Research Overview

I turn geophysical observations into model-ready constraints to improve predictions of ice sheet behavior and reduce uncertainty in poorly understood processes. My work often adapts techniques from other fields to answer glaciological questions, such as introducing the spectral analysis method from seismology to radar sounding. I used the full radar frequency spectrum to estimate ice temperature, its first application in this setting.

Crossing disciplinary boundaries drives new understanding. I led a synthesis that brought together Earth and planetary perspectives on radar attenuation in ice within a shared physical framework. These communities face similar technical challenges but have developed in parallel, and identifying common needs and opportunities across icy worlds is helping to bring them together.

I use geophysical observations, mainly radar sounding, to reveal the physical state of ice sheets from surface to bed. Radar data can capture variations in ice properties and basal thermal and hydrologic conditions, yet quantitative analyses remain uncommon. My work advances these capabilities by applying statistical and machine learning approaches to extract robust physical insight from radar echoes.

For example, I developed a statistical framework to classify frozen and thawed bed conditions beneath the outflow of the Wilkes Subglacial Basin in East Antarctica, along with confidence estimates for each classification. This analysis not only revealed variable basal conditions in a region critical to East Antarctica's stability but also marked the first radar-based assessment of subglacial thermal state in this region. These findings were covered in Stanford Engineering News. By turning radar observations into physically interpretable classifications, this work opens the door to using thermal state maps as direct constraints in ice sheet models.

I specialize in integrating geophysical observations into ice sheet models, using methods such as inverse modeling, data assimilation, and targeted sensitivity analysis to translate measurements into physically meaningful model constraints. This two-way connection is essential: observations are only as powerful as the insight they provide to models, and models are what allow us to project future mass loss, instability, and contributions to sea-level rise. For instance, the basal thermal state has a large influence on ice flow, yet leading ice sheet models give entirely different maps of basal temperature. By coupling radar and other field data with process-based simulations, my work anchors model behavior in observations while also guiding new data collection to the locations that will most improve predictions.

My research in Nature Communications is an example of how targeted modeling experiments can identify which observations will be most valuable for ice flow projections. By testing the sensitivity of Antarctic mass loss to basal thermal state in the Ice-Sheet and Sea-Level System Model (ISSM), I found that in regions where the bed is frozen but close to melting, even a small temperature increase can drastically reduce basal friction, accelerating ice discharge. This study, featured in Stanford Sustainability News, shows how better constraints on basal thermal conditions are key to improving future sea-level rise projections.