Motion Estimation of Aerosol Hotspots using a Polarization Lidar

Abstract

The accurate detection and spatial characterization of aerosol hotspots are essential for advancing air quality assessment, climate modeling, and environmental hazard mitigation. While conventional lidar systems provide valuable data for aerosol profiling, their reliance on high-power lasers often introduces ocular safety risks and operational constraints. To address these limitations, the Institute of Technical Physics at the German Aerospace Center has developed a polarization-sensitive eye-safe lidar system, which maintains high sensitivity for aerosol pre-detection while operating within safe-exposure limits. This thesis focuses on the integration of this lidar system into a self-referencing scanning setup, enabling precise three-dimensional mapping of the environment through controlled azimuth-elevation maneuvers. A key contribution of this work is the implementation of a cross-correlation based motion estimation algorithm, designed to quantify the dynamic behavior of aerosols from sequential lidar scans. The algorithm incorporates adaptive windowing techniques, robust peak-finding methods, and quality control criteria to ensure operation across varying signal conditions. The experimental system architecture combines the lidar with an altazimuth mount, an Inertial Measurement Unit (IMU) for attitude and heading reference, and a Global Navigation Satellite System (GNSS) receiver for precise geospatial positioning. The performance of the system is evaluated through measurement campaigns, including scans of natural and artificial aerosol plumes over extended ranges. Results demonstrate that the integrated lidar scanning setup achieves high-fidelity aerosol detection, including faint aerosol hotspots, while enabling preliminary aerosol classification via measured depolarization ratios. The synthesis of IMU and GNSS data enables the generation of georeferenced three-dimensional representations of aerosol distributions. Long distance scans of the TV tower in Stuttgart (6.5 km) demonstrate the optical system’s precision and the positioning system’s accuracy. The motion estimation algorithm effectively tracks low-velocity displacements of distant aerosol features, with sequential scans of cloud formations showing good agreement with ERA5 reanalysis data. However, the detection of movement for rapidly morphing structures remains a challenge due to the limitations in scan speed.