Estimating Snowfall Metrics with a Terrestrial Laser Scanner
This thesis examines lidar (light detection and ranging) scans collected from a Terrestrial Laser Scanner (TLS) during varying snowfall events with the goal of determining its capability to make measurements in a degraded visual environment and the feasibility of using it as a meteorological instrument. The ability to estimate visibility, snowfall intensity rate, particle size and velocity of hydrometeors are explored by comparing metrics derived from lidar scans with estimations obtained from an optical disdrometer. Statistics based on return counts, ranges and reflectance values from the TLS measurements of hydrometeors and static targets were used for comparisons and modeling the parameters of interest. Estimated hydrometeor sizes are much smaller than the laser footprint, preventing bulk statistical comparisons from revealing clear correlations. The TLS is capable of estimating hydrometeor velocities when conditions are conducive; however, results are sporadic and selection of returns must be considered as systematic errors yield unreasonable estimations. Regression between TLS metrics and the optical disdrometer estimates for visibility and snowfall intensity proved statistically significant. This indicates that a TLS provides informative measurements during varying atmospheric conditions of snowfall. Results show that TLS has potential to estimate visibility and a snowfall intensity rate, but has difficulty estimating hydrometeor size and speed. Visibility estimations from a laser scanner with a larger range of spatial measurements will be an improvement compared to the optical disdrometer, which samples a static position and extrapolates based on the assumption of atmospheric homogeneity. This would allow the ability to monitor a larger spatial extent with a higher degree of accuracy.