The Carruthers Mission Concept and Performance

The Carruthers Geocoronal Observatory (Carruthers), formerly GLIDE, is a NASA Heliophysics Science Mission of Opportunity implemented through the Solar Terrestrial Probes (STP) Program and launched as a rideshare in September 2025. Carruthers is the first spaceflight mission dedicated to continuous global imaging of Earth’s hydrogen exosphere through observations of geocoronal Lyman-α emission at 121.6 nm. The observatory operates at distances of 1.3–1.7 million km in a halo orbit about the Sun–Earth L1 point, enabling retrieval of the three-dimensional distribution of atomic hydrogen, the dominant constituent of the exosphere, on hourly timescales and unprecedented spatial resolution. These new data provide the key to understanding processes governing atmospheric escape, geospace coupling, and solar-wind interaction.

Carruthers carries the GeoCoronal Imager (GCI), a dual-channel ultraviolet imaging instrument comprising the Narrow-Field Imager (NFI) for high-resolution observations of the inner exosphere and the Wide-Field Imager (WFI) for synoptic imaging of the extended hydrogen halo. Along with a student-provided experiment, the instrument suite is the sole payload aboard a three-axis-stabilized spacecraft designed to support nadir viewing throughout the ∼ 178 day halo orbit about L1. Observations also measure the interplanetary Lyman-α background, enabling separation of heliospheric and geocoronal emissions. The student-led experiment monitors solar Lyman-α and extreme ultraviolet emission during portions of the orbit. Science data are returned through the Deep Space Network at data rates up to 1 Mbit s−1. The two-year baseline mission begins in March 2026, with propellant reserves capable of supporting more than ten years of orbit maintenance and maneuvers.

The measurement approach is designed to retrieve the global, three-dimensional distribution of neutral hydrogen in Earth’s exosphere by imaging resonantly scattered solar Lyman-α emission at 121.6-nm from a stable vantage point near L1. This observing geometry enables nearly continuous, high-cadence global imaging of the geocorona, overcoming the spatial, temporal, and line-of-sight limitations that have constrained previous measurements from low Earth orbit or highly elliptical Earth orbits. Carruthers’ observatory is launched as a secondary rideshare where the two other observatories (primary: NASA IMAP, second secondary: NOAA SWFO-L1) are also targeting an L1 halo orbit for operations. It attains orbit about L1 by performing a series of propulsive maneuvers after separation from the launch vehicle. Approximately one week after separation, an initial orbit shaping maneuver (OSM#1) corrects for small errors in the initial trajectory provided by the Falcon 9 launch vehicle. After this first maneuver, cruise phase begins, the payload ICP is turned on and the payload is warmed to bakeout any residual contaminant. Payload doors are opened 21 days after separation to begin commissioning of the instruments. At approximately 60 days from separation, a second maneuver places the observatory on track to intercept a point in the halo orbit about L1. This point is reached about 100 days after separation at which time the Lissajous orbit insertion takes place that inserts the spacecraft into a stable ∼ 178 day orbit about the L1 point. Stationkeeping maneuvers are planned on a 2-3 month basis after that and continue throughout the operational lifetime of the mission. Carruthers can maintain Earth-nadir pointing throughout the science phase, viewing the exosphere at angles between approximately 8◦ and 28◦ from the Earth–Sun line.

The primary payload, the GeoCoronal Imager (GCI), implements a dual-channel imaging strategy using co-aligned telescopes with nested fields of view optimized for different regions of the exosphere. One channel emphasizes high spatial resolution observations of the inner exosphere, where density gradients are steep and where thermal evaporation, charge exchange, and ion–neutral coupling strongly influence structure. The second channel provides wide-angle, high-sensitivity imaging of the extended hydrogen halo, enabling characterization of large-scale morphology shaped by solar radiation pressure and interactions with the heliosphere. These channels are termed the Narrow and Wide Field Imagers (NFI and WFI). Though the reflective optical designs differ, the imagers share key design features, with identical filter wheels, ultraviolet cameras and single-shot door mechanisms. Simultaneous operation of the two channels allows Carruthers to capture the full dynamic range of Lyman-α brightness in a single observation sequence. This nested approach ensures continuity between inner and outer regions of the exosphere and supports the retrieval of global hydrogen density profiles without relying on extrapolation between disparate datasets.

Carruthers measures the Lyman-α radiance distribution with a target absolute accuracy of approximately 30% within defined spatial and temporal resolution elements. The isolation of the geocoronal signal from instrumental and astrophysical backgrounds is achieved through a combination of spectral discrimination, observing geometry, and on-orbit calibration. A key advantage of the Carruthers measurement approach is the ability to measure the Ly-α interplanetary hydrogen (IPH) background, collect these observations into a continuous belt of observations near the ecliptic, and subtract it directly from science images. During the science phase, the interplanetary contribution is measured using the outer annulus of each wide-field image, spanning approximately 30–35 RE. During the initial commissioning phase, additional perimeter scans are performed to map the detectable exospheric boundary and verify the validity of using the outer WFI pixels to characterize the IPH. Absolute radiometric calibration is achieved through a combination of stellar observations, filter-based spectral characterization, and detector background measurements. Bright Lyman-α stars allow for direct measurement of instrument responsivity at the line center, while observations of stars with strong extreme-ultraviolet continua enable characterization of the wavelength-dependent response across the instrument passband. End-to-end performance analyses based on measured instrument characteristics indicate substantial margin relative to signal-to-noise and retrieval accuracy requirements across anticipated observing conditions.

The GCI payload consists of two co-aligned telescopes, ultraviolet cameras, filter wheel mechanisms, deployable doors, high-voltage power supplies, and instrument control electronics. The payload housing is an aluminum enclosure, where all of these elements are either inside the box or mounted to the outside. The enclosure is attached to the spacecraft by three bipods. Carruthers carries a student-built solar monitor (COSSMo) that measures solar Lyman-α and soft X-ray irradiance to provide context for exospheric variability. COSSMo is regularly exposed to sunlight, while the main instrument enclosure is typically shaded from illumination by the fixed solar panel except for flat field observations. The GCI consists of two co-aligned imaging telescopes—the Near-Field Imager (NFI) and the Wide-Field Imager (WFI)—mounted on a common optical bench and supported by mechanisms, electronics, and calibration subsystems. The payload design draws extensively on heritage from the ICON FUV instrument while incorporating mission-specific adaptations required for long-duration nadir pointing from the Sun–Earth L1 environment. The GCI is organized around an aluminum optical bench that provides a common reference for the NFI and WFI optical trains, filter wheels, cameras, and fiducial alignment elements. The spacecraft star trackers are also mounted to the payload to maintain tight coupling between the payload boresight and the spacecraft attitude control determination. Major payload elements include NFI and WFI telescopes, ultraviolet camera assemblies, filter wheel mechanisms, deployable doors, high-voltage power supplies, the Instrument Control Package (ICP), the Preamp and Motor Drive electronics, and thermal and contamination control hardware.

Both NFI and WFI employ all-reflective ultraviolet optical designs optimized for Lyman-α imaging. The two telescopes differ in field of view and focal length to address distinct regions of the exosphere. The optical layouts are implemented using high-heritage mirror substrates with AlMgF2 coatings applied at both NASA Goddard Space Flight Center and commercial vendors. The optical bench and mirror mounts employ spherical washer stacks and precision fiducial reference surfaces to enable deterministic alignment. Alignment proceeds from individual optic characterization through full instrument integration, with wavefront error and point spread function verified at multiple stages. The final ultraviolet alignment and calibration were performed under vacuum conditions to ensure flight-like optical performance. Each GCI channel is equipped with an ultraviolet camera consisting of a sealed-tube image intensifier coupled through a fiber-optic taper to a solid-state focal plane array. The ultraviolet converter employs a KBr photocathode deposited on a microchannel-plate stack followed by a phosphor screen and optical taper to the detector. By optimizing the intensifier configuration and optical coupling to the CMOS detector, the system achieves angular resolution better than one arcminute across the field of view, well matched to the intrinsic spatial resolution of the microchannel plates and the detector pixel scale.

Both NFI and WFI employ identical six-position filter wheels driven by stepper motors with heritage from the AWE and ICON FUV missions. Filter positions include nominal Lyman-α science filters, out-of-band filters used for background characterization, and open and blocked positions. Deployable instrument doors protect the optics during launch and early mission phases. The door mechanisms provide one-time actuation using pin-pullers, spring-driven deployment, and end-of-travel damping to control shock during opening. The GCI electronics architecture consists of three primary subsystems: the Instrument Control Package (ICP), the Preamp and Motor Drive electronics, and the High Voltage Power Supplies (HVPS). The ICP serves as the instrument computer and power distribution hub, providing command and telemetry interfaces with the spacecraft, and managing science data handling. The PMD electronics provide stepper motor drive for the filter wheels and preamplification for sun-sensor photodiodes. The HVPS units supply high voltage to the ultraviolet intensifiers and are derived from heritage designs developed for the ICON mission. The ICP controls the filter wheels and the cameras of the GCI and co-adds pixels to match a selected resolution. It polls the spacecraft for timestamps and pointing quaternions determined by the spacecraft attitude determination and control system. All of the camera, filter wheel and HVPS states are included in the telemetry stream, along with the pointing quaternion at the start and end of each camera integration as well as the spacecraft position held by the onboard ephemeris. This last innovation simplifies the science data pipeline, and eliminates the need for a separate ancillary data product to interpret the images.

Thermal control of the payload is achieved through a combination of passive design features and active heater control. External surfaces are protected with multilayer insulation, while the internal optical surfaces are black anodized to suppress stray light and control radiative coupling. Contamination control measures included molecular adsorber coatings, purge hardware, vacuum bake-outs of sensitive components, and controlled integration and test procedures. These measures preserve ultraviolet throughput throughout the mission lifetime. The integration and test of the payload followed a deterministic approach beginning with the characterization at the optical level and progressing through full instrument assembly, alignment, and environmental testing. Initial optical alignment was performed at the UC Berkeley Space Sciences Laboratory, while final ultraviolet calibration was conducted at the Centre Spatial de Liège using vacuum collimator facilities. Post-calibration vibration testing was followed by re-verification of instrument performance in vacuum conditions. No adjustment of the instrument focus was required following the final alignment of the payload.

The Carruthers spacecraft provides a stable, power-positive, and thermally controlled platform for continuous ultraviolet imaging of Earth’s exosphere from a halo orbit about the Earth–Sun L1 point. The spacecraft was developed and integrated by BAE Systems (formerly Ball Aerospace), leveraging heritage from the Small Satellite Evolve product line missions. The design emphasizes simplicity and operational margin consistent with NASA Class D mission objectives and the requirement for long-duration autonomous operations. A baseline small-satellite bus architecture was tailored to support the Earth–Sun L1 staring mission geometry and the continuous nadir-pointing science mode required for geocoronal imaging. Carruthers was launched on 24 September 2025 as a rideshare payload on the IMAP mission mounted to an ESPA-Grande ring. Following separation from the launch vehicle, the spacecraft operates as an independent observatory. The spacecraft employs a compact aluminum primary structure with machined decks and honeycomb panels forming a stiff, low-mass bus. The payload is mounted to a dedicated payload deck using precision interfaces designed to maintain optical alignment and thermal isolation. The overall configuration accommodates the GeoCoronal Imager field-of-view requirements, star tracker lines of sight, antenna coverage, and solar array illumination without obstruction during nominal observing attitudes.

The attitude determination and control system provides precise pointing knowledge and stability to support continuous Earth-centered imaging. Attitude knowledge is derived primarily from two Sodern star tracker optical heads mounted near the instrument boresight, supplemented by coarse sun sensors and an inertial reference unit. Spacecraft attitude control is provided by a reaction wheel assembly arranged to deliver full three-axis control authority. Thrusters are used for momentum management, orbit correction maneuvers, and initial detumble following separation from the launch vehicle if the tip-off rate exceeds the capability of the reaction wheel array. During nominal science operations, small propulsive maneuvers are performed weekly to reduce momentum stored in the reaction wheels. Electrical power is generated by a body-mounted solar array and distributed through a direct-energy-transfer architecture with heritage charge-control algorithms. Energy storage is provided by a lithium-ion battery sized to support science operations, communications, and maneuvering with margin. Orbit insertion, halo-orbit maintenance, and momentum unloading are accomplished using a blowdown monopropellant propulsion system with four 5-N thrusters. The propulsion system provides sufficient ∆V margin to support both the baseline two-year mission and potential extended operations. Thermal control is achieved through a combination of passive insulation, radiative surfaces, and heater-controlled components. Thermal analysis indicates that all spacecraft and payload components remain within allowable temperature limits throughout the mission.

The observatory concept of operations and architecture is designed to minimize thermal gradients in the payload to maintain stability of the optics and detectors. By default the instrument is expected to be operated with the-Z component of the observatory frame of reference intersecting the Sun-Earth line, with the payload effectively shaded by the spacecraft body and solar panel. Illumination of the +X or-X panels is also an option. Illumination of the +Z face is avoided, except during flat field observations, to reduce thermal input to the top of the GCI instrument. Around the…

Improvements to AI Systems:

1. Autonomous Mission Planning and Anomaly Response

• Train an AI scheduler to generate time-tagged command sequences that balance science observations, calibration windows, DSN passes, and momentum-unloading maneuvers, while respecting thermal constraints (e.g., avoiding +Z illumination except for flat fields).

• Implement a reinforcement learning agent that detects off-nominal telemetry patterns (e.g., filter wheel jams, HVPS drift, star tracker dropout) and autonomously triggers safe-mode transitions or corrective sequences, reducing reliance on human first responders.

2. Real-Time Data Fusion and Background Subtraction

• Develop a neural network that fuses simultaneous NFI/WFI images with spacecraft ephemeris, pointing quaternions, and solar irradiance (from COSSMo) to separate geocoronal Lyman-α from interplanetary hydrogen (IPH) background in real time, using the outer WFI annulus (30–35 RE) as a dynamic calibration reference.

• This enables onboard generation of IPH-subtracted images, reducing downlink data volume and enabling faster science alerts.

3. 3D Hydrogen Density Retrieval via Physics-Informed Neural Networks

• Replace traditional forward-model inversion with a physics-informed neural network (PINN) that ingests calibrated L1 images, observation geometry (8°–28° from Sun-Earth line), and solar Lyman-α flux to directly output global 3D hydrogen density distributions.

• The PINN can embed radiative transfer equations and charge-exchange physics as loss terms, achieving hourly tomographic reconstructions at unprecedented spatial resolution without iterative solvers.

4. Predictive Stationkeeping and Orbit Maintenance

• Use a transformer-based time-series model on DSN tracking data and MONTE-generated ephemerides to predict halo-orbit drift and optimal stationkeeping maneuver times (every 60–90 days), targeting zero X-velocity at X–Z plane crossings.

• This reduces fuel consumption and extends mission lifetime beyond the 10-year propellant reserve.

5. Automated Calibration and Instrument Health Monitoring

• Deploy a convolutional autoencoder to detect subtle degradation in UV camera response (e.g., KBr photocathode aging, filter wheel misalignment) by comparing stellar calibration images (bright Lyman-α stars) against baseline.

• Automatically adjust flat-fielding and radiometric correction coefficients in the Level 1 pipeline, maintaining the 30% absolute accuracy requirement without manual intervention.

6. Cross-Mission Data Assimilation for Space Weather Prediction

• Create a multi-source AI model that assimilates Carruthers geocoronal hydrogen densities with solar wind data (from IMAP and SWFO-L1 at the same halo orbit) and magnetospheric indices to forecast exospheric escape rates and geospace coupling dynamics.

• This enables predictive alerts for satellite drag variations and thermospheric density changes, benefiting low-Earth-orbit operations.

7. Optimized Science Planning Under Resource Constraints

• Implement a constrained optimization AI (e.g., mixed-integer programming with learned heuristics) to schedule observations that maximize scientific return (e.g., capturing solar radiation pressure effects, charge-exchange regions) while respecting DSN downlink windows (12 hours/week), power margins, and thermal limits (≤28° sun angle).

• The system can dynamically reprioritize observations based on real-time solar activity (from COSSMo) to capture transient exospheric events.

8. Autonomous Data Product Generation and Quality Control

• Build a generative AI pipeline that automatically produces Level 1 calibrated images and Level 2 density products, flagging anomalies (e.g., stray light, detector artifacts) using anomaly detection transformers.

• This reduces SDOC manual validation time and enables near-real-time public data release through NASA’s Space Physics Data Facility.

What the Improved AI System Can Do:

• Operate the Carruthers observatory with minimal human oversight, autonomously planning, executing, and adapting science operations for a decade-long extended mission.

• Produce real-time, globally resolved 3D hydrogen density maps of Earth’s exosphere every hour, directly usable for space weather forecasting and atmospheric escape studies.

• Seamlessly integrate with IMAP and SWFO-L1 data to create a multi-point L1 observatory network for heliospheric and geocoronal research.

• Reduce mission operations costs and latency by automating calibration, anomaly response, and data processing, while maintaining high scientific accuracy and open-data compliance.

The Carruthers Geocoronal Observatory (Carruthers), formerly GLIDE, is a NASA Heliophysics Science Mission of Opportunity implemented through the Solar Terrestrial Probes (STP) Program and launched as a rideshare in September 2025. Carruthers is the first spaceflight mission edicated to continuous global imaging of Earth's hydrogen exosphere through observations of geocoronal Lyman- alpha emission at 121.6 nm. The observatory operates at distances of 1.3-1.7 million km in a halo orbit about the Sun-Earth L1 point, enabling retrieval of the three-dimensional distribution of atomic hydrogen, the dominant constituent of the exosphere, on hourly timescales and unprecedented spatial resolution. These new data provide the key to understanding processes governing atmospheric escape, geospace coupling, and solar-wind interaction. Carruthers carries the GeoCoronal Imager (GCI), a dual-channel ultraviolet imaging instrument comprising the Narrow-Field Imager (NFI) for high-resolution observations of the inner exosphere and the Wide-Field Imager (WFI) for synoptic imaging of the extended hydrogen halo. Along with a student-provided experiment, the instrument suite is the sole payload aboard a three-axis-stabilized spacecraft designed to support nadir viewing throughout the about178 day halo orbit about L1. Observations also measure the interplanetary Lyman- alpha background, enabling separation of heliospheric and geocoronal emissions. The student-led experiment monitors solar Lyman- alpha and extreme ultraviolet emission during portions of the orbit. Science data are returned through the Deep Space Network at data rates up to 1 Mbit s-1. The two-year baseline mission begins in March 2026, with propellant reserves capable of supporting more than ten years of orbit maintenance and maneuvers.