Advancement in Snow Depth Estimation and Avalanche Detection from Microwave and Acoustic Remote Sensing Measurements

Abstract

Seasonal snow stored in mountainous regions is a crucial natural resource that accounts for about 70% of the water resources consumed by the western United States and serves as the primary freshwater supply for approximately two billion people worldwide. Consequently, seasonal snow has significant ecological and economic importance for regions dependent on snow for water resources. However, while snow accumulation offers immense benefits, it also creates the potential for snow avalanches, a natural hazard that can cause substantial economic disruption and pose significant safety risk. Observing and quantifying snow parameters using non-destructive methods presents unique challenges, and traditional in-situ observation methods for remote alpine regions are often impractical at high temporal and spatial resolution. This dissertation accessed remote sensing techniques for observing and quantifying snow properties. The remote sensing techniques used include microwave radar, radiometry, and low-frequency acoustics. On the one hand, I quantified the potential of X-band and Ku-band radar for estimating snow depth distribution. Additionally, I evaluated the responsiveness of X, K, and Ka band radiometry for snow depth estimation. The integration of these multi-frequency approaches offers a promising pathway for snow monitoring from space. On the other hand, using recorded infrasound waveforms, an acoustic remote sensing technique, I developed a deep learning categorization model for identifying high-quality signals in recorded infrasound waveforms. The developed model can be deployed in an embedded system installed at an infrasound station to provide event alerts. This approach reduces the time required for event discrimination and permit timely event alerts.