Photometry of Fireballs using High Frame Rate Cameras

1. Problem Statement

Characterizing the physical structure and fragmentation behavior of meteoroids requires high-temporal-resolution photometry. Current observational methods face significant limitations:

• Long-Exposure Imaging (e.g., Desert Fireball Network): Uses mechanical or electro-optical shutters to encode timing. While excellent for astrometry, this method inherently sacrifices photometric data during shutter gaps and suffers from saturation during bright events.
• Standard Video (30 FPS): Lacks the temporal resolution to resolve rapid fragmentation events (flares) and is prone to sensor saturation due to limited dynamic range (typically 8–10 bits).
• Radiometers (PMT/Photodiodes): Offer high sampling rates (500–5000 Hz) and high dynamic range but lack imaging capabilities. They require complex calibration against image-based systems, are sensitive to ambient light (moonlight/daylight), and cannot provide simultaneous trajectory data without a separate camera system.

There is a critical need for a low-cost, integrated solution that combines the high temporal resolution of radiometers with the imaging and calibration capabilities of cameras, capable of capturing the full dynamic range of fireballs (from faint to magnitude -15+) without saturation.

2. Methodology

The authors developed and tested a prototype system using modern CMOS machine vision cameras, evolving through two field deployments:

• Hardware:
  • Camera: Allied Vision Alvium 1800 U-052 (Sony IMX426 CMOS sensor, Global Shutter, 12-bit depth).
  • Optics: Initially a custom-sealed Fujinon fish-eye lens; later upgraded to an IP-rated Commonlands stereographic fisheye lens for easier deployment.
  • Frame Rates: Operated at 470 FPS (Perenjori prototype) and 500 FPS (Forrest prototype).
• Software & Algorithms:
  • Detection: Background subtraction using masked stacked images to filter out the Moon and static reflections. Contour detection (OpenCV) triggers event recording.
  • Region-of-Interest (ROI): To manage data throughput, images are cropped to 100x100 pixels centered on the fireball centroid.
  • Detection Localised Auto-brightness Control (DLAC): A novel real-time algorithm that adjusts camera gain and exposure based only on the brightness within the fireball's contour, rather than the whole frame. This allows the system to maintain a high Signal-to-Noise Ratio (SNR) for the meteor while preventing pixel saturation.
• Validation Strategy:
  • Photometric Accuracy: Validated against the Moon (as a stable bright source) and compared against co-located 30 FPS cameras and FRIPON stations.
  • Scientific Application: Applied to a bright fireball (DN250711_02) to perform semi-empirical fragmentation modeling using the Borovička et al. (2013) model.

3. Key Contributions

• DLAC Implementation: The successful integration of a localized auto-brightness control loop that dynamically adjusts exposure (20–1800 µs) and gain (1–30 dB) in real-time. This overcomes the native dynamic range limitations of the sensor.
• High-FPS Photometry: Demonstrated that high-frame-rate CMOS cameras (470–500 FPS) can produce calibrated photometry comparable to standard 30 FPS systems while capturing sub-second fragmentation dynamics.
• Integrated Instrumentation: Proved that a single, low-cost imaging system can replace the need for hybrid setups (camera + radiometer), providing both precise trajectory data (via astrometry) and complete, unsaturated light curves.
• Technology Readiness: The system achieved Technology Readiness Level 7 (TRL-7) by being successfully demonstrated in an operational field environment.

4. Results

• Dynamic Range Expansion: The DLAC algorithm increased the system's effective dynamic range by 8.51 magnitudes, resulting in a total system dynamic range of approximately 13.91 magnitudes (covering apparent magnitudes from -3 to -17).
• Photometric Accuracy:
  • Moon Calibration: Measured magnitude of -11.23 vs. JPL Horizons reference of -11.195, a deviation of only 0.035 magnitudes.
  • Cross-Validation: During the May 6, 2024 fireball, the high-FPS prototype agreed with 30 FPS and FRIPON cameras within ±0.5 magnitudes in unsaturated regions.
• Fragmentation Analysis (DN250711_02):
  • Captured a fireball with a peak magnitude of nearly -15 with minimal saturation (only 4 discrete data points saturated).
  • Enabled semi-empirical modeling revealing an initial mass of 80 kg and a bulk density of 2200 kg/m³ (suggesting a carbonaceous chondrite composition).
  • Identified three eroding fragment events and a terminal disruption at 2.08 MPa dynamic pressure, where 80% of the remaining mass crumbled into dust.
  • The model successfully reproduced the main light curve features, though fine-scale peak sharpness was limited by the model's 0.02s time step.

5. Significance

• Solving the Saturation Bottleneck: This system resolves the primary observational gap in current fireball networks: the inability to capture complete light curves for bright events without saturation or data gaps.
• Enhanced Physical Modeling: The high-cadence data allows for detailed reconstruction of meteoroid internal structure, mechanical strength, and fragmentation dynamics, which are impossible to derive from low-frame-rate or shuttered imagery.
• Future Deployment: The validated system is intended to become a core component of the Global Fireball Observatory (GFO). It will be deployed in a hybrid configuration alongside dedicated high-resolution astrometric cameras, enabling the simultaneous capture of precise orbital trajectories and high-resolution physical data.
• Cost-Effectiveness: By utilizing off-the-shelf CMOS sensors and eliminating the need for complex PMT radiometers or secondary timing hardware, this approach offers a scalable, low-cost solution for global meteor monitoring.