Investigation of meteors and their impact on the sodium layer in the mesosphere and lower thermosphere
Open Access
- Author:
- Li, Yanlin
- Graduate Program:
- Electrical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 12, 2023
- Committee Members:
- Madhavan Swaminathan, Program Head/Chair
William Bristow, Outside Unit & Field Member
Qihou Zhou, Special Member
Tim Kane, Major Field Member
Julio Urbina, Chair & Co-Dissertation Advisr
Tai-Yin Huang, Major Field Member & Dissertation Advisor - Keywords:
- meteor
atmospheric sodium layer
atmosphere chemistry
radar
radar signal processing
deep learning - Abstract:
- Meteoroids are small interplanetary and interstellar particles that enter Earth's atmosphere and ablate. The observable phenomena associated with meteoroid ablation are commonly referred to as meteors. The plasma surrounding the meteoroid, or the ionized trail left behind by the meteoroid, can be detected by radar. By analyzing the physical characteristics of these meteors, such as their size and velocity, researchers can gain insights into the composition and distribution of meteoroids in our solar system. Radar detection is a valuable tool for studying meteoroids as it provides a relatively low-cost, non-intrusive, and high-resolution method for detecting these elusive objects. This dissertation introduces three radar meteor detection algorithms. The newly developed detection algorithms and signal processing workflows represent a significant advancement compared to existing approaches, demonstrating superior sensitivity and/or automation levels. These innovative methods constitute a substantial improvement in the field, promising to enhance the quality and accuracy of meteor detection and analysis. The first algorithm, the Statistical-based Head Echo Detection algorithm (SHED), employs a probabilistic model utilizing coded long pulse and meticulously defined probabilities to distinguish meteor echoes from noise. The second algorithm, known as AHEAD, is a sensitivity-prioritized method that can detect more than 200 meteors per minute from data obtained from the Jicamarca incoherent scatter radar. The third algorithm presented in this dissertation, MADAME, incorporates state-of-the-art machine-learning computer vision techniques into radar digital signal processing. By utilizing a variety of deep learning architectures, including both supervised and unsupervised learning, MADAME offers a fast yet effective approach to meteor echo detection and analysis. Both AHEAD and MADAME algorithms are capable of performing interferometric analysis on meteor echoes. Notably, AHEAD's variant workflow can resolve interferometric aliasing, resulting in a more accurate determination of the vector direction and speed of the meteors. Overall, these three cutting-edge algorithms demonstrate significant potential in enhancing the accuracy and precision of meteor detection and analysis. Furthermore, this dissertation presents the first sporadic micro-meteor radiant source distribution derived from the observations conducted at the Arecibo Observatory. The findings unveil all six sporadic meteor radiant sources, namely North and South Apex (NA and SA), North and South Toroidal (NT and ST), and Helion and Anti-Helion sources (H and AH). Moreover, this research investigates the impact probability among various orbits. Meteoroids are generally acknowledged as the source of the metal layers in the Earth's upper atmosphere. Meteoroid material input can impact various atmospheric processes and contribute to our comprehension of the physical processes that occur in the upper atmosphere. Sodium is one of the species of the atmospheric metal layer. In this dissertation, we present a new numerical sodium chemistry model, referred to as NaChem, that treats all involved species with continuity equations instead of using the steady-state approximation. The study aims to investigate the relationship between the global meteoroid material input and the seasonal variation of the sodium layers in the MLT by comparing simulation and observational data. The simulation incorporates components of the Whole Atmosphere Community Climate Model (WACCM) for background gas species. Our findings are consistent with the general features of the measurements obtained from the Colorado State University Lidar (CSU) and the Andes Lidar Observatory (ALO). Overall, this dissertation contributes to our understanding of meteor detection, the distribution of their radiant sources, and their impact on the atmospheric sodium layer in the MLT.