A Multi-station Meteor Monitoring (M$^3$) System. II. system upgrade and a pathfinder network

Zhenye Li, Hu Zou, Ji Liu, Jiahui Ma, Qingkai Meng, Yongbing Cai, Xin Zhao, Xiangwen Li, Zhoudunming Tu, Baowei Zhang, Ran Wang, S. Wang, Feng LuarXiv 2024 · DOI · arXiv

Abstract

Meteors are vital indicators of the properties and origins of interplanetary dust particles. To facilitate large-scale data collection, we present the design and upgrade of the Multi-station Meteor Monitoring ($M^3$) system in China. The upgraded system transitions from single-camera setups to four-camera stations for full-sky coverage, supported by an integrated software pipeline and a cloud-based data platform. During its first 10 months of operation, a four-station pathfinder network detected 8,683 orbits. Analysis of these orbits reveals a bimodal velocity distribution and distinct orbital populations categorized by the Tisserand parameter, demonstrating the system's capability for high-precision automated meteor science.

Introduction

While meteor networks in Europe and North America are well-established, China has historically lacked a coordinated, high-precision automated network. Many regional observations have relied on amateur setups using fragmented software, which often results in degraded image quality and poor timing precision. To study the origin, mass distribution, and orbital evolution of meteoroids, scientists require massive volumes of high-quality data. The $M^3$ project aims to fill this geographic gap by creating a scientifically rigorous system that automates the entire pipeline—from initial detection to orbital calculation.

System Architecture and Hardware Upgrades

The $M^3$ system underwent a comprehensive overhaul to improve scalability and precision. The architecture is designed as an end-to-end pipeline that minimizes manual intervention.

The M³ system architecture, illustrating the data flow from hardware sensors through the meteorWorker processing units to the cloud-based Data Platform and User Experience interfaces.

As shown in the architecture diagram, the system is divided into four layers. The meteorWorker component handles real-time video detection (via C++ meteorThread) and detailed measurements (via Python meteorExtractor). A key innovation is the use of computer memory for data transfer between these components, rather than writing intermediate files to disk, which significantly accelerates processing.

Comparison of hardware specifications between the original and upgraded M³ meteor stations.

The hardware transition, detailed in the table above, included moving from 6.5mm to 8mm lenses. Although the 8mm lens has a narrower field of view, its larger aperture provides a deeper limiting magnitude (detecting stars down to magnitude 5.58 compared to 4.07 for the 6.5mm lens). This allows the system to detect more meteors overall despite the smaller sky coverage. Additionally, the upgraded stations feature significantly more powerful CPUs and increased RAM to handle the multi-camera data load.

Cloud-Based Data Processing

The system utilizes a Software as a Service (SaaS) model hosted on Aliyun (Alibaba Cloud). The Object Storage Service (OSS) stores raw and processed data, while "Function Compute" (serverless computing) automatically triggers multi-station orbit calculations whenever new data is uploaded. This automation allows 70% of detected meteors to be processed within 150 seconds, with cloud-based orbit determination taking only about 10 seconds once multi-station data is synchronized.

Key Results and Scientific Insights

The pathfinder network, consisting of four stations arranged in a diamond shape around Beijing, has proven highly effective:

  • Orbit Cataloging: In its first 10 months, the network calculated 8,683 orbits. Approximately 54% of these were linked to 265 known meteoroid streams, such as the Perseids and Geminids.
  • Orbital Classification: Analysis of the Tisserand parameter ($T_j$) showed a bimodal distribution with peaks at $T_j \approx 0$ and $T_j \approx 3$. This acts as a "fingerprint" for parent bodies, effectively separating isotropic (cometary) orbits from ecliptic (asteroidal) ones at the $T_j = 2$ threshold.
  • Astrometric Precision: The 8mm lens achieved a median fitting residue of 0.28 arc-minutes, a marked improvement over the 0.42 arc-minutes of the previous design.
  • Dark Flight Prediction: The system successfully tracked a bright fireball in early 2024, combining observational data with real-time weather balloon data to predict "dark flight" trajectories and potential meteorite landing sites.

Conclusion and Future Work

The $M^3$ system represents a significant step forward for meteor science in China. While environmental factors like severe weather and sensor saturation from extremely bright fireballs remain challenges, the team is already testing secondary lenses with smaller apertures to recover data from bright events. Future work will focus on expanding the network further and performing deeper physical analyses of meteoroid densities to refine engineering models for spacecraft shielding.

Summary updated

Abstract

Meteors are important phenomenon reflecting many properties of interplanetary dust particles. The study of their origin, mass distribution, and orbit evolution all require large data volume, which can only be obtained using large meteor networks. After meteor networks in Europe and America, we present our designs and upgrades of a proposing network in China. The new designs are mainly aimed for facilitating data gathering process. Each of the newly designed meteor stations now can support up to 4 cameras to cover the full sky. Newer version of meteor station software now works as an integral system, which can streamline the process of detecting, measuring and uploading meteors. We have built a meteor data platform to store, process and display the meteor data automatically. The software and data platform are designed to be easy to learn and use, so it can attract more people to join and operate meteor stations. Four stations are installed as the first phase of the network, and during the operation in 10 months, the network detected 8,683 orbits, and we find that half of the orbits can be related to established meteoroid streams. The statistical analysis of sporadic meteoroids shows a bimodal distribution of the velocities, which coincides with previous studies. The distribution of Tisserand parameters, $T_j$, shows the two peaks at $T_j=0$ and 3, indicating the different orbits of parent bodies (isotropic and ecliptic), which are divided by $T_j=2$. The falling trajectory of a meteorite was also predicted using observational data of the network. We are currently expanding the network, and in the future we will carry out detailed analysis of the key parameters of the distribution of the meteoroids.
BibTeX
@misc{li2024multi,
  author = {Zhenye Li and Hu Zou and Ji Liu and Jiahui Ma and Qingkai Meng and Yongbing Cai and Xin Zhao and Xiangwen Li and Zhoudunming Tu and Baowei Zhang and Ran Wang and S. Wang and Feng Lu},
  title = {{A Multi-station Meteor Monitoring (M$^3$) System. II. system upgrade and a pathfinder network}},
  year = {2024},
  eprint = {2410.08103},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2410.08103},
}
Figure