReVelle, 1976), rocket launches (Pilger et al. 2021), controlled re‐entries (ReVelle et al., 2005; Silber, Bowman, et al., 2024), and explosions (Bowman et al., 2025).

The OSIRIS‐REx SRC re‐entry, the first interplanetary capsule to return over the United States since Stardust in 2006, presented an exceptional opportunity for a comprehensive geophysical observational campaign. Launched by the National Aeronautics and Space Administration (NASA) on 8 September 2016, OSIRIS‐REx was designed to collect and return samples from the near‐Earth asteroid Bennu, a primitive body that contains clues about the early solar system (Lauretta et al., 2017). After successfully collecting samples in 2020, the spacecraft approached Earth on 24 September 2023, releasing its SRC for re‐entry at 10:41:55 UTC. The capsule entered Earth’s atmosphere at a velocity of 12.36 km/s and landed at the Utah Test and Training Range (UTTR) at 14:52:09 UTC (Francis et al., 2024).

A distributed sensor network can capture acoustic signals from distinct trajectory points during hypersonic re‐entry (e.g., Silber, 2024a). In particular, for hypersonic sources, the geometry of the Mach cone plays a critical role in signal propagation (Anderson, 2006). For example, at Mach 30 (Mach number is the ratio of the object’s speed and the local speed of sound), the Mach cone half‐angle is only 1.9°, resulting in predominantly ballistic shocks and focused energy emission pattern (ReVelle, 1976; Anderson, 2006; Silber et al., 2018). In near field (<300 km), ballistic shocks typically travel the shortest path and arrive nearly perpendicular to the trajectory, making them dominant contributors to observed signals. Therefore, an infrasound sensor network positioned in the proximity of the direct path of a high‐altitude moving source could, in principle, capture signals generated at distinct points along the trajectory, as for the first time practically demonstrated through observations of an earth‐grazing fireball (Silber, Ronac Giannone, et al., 2024). Building on this theoretical framework, we deployed a dense sensor network over a wide geographical region to experimentally investigate the extent to which signals can be captured from distinct points along the trajectory and how signal properties, such as period and amplitude, vary with altitude and propagation geometry. Our study demonstrates the potential of controlled SRC re‐entries to advance our understanding of natural meteoroid dynamics, refine shock wave propagation models, and improve methods for planetary defense and atmospheric monitoring.

Here, we present the analysis of infrasound signatures from the OSIRIS‐REx SRC re‐entry, recorded by a dense network of ground‐based sensors spanning Nevada and Utah, United States, as part of the largest geophysical campaign for a hypersonic re‐entry to date (Silber et al., 2023; Carr et al., 2024; Fernando, Charalambous, Saliby, et al., 2024; Fernando, Charalambous, Schmerr, et al., 2024; Silber, Bowman, et al., 2024). By leveraging the predictable dynamics of a controlled re‐entry, this study provides a unique framework for examining how infrasonic signal parameters, such as amplitude and period, depend on source altitude, SRC trajectory geometry, and propagation path.

A comprehensive geophysical observational campaign was prepared in advance of the OSIRIS‐REx SRC re‐entry (Silber, Bowman, et al., 2024). Leveraging NASA’s Entry, Descent, and Landing (EDL) trajectory data (Francis et al., 2024), sensors were strategically deployed to maximize the detection of geophysical signals. Instrument placement focused on two key regions: the vicinity of Eureka, Nevada, corresponding to the predicted location of peak heating, and the Nevada–Utah state line, ∼200 km downrange (Fig. 1). This latter region included the SRC’s transition from hypersonic to supersonic flight and the subsequent subsonic “dark flight” phase before landing at the UTTR (Ajluni et al., 2015; Francis et al., 2024; Silber, Bowman, et al., 2024). Deployment details are thoroughly discussed in Silber, Bowman, et al. (2024); here, we summarize the most relevant aspects.

The analysis focuses on infrasound signals recorded by a ground‐based network of 39 single‐sensor stations. All stations were equipped with the same sensors (Gem infrasound microbarometers) to eliminate potential instrument bias. These portable, low‐power sensors operate at a sampling rate of 100 Hz (Anderson et al., 2018). The network was arranged into three distinct lines and was designed to target specific regions of interest while remaining logistically feasible (Fig. 1, Table S1, available in the supplemental material to this article):

• T line: positioned nearly beneath the trajectory to capture near‐vertical arrivals from the hypersonic flight path.

• A line: an orthogonal transect near Eureka, Nevada, positioned to capture signals associated with the region of maximum energy deposition during peak heating.

• C line: a second orthogonal transect near the Nevada–Utah state line, designed to observe signal propagation up to dark flight (i.e., the SRC becomes subsonic and is no longer luminous).

The SRC entered the Earth’s atmosphere at 14:41:55.57 UTC above San Francisco, California, at a shallow angle of 8.2° (±0.08°) and a velocity of 12.36 km/s (Ajluni et al., 2015; Francis et al., 2024). The atmospheric interface was at 125 km over the equatorial radius, which corresponds to a geodesic altitude of ∼132.89 km. The SRC followed a southwest‐to‐northeast trajectory and landed at the UTTR. Because of a lack of detailed telemetry during the SRC descent, real‐time trajectory information is unavailable. NASA is currently in the process of reconstructing the trajectory based on limited observational data. At present, a modeled postflight reconstruction of the EDL trajectory has been made available (Francis et al., 2024).

The first step was to identify infrasound detections on all stations, which proved to be relatively straightforward because the signals were readily apparent in raw waveforms. The recorded time‐series data were processed using a Butterworth filter to isolate the infrasound signals from background noise. Details about this approach for processing and analyzing signals from high‐altitude hypervelocity sources can be found in Silber (2024a). Given the broadband nature of the acoustic energy, data were filtered using a fourth‐order zero‐phase Butterworth filter across multiple frequency bands to analyze signal characteristics comprehensively (Figs. S1–S3). The lower frequency cutoffs were varied between 0.1 and 1 Hz (0.1, 0.2, 0.25, 0.3, 0.4, 0.5, and 1 Hz), and the upper frequency cutoffs were varied between 20 and 45 Hz (20, 25, 30, 40, and 45 Hz). For each frequency band, we measured the following signal parameters: maximum amplitude, peak‐to‐peak amplitude, and signal period. Analyzing signals across multiple frequency bands enabled us to assess the sensitivity of period and amplitude measurements to filter bounds and to quantify uncertainties in the derived signal parameters.

Because of the short signal duration, along with shock wave attenuation and relaxation influencing negative phase features, we developed a robust and repeatable method for signal period measurement tailored to the characteristics of the recorded signals. Specifically, the period was determined using zero crossings at the point of maximum amplitude, focusing on the first half‐cycle because it consistently exhibited the most well‐defined and stable signal structure. This value was then multiplied by two to estimate the full‐cycle period. The same methodology was uniformly applied across all recorded signals and frequency bands, ensuring consistency, comparability, and reproducibility in the measurements.

Propagation modeling was conducted using InfraGA, an open‐source computational ray‐tracing tool for simulating infrasound propagation in a 3D, spherical Earth framework (Blom, 2014). Using geometric ray‐tracing techniques, it calculates acoustic paths, including eigenrays that connect the source to the receiver (Blom and Waxler, 2017). InfraGA incorporates realistic atmospheric parameters (e.g., temperature, pressure, and wind), enabling detailed simulations over regional and global distances and accounting for the Earth’s curvature. We used the ground‐2‐space (G2S) atmospheric specifications (Hetzer, 2024) corresponding to the hour of the OSIRIS‐REx SRC re‐entry, extracted for two locations: Eureka, Nevada (39.7480° N, −115.7730° E), near the peak heating point, and the Nevada–Utah state line, farther downrange (40.1052° N, −113.9700° E) (Fig. S4). G2S combines real‐time observational data and climatological models to generate layered profiles of temperature, pressure, and wind from the surface to the lower thermosphere (Drob et al., 2003).

We used the EDL trajectory of the SRC with propagation modeling to systematically search for all potential eigenrays connecting the source to each infrasound station. This was inspired by the approach previously used in the context of meteor infrasound (Brown et al., 2011; Silber and Brown, 2014; Silber, 2024a). The eigenray search parameter space extended from −10° (nearly horizontal) to −89° (nearly vertically downward). Rays were traced from ∼2000 densely spaced points along a linear segment of the trajectory. Each point was treated as an independent source. This accounts for variations in source altitude, trajectory geometry, and atmospheric conditions, and enables a comprehensive representation of potential signal arrivals at each station.

To estimate the most probable source altitudes, the modeled propagation paths and signal travel times were compared with the observed signal travel times and station distances from the SRC trajectory segment (3D distance and ground track distance). Because a hypersonic source on a linear trajectory is likely to produce ballistic shocks, which typically arrive at a near‐perpendicular angle to the trajectory and follow the shortest path, we collected all eigenrays predicted to arrive at such geometry. It should be noted that inherent uncertainties might persist due to atmospheric variability because G2S profiles may not fully resolve transient or localized conditions during the event. In addition, minor inaccuracies in the EDL trajectory could introduce errors in source location inputs. Cross validation against observations from multiple stations mitigates these effects and improves the reliability of the results.

Figure 2 illustrates the method used to associate eigenrays with a potential shock source, using station A19 as an example. All stations were processed using the same approach. The three panels show the relationships between the predicted signal travel time and the parameters of interest for all existing eigenrays: altitude (Fig. 2a), ground distance (Fig. 2b), and total 3D distance (Fig. 2c). Although the results include a larger portion of the trajectory (travel times >245 s), the plots are zoomed in to the relevant region for clarity. The vertex in each panel, denoted by the red star, corresponds to the ballistic arrival, which represents the shortest possible travel time for the shock wave to reach the station. In Figure 2a, the source altitude is inferred as 56.2 km. The symmetry in the ray paths around the vertex indicates the consistency of raytracing with ballistic arrivals. The yellow points indicate a one‐second uncertainty in travel time, which translates to an altitude uncertainty of ∼ ±0.75 km. In Figure 2b, the altitude is consistent with that inferred in Figure 2a (56.4 km), emphasizing agreement between ground distance and travel‐time results. In Figure 2c, similar to Figure 2a,b, the results reaffirm the inferred source altitude of ∼56.1 km. In the bottom two panels, the color gradient, ranging from lower altitudes (green) to higher altitudes (orange), reflects the altitudinal variation of eigenrays between 51 and 60 km altitude (the altitude range probed with raytracing extended 50–75 km for this station). To validate the consistency of this approach, the predicted arrival times were compared with the observed arrival times across all stations, showing excellent agreement. Finally, to derive the most probable source altitude and the associated error, we collated all eigenrays arriving within the uncertainty window (see Fig. 2a) and calculated mean and standard deviation.

InfraGA outputs attenuation in the form of geometric spreading (⁠$Tg$⁠) and absorption (⁠$Ta$⁠) in decibels (dB) for each eigenray. To account for propagation effects, we used these attenuation factors to convert the signals observed on the ground to their effective amplitude (⁠$As$⁠) at 1 km from the source. We conducted four suites of propagation modeling simulations at central frequencies of 1, 5, 10, and 20 Hz. The frequencies of 5 and 10 Hz encompass the infrasonic range of the SRC’s dominant frequencies. The correction factor (⁠$Acf$⁠) used to convert total attenuation (⁠$Tatt$⁠) from dB to linear amplitude is given by $Acf=10(Tatt/20)$⁠, in which $Tatt$ is the combined effect of geometric spreading and atmospheric absorption, that is, $Tatt=Tg+Ta$⁠. The source amplitude is then calculated as $As=Aobs/Acf$⁠, with units of pascals. To account only for geometric spreading, the conversion becomes $As=Aobs/10(Tg/20)$⁠. We note that the gain in pressure amplitude due to the specific acoustic impedance contrast as the signal travels from the SRC altitude to the ground nearly cancels out geometric attenuation in these cases.

The recorded signals exhibit a characteristic N‐wave signature (Dumond et al., 1946), with a prominent initial half‐cycle followed by an extended and more dispersed negative phase. In addition, a coda is present following the main signal, possibly arising from atmospheric scattering or reflections. An apparent precursor energy seen in the signals is a digitizer artifact. Figure 3 shows representative filtered signals recorded on sensors along the T line, which was situated nearly vertically beneath the SRC’s flight path. The lateral offset of the stations from the trajectory ranged between 1.1 and 2.2 km (Table S1). Relative ground distances (west to east) from T1 to other stations are as follows: 56.5 km (T7), 64.7 km (T8), 159.4 km (T10), 164.4 km (T11), and 215.9 (C7). Notably, the earliest signal was recorded at C7 and the latest at T1, opposite the flight direction. This reversal is attributed to the SRC’s lowest altitude during the overflight near C7 (∼44 km) and highest at T1 (∼62 km), resulting in a shorter propagation distance for the signals to reach the eastern stations. Notably, the signal amplitude is evidently lower and the period longer for higher altitudes, as observed near T1, whereas signals originating from lower altitudes, such as near C7, exhibit higher amplitude and shorter periods, and a greater contribution of higher frequencies. The signal measurements for all stations are given in Table S2. The signal properties for all the stations are listed as measured.

Raytracing results confirm that the observed infrasound arrivals across all stations align remarkably well with ballistic shock predictions, having virtually one‐to‐one (to a fraction of a second) correspondence to the EDL trajectory timing at derived altitudes. This alignment persists throughout the entire trajectory segment sampled by the infrasound network, spanning altitudes from 62.2 km down to 43.7 km. This agreement reflects the self‐consistency of the raytracing methodology and supports the hypothesis that the recorded signals are dominated by ballistic arrivals from hypersonic shock sources. This enabled a reconstruction of the overflight times associated with specific altitudes, further reinforcing the reliability of the inferred shock source altitudes. The source altitudes derived through raytracing are given in Table S3.

Figure 4 provides a 3D visualization of eigenrays for two representative stations from each of the three sensor lines: line A (Fig. 4a,b), line C (Fig. 4c,d), and line T (Fig. 4e,f). Eigenrays are color coded according to travel time, with yellow representing the shortest times corresponding to ballistic arrivals and dark blue indicating longer travel times (and consequently nonballistic arrivals). In each panel, the most probable source altitude, which is also associated with the ballistic shock arrival, is denoted with a red star. For clarity, only every tenth eigenray is displayed. We also supply GIF animations with 3D results in the supplemental material. The gradient from faster (yellow) to slower (blue) travel times reflects altitudinal and temporal variations in eigenray paths, with the color scale adjusted for each station to capture trajectory‐specific differences.

Having derived the most probable source altitudes, we plotted these points as a function of sensor distance along the SRC ground track (Fig. 5a) and station arrangement across three sensor lines (A, T, and C) (Fig. 5b). The data points are color coded according to the signal peak‐to‐peak amplitude, indicating the relationship between altitude, geographic distribution, and signal amplitude. Figure 5 demonstrates that each sensor captured signals from a slightly different part of the SRC’s trail. The steady decrease in source altitude with an increasing ground track distance reflects the gradual descent of the SRC along its trajectory (Fig. 5a). Sensors closer to Eureka, Nevada, captured signals originating from higher‐altitude portions of the trail (∼62–55 km), whereas those near the Nevada–Utah border recorded signals from lower‐altitude regions (∼50–44 km). The amplitude variation, with lower‐altitude signals exhibiting higher peak‐to‐peak amplitudes, further emphasizes the distinct phases of the SRC’s flight path captured by the network. The arrangement of stations across the three sensor lines (A, T, and C) illustrates the spatial distribution of captured signals (Fig. 5b). Stations in the Nevada region recorded signals corresponding to the earlier, higher‐altitude segments of the trail, whereas stations near the Nevada–Utah border captured signals from later, lower‐altitude segments. The color‐coded amplitudes show a consistent pattern: signals from higher altitudes have lower amplitudes, whereas those from lower altitudes have higher amplitudes.

Figure 6 shows the relationship between signal peak‐to‐peak amplitude and signal period, with data points color coded according to the inferred source altitude. The altitude data provide the context into how the shock source height affects signal characteristics. The signals cluster into two distinct groups corresponding to the two primary deployment regions: (1) high‐altitude sources near the peak heating region (∼61 km), and (2) lower‐altitude sources near the maximum dynamic pressure region (∼44 km). Signals from higher altitudes (Eureka, Nevada) exhibit longer periods and lower amplitudes, forming a cluster in the lower‐right portion of the plot. In contrast, signals originating from lower altitudes (Utah–Nevada) have shorter periods and higher amplitudes, forming a cluster in the upper‐left region.

We performed regression analysis using the software package OriginPro to evaluate the sensitivity of signal period and amplitude to source altitude and distance from the trajectory (Fig. 7, Table S4). We investigated the relationships between signal period and amplitude and the following parameters: source altitude (Fig. 7a,b), total 3D distance from source to sensor (Fig. 7c,d), and ground distance from the trajectory to sensor (Fig. 7e,f). The equations and regression statistics, including Pearson’s r, $R2$ coefficient of determination (COD) and adjusted $R2$⁠, are provided within each panel and in Table S4. The regression analyses show that source altitude and total 3D distance are the dominant factors influencing signal amplitude and period, with high correlations in Figure 7a–d. A clear positive correlation is observed between signal period and source altitude (⁠$R2=0.85$⁠, Pearson’s r = 0.92) and a negative correlation between amplitude and source altitude (⁠$R2=0.80$⁠, Pearson’s r = −0.89). In terms of the total 3D distance, there is a positive correlation between the signal period and 3D distance (⁠$R2=0.73$⁠, Pearson’s r = 0.85), and a negative, albeit weaker, correlation between amplitude and 3D distance (⁠$R2=0.64$⁠, Pearson’s r = −0.80). Finally, there is a weak correlation between signal period (⁠$R2=0.17$⁠, Pearson’s r = 0.41) and amplitude (⁠$R2=0.11$⁠, Pearson’s r = −0.34), respectively, and ground distance from the trajectory. The source amplitudes corrected for attenuation show similar trends.

Before we delve into the discussion and implication of our results, it is worth mentioning the fundamental differences between shock waves produced by ablating bodies (e.g., meteoroids) and nonablating bodies (e.g., SRCs and similar artificial objects). Unlike SRCs, meteoroids, composed of naturally occurring materials with varying porosity, composition, and structural integrity, undergo intense ablation and often fragmentation during atmospheric entry (Romig, 1964; Bronshten, 1983; Ceplecha et al., 1998). Ablation‐driven shock waves are significantly stronger than and overtake the initial cylindrical shock waves originating from the bow shock envelope (see Silber et al., 2018, for discussion). Depending on the atmospheric density and composition, and typically within the distance corresponding to the blast wave radius, the two waves coalesce into a stronger, unified shock wave, which continues to expand radially (Tsikulin, 1970; Bronshten, 1983). Sufficiently strong shock waves are often accompanied by secondary phenomena, including luminosity, plasma formation, hyperthermal chemical reactions, and the introduction of ablation‐derived materials into the upper atmosphere (Ceplecha et al., 1998; Silber et al., 2018). The coupling of these processes with the surrounding atmosphere introduces variability in the amplitude and frequency content of the resulting acoustic signal. Although this variability can complicate signal interpretation, it also produces distinct acoustic signatures that may be used to distinguish between different hypersonic sources.

In contrast, SRC re‐entries, such as OSIRIS‐REx, are designed to maintain structural integrity and shape along their atmospheric trajectory. A lower velocity regime and the use of engineered materials minimize ablation and prevent fragmentation during hypersonic descent. Under nominal conditions, SRC‐generated acoustic signatures are not contaminated by shock waves from more complex, ablating sources with highly varying and often poorly known parameters, which allows for more comprehensive interpretation of information content from unknown acoustic signals originating in the Earth’s atmosphere. As such, these events serve as an excellent platform for study of the fundamentals of acoustic signals generated by nonablational shock waves. Such signals may serve as a control or a well‐defined baseline, enabling direct comparison with the signals from more complex sources.

The shock waves generated by SRCs are predominantly ballistic (expanding radially and perpendicularly to the axis of propagation). The comprehensive study of these shock waves is fundamentally important because they arise from the ballistic shock envelope of aerodynamically predictable and steady hypersonic motion of an SRC through the atmosphere. In principle, cylindrical shock waves, originating from a nonablating body with a bow shock envelope are simpler to interpret. They are often modeled as conical wavefronts, with propagation dominated by the Mach cone geometry. These shocks exhibit predictable characteristics, allowing for more straightforward modeling and interpretation of the resulting infrasound signals (ReVelle and Edwards, 2006; Silber et al., 2015). This comparison underlines the stochastic and highly variable nature of meteoroid‐generated shocks versus the controlled and repeatable dynamics of SRC re‐entries.

The dense network of infrasound sensors deployed across Nevada and Utah enabled detailed analysis of the acoustic wavefield generated by the OSIRIS‐REx SRC re‐entry as a function of source altitude, propagation path, and distance from the trajectory. The observed signals exhibited clear trends correlating with source altitude with signals originating from higher altitudes displaying lower amplitudes and longer periods (Fig. 3). This behavior aligns with theoretical predictions rooted in the physics of shock wave propagation through a rarefied and stratified atmosphere before it transitions into an acoustic wave. At higher altitudes, where the atmospheric density and pressure are significantly lower, the energy carried by the shock wave is spread over a larger volume, resulting in greater attenuation and reduced peak amplitudes. Fundamentally, the amplitude of the shock wave decreases approximately as a function of the inverse of the distance from the source, modified by atmospheric absorption factors dependent on altitude (e.g., Plooster, 1968; ReVelle, 1976; Pierce and Beyer, 1990; Zel’dovich and Raizer, 2002; Silber et al., 2018). Accounting for the effect of attenuation, we have empirically demonstrated that the amplitude is predominantly governed by the source region rather than the propagation path, at least within the ranges investigated.

The observed N‐wave (Dumond et al., 1946) signatures (Fig. 3) in the recorded signals were consistent across all stations, characterized by a prominent positive half‐cycle (compression phase) followed by an extended negative phase (rarefaction phase). This is typical of shock‐generated acoustic waves propagating in stratified atmospheres (Plooster, 1970; ReVelle, 1976). The asymmetry in the wave arises because nonlinear steepening during shock generation is gradually offset by relaxation as the wave propagates. The negative phase, in particular, becomes more diffuse due to shock attenuation and frequency‐dependent relaxation effects. At higher altitudes, where air density and pressure are lower, these effects are amplified by greater geometric spreading and enhanced vibrational relaxation of atmospheric molecules, resulting in further attenuation and broadening of the waveform (e.g., Plooster, 1968; Bass et al., 1972, 1995; Zel’dovich and Raizer, 2002).

The longer periods observed for signals originating from higher altitudes can be attributed to the distribution of acoustic energy from the attenuating shock wavefront. In rarefied regions of the atmosphere, the initial shock envelope is larger and shock wave transitions into a weak shock or acoustic wave at larger radial distances, affecting temporal spreading. The acoustic signal is further influenced by the attenuation of high‐frequency components, which are preferentially absorbed in low‐density conditions due to molecular relaxation processes (e.g., Plooster, 1970; Sutherland and Bass, 2004). Consequently, the signals detected at ground stations from higher altitudes are dominated by lower frequencies, leading to longer periods. Energy dissipation also plays an important role. As the shock wave propagates downward, the cumulative energy flux diminishes with altitude due to viscous dissipation and thermal conduction. This attenuation is quantified by the absorption coefficient, which increases with decreasing density at higher altitudes (Sutherland and Bass, 2004). The relationship between amplitude, period, and altitude in the near‐source region is thus governed by the combined effects of rarefaction, stratification, and molecular relaxation, which collectively influence the observable signal characteristics. These findings are consistent with prior studies on meteoroid‐generated infrasound, in which altitude‐dependent signal attenuation was observed (ReVelle, 1976; e.g., Silber and Brown, 2014).

The clustering of signal parameters near the two distinct trajectory regions (Fig. 6)—the peak heating zone (higher altitude) and the dynamic pressure maximum (lower altitude)—illustrates the interplay of competing effects on shock wave propagation. The amplitude correction applied through InfraGA assumes a reference of 0 dB at an arbitrary distance of 1 km from the source, irrespective of altitude (Blom and Waxler, 2017), simplifying the representation of wave propagation dynamics. This assumption does not fully capture the complexities of source conditions, particularly the coupling of shock waves with the atmosphere at different altitudes. At higher altitudes, the shock wave is stronger at the source due to greater velocities and higher Mach numbers associated with the peak heating region. However, in the rarefied atmosphere, this is offset by an exponentially lesser number density of molecules and atoms in the shock front. Consequently, the energy coupling between the shock wave and the surrounding medium is less efficient, and the shock weakens because it propagates outward (Zel’dovich and Raizer, 2002). Thus, even at the conceptual 1 km reference distance, the corrected amplitude is lower, reflecting the combined effects of weaker energy coupling and greater geometric spreading in the first kilometer of propagation. In contrast, at lower altitudes near the dynamic pressure maximum, the shock wave is initially weaker due to lower velocities and Mach numbers. However, the denser atmosphere increases energy coupling, allowing more of the energy to propagate as acoustic waves (ReVelle, 1976). Our observations are consistent with theoretical expectations of shock wave decay in stratified atmospheres and validate the ability of ground‐based infrasound networks to capture the nuanced effects of altitude on shock wave characteristics. The amplitude trends encode information about both the source conditions and propagation dynamics, demonstrating the utility of infrasound sensing for trajectory analysis. The apparent higher amplitudes at lower altitudes, despite the initially weaker shocks, are consistent with theoretical predictions and provide a robust framework for studying shock wave dynamics in both controlled and natural atmospheric entry events.

The raytracing results showed excellent agreement between the predicted and observed arrival times, affirming the robustness of the propagation modeling methodology. The observed ballistic arrivals, corresponding to the shortest travel paths, provided reliable estimates of source altitudes along the SRC trajectory. This consistency supports the fundamental assumption that hypersonic shocks propagate in a ballistic manner, traveling near perpendicularly to the trajectory with minimal deviation due to atmospheric effects in the near field. The agreement between raytracing‐derived altitudes and the clustering of signal amplitude and period further supports the capability of distributed infrasound networks to capture trajectory‐dependent variations in shockwave propagation. As mentioned earlier, stations closer to the trajectory consistently detected signals originating from lower altitudes, which exhibited higher amplitudes and shorter periods due to increased atmospheric density and stronger coupling of the shock wave to the medium. Conversely, stations farther away captured signals from higher altitudes, characterized by lower amplitudes and longer periods as a result of reduced air density and greater propagation distances. These altitude‐dependent signal characteristics validate the reliability of ballistic shock assumptions and demonstrate the feasibility of using infrasound to perform accurate trajectory reconstructions. This will be the subject for a future article.

Finally, because the frequency of space debris re‐entries is expected to increase in the coming decades, understanding the differences between natural and artificial objects becomes increasingly critical. Unlike SRCs, space debris are composed of different materials, some of which may fragment or partially ablate during re‐entry, resulting in hybrid shock wave characteristics that lie between the extremes of meteoroids and engineered capsules. Studies like this, leveraging the controlled conditions of SRC re‐entries, provide a framework for interpreting the dynamic behavior of space debris and assessing its potential risks. Moreover, they contribute to the refinement of observational techniques and propagation models, offering tools to monitor both natural and artificial high‐energy atmospheric entries effectively.

The OSIRIS‐REx SRC re‐entry presented an unparalleled opportunity to investigate infrasound propagation dynamics and their dependence on source altitude and trajectory parameters. Utilizing a network of 39 ground‐based infrasound sensors across Nevada and Utah, this study constitutes the most extensive geophysical campaign ever conducted for a hypersonic reentry. Raytracing results, rigorously validated against observed travel times, confirmed ballistic arrivals across all stations, with source altitudes inferred to range from 44 to 62 km. These findings demonstrate the reliability of raytracing methodologies in characterizing hypersonic re‐entries. The infrasound signals exhibited a clear dependence on source altitude: higher‐altitude sources produced signals with lower amplitudes, longer periods, and diminished high‐frequency content, reflecting the interplay between shock wave strength and the shock source region. Regression analysis further reinforced the strong correlations between signal parameters and source altitude. Our findings indicate that, after accounting for attenuation, the observed amplitude is primarily dictated by the source characteristics, with minimal influence from the propagation path within the ranges studied.

The controlled conditions of the OSIRIS‐REx SRC re‐entry enabled a robust experimental investigation of atmospheric entry phenomena, offering benchmark data for direct comparisons with natural meteoroid events. Unlike the stochastic nature of meteoroid entries, SRC re‐entries provide a well‐characterized analog for validating theoretical models under repeatable conditions. The comparison between predictable ballistic shocks from SRCs and ablational shocks from meteoroids illuminates the broader utility of SRC re‐entries in refining shock wave propagation models. This study demonstrates the utility of SRC re‐entries in advancing planetary defense strategies, refining shock wave propagation models, and improving global monitoring systems for high‐energy atmospheric phenomena, including space debris re‐entries. With the anticipated rise in orbital traffic, these findings are crucial for addressing the increasing challenges of managing re‐entries in Earth’s atmosphere. Moreover, the results provide a foundation for understanding atmospheric entry dynamics in planetary atmospheres, contributing to the study of high‐energy phenomena in extraterrestrial environments.

InfraGA is freely available at https://github.com/LANL-Seismoacoustics/infraGA. Ground‐2‐space (G2S) is available on request through the National Center for Physical Acoustics (NCPA) at the University of Mississippi. The supplemental material includes additional figures (Figs. S1–S4) and all tables generated as part of this work (Tables S1–S4). The regression analysis was done using the proprietary graphing and analysis software package OriginPro, developed by OriginLab: https://www.originlab.com. The infrasound data were collected by an experimental network of the U.S. Government and are not currently available for release to the public. All websites were last accessed in September 2024.

The authors acknowledge that there are no conflicts of interest recorded.

The authors thank the two anonymous reviewers and the Editor‐in‐Chief Allison Bent for their helpful feedback. This work was supported by the Nuclear Arms Control Technology (NACT) program at the Defense Threat Reduction Agency (DTRA). The authors thank M. Moreau (NASA) and S. Francis (Lockheed Martin Space) for sharing the Entry, Descent, and Landing (EDL) trajectory. The authors gratefully acknowledge the Board of Eureka County Commissioners, the Eureka Municipal Airport personnel, the Bureau of Land Management Ely District Office, and the Bureau of Land Management Salt Lake Field Office. This article has been authored by an employee of National Technology and Engineering Solutions of Sandia, LLC under Contract Number DE‐NA0003525 with the U.S. Department of Energy (DOE). The employee owns all rights, title, and interest in and to the article and is solely responsible for its contents. The U.S. Government, and the publisher, by accepting the article for publication, acknowledges that the U.S. Government retains a nonexclusive, paid‐up, irrevocable, worldwide license to publish or reproduce the published form of this article or allow others to do so, for U.S. Government purposes. The DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan: https://www.energy.gov/downloads/doe-public-access-plan (last accessed September 2024). This article describes objective technical results and analysis.

Any subjective views or opinions that might be expressed in the article do not necessarily represent the views of the U.S. DOE or the U.S. Government.